Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The concept of deference in substantive review of administrative decisions in four common law countries Freckelton, Alan 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2013_spring_freckelton_alan.pdf [ 1.55MB ]
JSON: 24-1.0073508.json
JSON-LD: 24-1.0073508-ld.json
RDF/XML (Pretty): 24-1.0073508-rdf.xml
RDF/JSON: 24-1.0073508-rdf.json
Turtle: 24-1.0073508-turtle.txt
N-Triples: 24-1.0073508-rdf-ntriples.txt
Original Record: 24-1.0073508-source.json
Full Text

Full Text

 THE CONCEPT OF DEFERENCE IN SUBSTANTIVE REVIEW OF ADMINISTRATIVE DECISIONS IN FOUR COMMON LAW COUNTRIES   by  Alan Freckelton  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF LAWS in  The Faculty of Graduate Studies (Law)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   January 2013  © Alan Freckelton, 2013        i  ABSTRACT  This thesis examines the concept of “deference” in relation to judicial review of administrative decisions in Canada, and then compares this approach to judicial review to that which exists in the United Kingdom, New Zealand and Australia. Canadian courts have adopted a system of “substantive review” of administrative decisions, at least since 1979 (if not earlier), and will generally show deference to the decision-maker. It is important to note that Canadian courts have interpreted the word “deference” not as subservience (an approach that would make judicial review pointless), but as a form of “respectful attention” to the decision under review. Canadian courts recognise that they do not have a monopoly of wisdom on matters of statutory interpretation, but will step in to set a decision aside when that decision is unreasonable in some sense.  Courts in the United Kingdom have recognised at least since 1987 that the classic standard of Wednesbury unreasonableness – that the decision is “so unreasonable that no reasonable person could have made it” – is not suitable for all kinds of administrative decisions, and have moved to a system whereby there is a “variegated standard” of reasonableness on judicial review for matters not covered by the Human Rights Act 1998, and a proportionality approach for those that are. The law in New Zealand is not as clear, because the Supreme Court has yet to squarely approach the issue, but the lower courts certainly appear to be moving in a similar direction.  However, Australian courts vehemently deny that they show any deference to administrative decision-makers, and Australian academic commentators are equally insistent that such an approach is legally suspect at best and mere obsequiousness to government at worst. This is despite the fact that Australia has always recognised Wednesbury unreasonableness as a ground of judicial review. This thesis attempts to dispel some of the Australian arguments against a deference approach, particularly in relation to s.75 of the Australian Constitution, and concludes that Australia would be best off adopting a form of substantive review of administrative decisions, similar to that which exists in Canada.  ii  TABLE OF CONTENTS  Title  Page   Abstract………………………………………………………………………… i  Table of Contents………………………………………………………………. ii  List of Figures………………………………………………………….............. vi  Acknowledgements…………………………………………………………….. vii   Introduction – Deference and Substantive Review………………………….. 1  “Substantive review”, “merits review” and “review of the merits”…………….. 2  Structure of this thesis…………………………………………………………... 6  1 Basic Principles of Judicial Review…………………………………………... 7 1.1 What is judicial review and why do we need it?................................................... 9 1.1.1 Why does judicial review exist at all?................................................................... 12 The shared common law………………………………………………………… 13 The constitutional baseline set by Marbury v Madison………………………… 15 A consequence of constitutional entrenchment…………………………………. 17 1.2 The “Administrative State”……………………………………………………... 20 1.3 The tension between Parliamentary supremacy and the constitutional role of courts…………………………………………………………………………….  23 1.3.l Parliamentary supremacy……………………………………………………….. 24 1.3.2 The rule of law defined…………………………………………………………. 25 1.3.3 Embodying the tension: privative clauses………………………………………. 28 1.4 Summary………………………………………………………………………... 31  2 Substantive Review in Canada to 1975……………………………………….. 33 2.1 Grounds of judicial review in Canada…………………………………………... 34 2.2 Early exemplars: Canadian Supreme Court jurisprudence prior to 1949……….. 35 2.2.1 Prince Edward Island (Commissioner of Public Lands) v Sulivan……………... 36 2.2.2 Other pre-CUPE cases in which “deference” was shown………………………. 37 Labour tribunals………………………………………………………………… 38 Wartime and national security issues…………………………………………… 40 2.3 The denial of deference in the Supreme Court, 1949-1975……………………..  41  iii  2.4 Prelude to the Canadian decisions – Anisminic Pty Ltd v Foreign Compensation Commission……………………………………………………...  43 2.5 “Complete lack of deference” cases in Canada…………………………………. 47 2.5.1 Metropolitan Life………………………………………………………………... 47 2.5.2 Bell v Ontario Human Rights Commission……………………………………... 51 2.6 Conclusion: the turning of the tide……………………………………………… 54  3 The Return of Deference in Canadian Substantive Review………………… 55 3.1 Canadian Union of Public Employees, Local 963 v New Brunswick Liquor Corp……………………………………………………………………………...  55 3.2 The four standards of review in Canadian administrative law………………….. 62 3.2.1 Flaws in the (pre-Dunsmuir) reasonableness standards………………………… 64 A “sliding scale” of review?.................................................................................. 65 “Weight”………………………………………………………………………… 67 3.2.2 Judicial criticism of the “three standards” approach……………………………. 71 3.2.3 Limits on the principle of deference…………………………………………….. 73 3.3 A new high water mark? Dunsmuir v New Brunswick………………………….. 74 3.3.1 Facts……………………………………………………………………………... 75 3.3.2 Majority judgement……………………………………………………………... 76 3.3.3 Just moving traffic? The judgement of Binnie J………………………………... 81 3.3.4 The “just the facts approach”: Deschamps J……………………………………. 85 3.4 Dunsmuir reviewed……………………………………………………………... 87 3.4.1 What Dunsmuir has achieved…………………………………………………… 87 3.4.2 The “spectrum of reasonableness”……………………………………………… 88 3.4.3 “True questions of jurisdiction or vires”………………………………………... 90 3.4.4 To what should courts defer?................................................................................ 92 3.5 “Substantive review” in Canada defined……………………………………….. 95  4 Deference in the United Kingdom…………………………………………….. 97 4.1 Reasonableness and irrationality: the GCHQ Case……………………………... 98 4.2 Varying the Wednesbury principle: “anxious scrutiny”………………………… 101 4.2.1 Application of the “anxious scrutiny” principle………………………………… 102 4.2.2 Cases where “anxious scrutiny” is not necessary……………………………….. 107 4.2.3 Use of proportionality in pre-Human Rights Act cases?....................................... 109 4.2.4 Indicia of deference – common features and key differences………………….. 110 4.3 The Human Rights Act 1998……………………………………………………. 111 4.4 Irrationality after the Human Rights Act………………………………………... 114  iv  4.4.1 Definition of “rationality”………………………………………………………. 114 4.4.2 Moves towards a single test of judicial review?.................................................... 115 4.4.3 The role of deference……………………………………………………………. 118 4.5 Proportionality…………………………………………………………………... 120 4.5.1 Origins of the principle………………………………………………………….. 120 4.5.2 Differences between the irrationality and proportionality approaches…………. 122 4.5.3 “Merits review”…………………………………………………………………. 125 4.5.4 Deference and the “margin of appreciation”……………………………………. 127 4.5.5 Deference illustrated – the Denbigh High School case…………………………. 129 4.6 Merits review, variable review and proportionality: is there really any difference?.............................................................................................................  132 4.7 Summary………………………………………………………………………... 138  5 Deference in New Zealand…………………………………………………….. 141 5.1 Deference and variable unreasonableness in New Zealand law – the beginnings 143 5.1.1 Wellington City Council v Woolworths New Zealand Ltd (no 2)……………….. 143 5.1.2 Origins of deference and variegated review in New Zealand …………………... 144 5.2 A sudden rush – the Court of Appeal in 1999-2000…………………………….. 146 5.3 A revival of variegated reasonableness? Wolf v Minister for Immigration……... 150 5.4 Increasing intensity of anxiety?............................................................................. 154 5.4.1 Discount Brands………………………………………………………………… 154 5.4.2 Unison Networks Ltd v Commerce Commission………………………………... 158 5.4.3 Ye v Minister for Immigration ………………………………………………….. 160 5.5 Conclusions……………………………………………………………………... 164 6 Deference in Australia 166 6.1 The Enfield decision…………………………………………………………….. 167 6.1.1 Prelude – Chevron………………………………………………………………. 167 6.1.2 Enfield – the facts……………………………………………………………….. 170 6.1.3 High Court decision……………………………………………………………... 173 Majority decision………………………………………………………………... 173 Gaudron J……………………………………………………………………….. 178 Academic comment……………………………………………………………... 179 6.2 Merits review in Australian courts……………………………………………… 180 6.2.1 The judicial power of the Commonwealth……………………………………… 180 6.2.2 “Judicial power” and “merits review”…………………………………………... 183 6.2.3 “Merits review” and “review of the merits” distinguished……………………... 187 Sun v MIEA……………………………………………………………………… 188  v The Guo litigation………………………………………………………………. 190 6.3 Deference and standards of review in Australia………………………………… 193 6.3.1 Jurisdictional facts………………………………………………………………. 194 Definition………………………………………………………………………... 194 M70 v Minister for Immigration………………………………………………… 195 “The Minister is satisfied that …”………………………………………………. 198 6.3.2 Deference on questions of law generally……………………………………….. 201 6.3.3 Expertise………………………………………………………………………… 202 6.3.4 Discretion and fact-finding generally…………………………………………… 207 6.4 Privative clauses in the High Court……………………………………………... 210 6.4.1 Cases from federation to 1945………………………………………………….. 210 6.4.2 Reconciliation between the rule of law and Parliamentary supremacy: the Hickman case…………………………………………………………………….  211 6.4.3 The jurisdictional error qualification: the evisceration of Hickman in S157…… 212 6.4.4 Why the change?………………………………………………………………... 217 6.5 A variable standard of reasonableness review in Australia?................................. 218 6.6 Australia – the height of intervention…………………………………………… 225  7 Conclusions…………………………………………………………………….. 227  7.1 Academic comment opposed to deference in Australia………………………… 227 7.1.1 Deference as a failure to exercise judicial power……………………………….. 227 7.1.2 Deference as obsequiousness to governments…………………………………... 229 7.1.3 Deference as an affront to the rule of law………………………………………. 231 7.1.4 Judicial review that affirms a tribunal decision is mere “deference”…………… 234 7.2 Pro-deference writers…………………………………………………………… 237 7.3 Why the vitriol?..................................................................................................... 241 7.4 The way forward for Australian administrative law……………………………. 247 7.4.1 Problems with the Canadian approach………………………………………….. 248 7.4.2 Problems with the Australian approach…………………………………………. 251 7.4.3 Substantive review in Australia?........................................................................... 252 Substantive review and section 75 of the Constitution…………………………. 253 The lack of an Australian bill of rights………………………………………….. 254   Bibliography 257    vi  LIST OF FIGURES  Figure           Page  Figure 7.1 Chart of judicial decision-making process…………………………..252   for review of administrative decisions in Australia  vii  Acknowledgements  I would like to acknowledge the following people who assisted and supported throughout this task. Without them, it would not have come close to completion.  Firstly, my supervisor, Dr Mary Liston, coped with several drafts of this thesis and innumerable inquiries. She also had to deal with my accent and initial lack of knowledge of Canada, and my inability to even find her office the first time we met. Thanks for everything, especially for helping me shape an initial idea into something that might be of some value. You’re welcome to visit Squamish any time, although I don’t know about actually going into the Ocean Port Hotel.  Secondly, the second reader of this thesis, Dr Hudson Janisch, probably was not sure what he had let himself in for after the first few pages. Thanks for your valuable comments and fitting in our meetings with my bar exams.  Thirdly, the Sub-Dean of the Australian National University’s Graduate Certificate in Migration Law, Marianne Dickie, appreciated my work on this thesis and my other LLM studies in allocating work to me and pointing me to a few interesting Australian cases. Thanks for your patience.  Finally, but by no means least, to my wife Vikki and the five children we have with us in Canada for putting up with me sitting at computer screens for hour after hour. I might note that I also had to put up with four kids reading over my shoulder and the three-year-old sometimes on my lap. Thanks for giving me a reason to keep going, just by being there and being my sounding boards. There is no way I could have done this without you.  Alan Freckelton   1  INTRODUCTION Deference and Substantive Review  Strikingly to Australian lawyers, Canadian administrative law lacks the strict prohibition on “merits review” that exists in Australian law. Since the year 2000, the Australian High Court has twice considered – once in detail and once briefly – whether a North American concept of “deference” to administrative decision-makers should be introduced into Australian law1.  In Corporation of the City of Enfield v Development Assessment Commission 2  the High Court roundly rejected any endorsement of a common law principle of deference, claiming that such an approach involves an abdication of the court’s responsibility, a theme later taken up by commentators, including Hayne J, one of the judges in Enfield 3 .  The Enfield judgement, criticising any notion of deference to administrative decision-makers, was a direct response to the arguments raised by counsel. However, in Minister for Immigration and Citizenship v SZMDS 4 , the issue was raised again, this time seemingly on the volition of Gummow ACJ and Kiefel J. Although the deference approach was rejected again, the concept this time was not dismissed out of hand, and perhaps left the way open for its application in the future.  1  The term “deference” must, by its nature, await a more detailed explanation, but for present purposes, it is sufficient to note that it is concerned with the way that a court approaches the exercise of a statutory power by an administrative decision-maker, whether this decision-maker is a person (such as a Minister, acting either personally or through his or her delegates), or a statutory tribunal. When a court gives a high degree of deference to a decision- maker, it means that it is reluctant to interfere with the administrative decision, and a low degree of deference obviously means the converse. This thesis will provide greater depth to this definition as we proceed. 2  (2000) 199 CLR 135. 3  Justice Kenneth Hayne, “Deference: An Australian Perspective”, [2011] Public Law 75. 4  [2010] HCA 16.  2   “Substantive review”, “merits review” and “review of the merits”  The central argument I will make in this thesis is that the distinction between “merits review” and review for an error of law is ultimately untenable, and that the Canadian approach to substantive review, while it has its own difficulties, is preferable to that which exists in Australia. To that end, this thesis will undertake a comprehensive examination of the concept of deference in Canada, and explain how this approach, despite arguments to the contrary, is consistent with the Australian constitutional framework.  This thesis will analyze the jurisprudential relationship between “substantive review” and “deference”, focusing on Canadian law. The question of the level of deference, if any, to be given by a court to an administrative decision-maker is one that appears at first glance to be a uniquely Canadian approach to the vexed issue of curial review of administrative decisions. Canadian administrative law has included a doctrine of deference to administrative decision- makers on judicial review of administrative decisions since at least the 1979 decision of Canadian Union of Public Employees, Local 963 v New Brunswick Liquor Corp 5 , although the approach may actually have a much longer lineage than that 6 . The deference approach has been restated and updated (at least somewhat) in the seminal case of Dunsmuir v New Brunswick 7 , and this case will be considered in some detail. However, it appears that many common law jurisdictions have attempted to address the same kinds of issues that the Supreme Court of  5  [1979] 2 SCR 227. 6  David Mullan, “The Supreme Court of Canada and Tribunals – Deference to the Administrative Process: A Recent Phenomenon or a Return to Basics?”, (2001) 80 Canadian Bar Review 399. 7  [2008] 1 SCR 190.  3  Canada addressed in Dunsmuir. Why is it that some jurisdictions regularly expressly “defer” to administrative decision-makers and others do not? In the case of those that do not, do they effectively engage in the same kind of judicial analysis by means of alternative legal formulations?  The term “deference” is best described in Dunsmuir v New Brunswick as follows8:  Deference is both an attitude of the court and a requirement of the law of judicial review. It does not mean that courts are subservient to the determinations of decision makers, or that courts must show blind reverence to their interpretations, or that they may be content to pay lip service to the concept of reasonableness review while in fact imposing their own view. Rather, deference imports respect for the decision-making process of adjudicative bodies with regard to both the facts and the law. The notion of deference “is rooted in part in a respect for governmental decisions to create administrative bodies with delegated powers”9 … Deference in the context of the reasonableness standard therefore implies that courts will give due consideration to the determinations of decision makers.  Deference is an approach to judicial review taken by the courts, and effectively acts as a form of reconciliation between the rule of law and Parliamentary supremacy. That is, deference to administrative decision-makers balances the courts’ constitutional requirement to review the decisions of administrative decision-makers to ensure that they are both constitutionally valid and within the decision-maker’s power to make, and the power of the Parliament to allocate certain decision-making powers to persons authorised by or bodies created by statute. As Cromwell J of the Supreme Court, writing extrajudicially, has stated, “[t]he courts recognized that the rule of law did not require a judicial monopoly on adjudication or an unduly expansive  8  Ibid at paragraph 47. 9  Referring to Canada (Attorney General) v Mossop [1993] 1 SCR 554 at 596.  4  approach to jurisdictional error”10, and “[t]he exercise of delegated power by the executive and tribunals has been accepted as normal and the courts have recognized that other decision-makers have central roles to play in interpreting legislation, an exercise that may well admit of more than one legally acceptable answer”11.  Having established what “deference” and “substantive review” mean in Canada, I will then briefly discuss the fact that Canada is not unique in its approach to judicial review of administrative decisions. Both the United Kingdom (UK) 12  and United States of America (USA) 13 , to varying extents and using varying terminology, provide for deference to administrative decision-makers, and New Zealand appears to be moving in that direction. This thesis will focus primarily on Australia, with consideration given to the position in the UK and New Zealand. I will not consider the deference approach in the USA, except where necessary to explain the law in these other countries.  In particular, I will examine the claim made by courts, particularly in the UK and Australia, that they do not engage in “merits review”. Canadian courts do not make this argument, and instead simply focus on whether an administrative decision is “reasonable” or “correct”, depending on the applicable standard of review (another concept that will be discussed more fully later in this thesis). UK and Australian courts would regard this as a form of “merits review”. However, I will argue that UK and Australian courts already engage in review of the merits of a decision,  10  Justice Thomas Cromwell, “From the New Despotism to Dunsmuir: A Funny Thing Happened on the Way to the Apocalypse”, (2011) 24 Canadian Journal of Administrative Law and Practice 285 at 287. 11  Ibid at 289. 12  See in particular R v Secretary of State for the Home Department, ex parte Daly [2001] 2 AC 532. 13  See in particular Chevron USA Inc v Natural Resources Defence Council Inc (1984) 467 US 837.  5  regardless of their protestations to the contrary, especially when one considers the related grounds of review of Wednesbury unreasonableness 14, “variegated” or “sliding scale” unreasonableness, and proportionality.  I will argue that, by permitting review of administrative decisions on the basis of “unreasonableness”, the courts in all four countries are in fact intruding on the merits of administrative decisions, the primary difference between Canada and the other jurisdictions being that Canada does not pretend otherwise. Indeed, in Halifax (Regional Municipality) v Canada (Public Works and Government Services) Cromwell J, writing for the court, stated that “at issue in this appeal are the scope of the Minister’s discretion to determine that value [of certain property], the standard of review applicable to the exercise of this discretion, and the ultimate merits of the Minister’s valuation of the land in this case”15.  I will argue that the only difference between review of the merits of a decision and Wednesbury unreasonableness is the degree of deference afforded to the decision-maker – a difference of degree and not substance. I will also argue that there is a difference between “merits review” and “review of the merits”, and that the latter position describes the current law of judicial review in all four countries, despite the insistence by the UK and Australian courts in particular that they do not undertake such review. The former involves a de novo reconsideration of a case and the substitution of a new decision for that of an administrator and is an executive function, while the latter involves review of a decision for jurisdiction, procedural fairness and reasonableness, with the matter being remitted to a decision-maker in the case of any flaw, and can properly be  14  Referring to Associated Picture Houses Ltd v Wednesbury Corporation [1948] 1 KB 223. 15  [2012] SCC 29 at paragraph 37.  6  regarded as a judicial function. There is a particular convergence in the law in Canada and the UK, with the lower courts at least in New Zealand following suit. Australia, in insisting that there is a clear difference between the legality and the merits of a decision, remains the outlier.  Structure of this thesis  1.  I will first consider a number of issues and concepts that are of importance to judicial review generally, and how they relate to this thesis in particular. These are as follows:  a. The definition of “judicial review”  b. The reasons for the existence of judicial review  c. Remedies available from judicial review  d. The rule of law – deference as a constitutional principle  e. Privative clauses and how they conflict with the rule of law  2. I will then discuss the principle of deference as part of the substantive review “package” in Canada, focusing on the following cases:  a. Metropolitan Life Insurance Co v International Union of Operating Engineers, Local 796 16  b. Bell v Ontario Human Rights Commission17 c. Canadian Union of Public Employees, Local 963 v New Brunswick Liquor Corp18  16  [1970] SCR 425. 17  [1971] SCR 756. 18  Supra n5.  7  d. Dunsmuir v New Brunswick19  3. I will discuss the law of judicial review as it currently exists in the UK, particularly the move from Wednesbury unreasonableness to a “sliding scale” of reasonableness, to proportionality (in cases considering the Human Rights Act 1998 20  at least). I will consider whether the British proportionality approach has any lessons for Canada, given the recent Supreme Court of Canada decision in Doré v Barreau du Québec 21 .  4. I will then consider the law of judicial review in New Zealand. The situation in this country is interesting in that the lower courts appear to be enthusiastic about a variable reasonableness approach to judicial review, but the Supreme Court is not.  5. Finally, I will consider the following issues in relation to Australia.  a. The decision in Enfield, to explain why the deference approach was rejected for Australia.  b. Whether a Canadian kind of substantive review could operate in Australia, with particular attention to s.75 of the Australian Constitution and the definition of the “judicial power of the Commonwealth”. This will require an examination of the  19  Supra n7. 20  1998 c 42. 21  [2012] SCC 12.  8  Australian principle that reviewing courts must never examine the “merits” of an administrative decision, and an explanation of what this means.  c. Whether, despite the frequent denials of the High Court, Australian courts do in fact engage in “merits review”. I will argue that, at the very least, the “Wednesbury unreasonableness” ground of review is a form of review of the merits of an administrative decision, and I will distinguish between “merits review” and “review of the merits” of a decision. I will argue that while the former is an exercise of executive power, the latter is an exercise of judicial power and is not prevented by s.75 of the Constitution.  d. Finally, I will consider the fact that Australian commentators, generally speaking, have not merely rejected the deference concept, but have done so in a particularly vehement manner. I will consider why this might be the case, and argue that such commentators have misunderstood the meaning of the principle.  My primary conclusion will be that Australian (as well as Canadian, UK and New Zealand) courts already engage in review of the merits of administrative decisions, and they would be best off recognising this fact. The courts could then move to a more formalised deference approach similar to Canada, an approach which I will argue is consistent with the Australian constitution.  9  CHAPTER ONE – BASIC PRINCIPLES OF JUDICIAL REVIEW  Before proceeding further, it is necessary to understand some of the basic principles and problems of judicial review. The entire concept of judicial review of administrative decision- making represents a clash between the legislature and the executive on one hand, and the judiciary on the other. It also represents a fundamental problem for the rule of law – how can we reconcile the constitutionally mandated supremacy of Parliament, which creates administrative decision-making bodies and powers, with the role of the judiciary in ensuring that decision- makers remain within the bounds of the powers given to them by statute? This issue becomes particularly vexed when one considers the phenomenon of privative clauses, which is an attempt by the Parliament to prevent courts from reviewing administrative decisions that they otherwise could review.  This chapter will discuss the definition and origins of judicial review, and the constitutional basis for deference to administrative decision-makers. I will argue that deference, as opposed to subservience, to decision-makers is the best way in which the competing factors relevant to the rule of law outlined above can be reconciled.  1.1 What is judicial review and why do we need it?   Perhaps the best place to start is to define what judicial review actually is. WJ Waluchow has defined it as follows 22 :   22  WJ Waluchow, “Judicial Review”, (2007) 2 Philosophy Compass 258 at 258.  10  A practice whereby courts are sometimes called upon to review a law or some other official act of government (e.g. the decision of an administrative agency such as a state or provincial labour relations board) to determine its constitutionality, or perhaps its reasonableness, rationality, or its compatibility with fundamental principles of justice.  It is important to note that judicial review is not new. Judicial review has existed at least since the time of the English prerogative writs, and in the former British colonies of Australia, Canada and New Zealand since the time those States came into existence 23 .  It has always been a part of the common law of those countries, whether this has been acknowledged or not. The ultra vires principle operates, in the case of federal states, to invalidate state or provincial laws that are inconsistent with federal laws 24 , or actions of municipal governments that are inconsistent with their enabling legislation, as well as those of administrative decision-makers.  Bastarache J of the Supreme Court of Canada, writing extrajudically, has defined the term “judicial review”, in terms of review of administrative decisions, as “the means by which the courts supervise those who exercise statutory powers, to ensure that they do not overstep their legal authority”25, quoting paragraph 28 of the judgement in Dunsmuir v New Brunswick26. That paragraph went on to say that “the function of judicial review is therefore to ensure the legality, the reasonableness and the fairness of the administrative process and its outcomes”. From these definitions, we can see that judicial review in administrative law involves the review by one branch of government, the judiciary, of a decision made by another branch of government, the executive. It can therefore be distinguished from an appeal from a lower court. The purpose of  23  For early cases involving judicial review of administrative decisions, see Prince Edward Island (Commissioner of Public Lands) v Sulivan [1877] 1 SCR 3 (Canada) and Potter v Minihan (1908) 7 CLR 277 (Australia). 24  See for example s.109 of the Commonwealth of Australia Constitution Act. 25  Justice Michel Bastarache, “Modernizing Judicial Review”, (2009) 22 Canadian Journal of Administrative Law and Practice 227 at 228, quoting from Dunsmuir, supra n7 at paragraph 28. 26  Supra n7.  11  judicial review is for the court to examine whether the decision-maker had the power to make the decision or grant the remedies in question (or even embark on the decision-making process in the first place), the process by which the decision was reached (for example, failure to afford procedural fairness), or the substance of the decision itself (review on the basis of “reasonableness”, however this term is defined).  De Smith notes two essential principles of true judicial review, as follows. Firstly, the issue of a prerogative writ (a term that will be defined shortly) is never given as of right – instead, the applicant must demonstrate why the writ should be granted 27 . Secondly, the remedies on judicial review are discretionary, meaning that a remedy will not be granted if, for example, there has been delay on the part of the applicant, or the applicant has been guilty of misconduct of some kind 28 .  Judicial review can be contrasted with a statutory appeal. In some cases, an Act itself will provide for an appeal against an administrative decision to a court. For example, s.476 of the Migration Act 1958 provides for an appeal against a decision of the Migration Review Tribunal (MRT) or Refugee Review Tribunal (RRT) to the Federal Magistrates Court. The distinction between judicial review and statutory appeals in Canada is now largely one of form and not substance, after the Supreme Court, in Canada (Citizenship and Immigration) v Khosa 29 , extended the Dunsmuir “standard of review” analysis to statutory appeals. In Australia, the two terms are generally used interchangeably by lower courts. However, the distinction may prove to  27  SA De Smith, “The Prerogative Writs”, (1951-53) 11 Cambridge Law Journal 40 at 42 28  Ibid at 44 29  [2009] 1 SCR 339  12  be of more significance in the High Court, because s.75 of the Australian Constitution guarantees some form of judicial review of administrative decisions, at least in that court.  One matter that is omitted from these definitions is what the courts can do if they find that a decision fails to meet any one of the criteria specified above. In general, a court will have the power to set aside (“quash”) a decision made without jurisdiction, or in breach of requirements of procedural or substantive fairness, and remit that decision to the individual or body concerned for reconsideration. If a court is unable to actually do anything about a decision that falls foul of these principles, one wonders whether it is a true “judicial review”30. The issue of remedies will be discussed in more detail later in this chapter.  1.1.1 Why does judicial review exist at all?   The next question is “why does judicial review of the decisions of administrators exist in the first place”? Why does a court have the authority to overturn a decision of the executive? There are three reasons – the historical prerogative writs and the background of the common law, decisions of the courts themselves, and constitutional entrenchment.   30  This was an argument raised by Hill J (sitting at first instance) in Eshetu v Minister for Immigration and Ethnic Affairs [1997] FCA 19, and by the applicants in Abebe v The Commonwealth; Re Minister for Immigration and Multicultural Affairs, ex parte Abebe (1999) 162 ALR 1.  13 The shared common law  The superior courts themselves of many common law jurisdictions have found that they have the power to set aside an administrative decision on the basis that it was contrary to the constitution, ultra vires the enabling act, resulted from a misapprehension of the decision-maker’s jurisdiction or was unreasonable in some sense.  Modern administrative law in common law countries is based on the old English “prerogative writs”. The writs were really a form of remedy for inappropriate administrative action or inaction as the case may be. The most important of these writs were mandamus, or an order to compel a decision-maker to exercise their jurisdiction; prohibition, which was effectively a form of injunction preventing a decision-maker from hearing a case; certiorari, which was originally “essentially a royal demand for information”31, but became a means by which a court could quash the decision of an administrator and order that the decision be made again, and habeas corpus, literally “produce the body”, which was usually an inquiry as to the legality of the detention of an individual. As can be seen from the name “prerogative writs”, the writs are discretionary in nature, and courts today still retain the ability to refuse relief in situations such as the applicant not coming to the court with “clean hands”32.  The usefulness of the old writs varies from jurisdiction to jurisdiction. At one extreme, they have been abolished in British Columbia, and replaced by the general application for judicial review,  31  De Smith, supra n27 atat 45. 32  See for example Homex Realty Pty Ltd v Wyoming (Village) [1980] 2 SCR 1011.  14  through which a reviewing court may grant relief “in the nature of” the old remedy33. On the other hand, the writs still have significant currency in Australia, at least at the High Court level, because they are specifically provided for the in the Constitution.  It can also be seen that the remedies available on judicial review have substantially influenced the substantive law of judicial review. The question of whether a particular administrative decision should be set aside has influenced the law relating to the grounds on which this can happen. A striking modern example of this problem can be seen in Blencoe v British Columbia (Human Rights Commission) 34 , in which the applicant, a former British Columbia cabinet Minister, sought a permanent stay of sexual harassment proceedings against him because of excessive delay in hearing the matter.  The Supreme Court found that the delay was not so egregious as to warrant a permanent stay of proceedings, and therefore found that he had not been prejudiced by the delay – a clear example of the “tail wagging the dog”, to use the popular phrase 35. The minority’s approach, which was to find that the applicant had been substantially prejudiced by the delay but to decline to grant the order he sought, is preferable and avoids the circularity present in the majority judgement.  It is apparent that Canadian courts have been more inventive in granting remedies than their Australian counterparts. Remedies such as delayed declarations of invalidity 36  are unknown in Australia, and the unwillingness of Australian courts to do anything other than simply invalidate  33  Subsections 2(1) and (2) of the Judicial Review Procedure Act (BC), RSBC 1996 c 241. 34  [2000] 2 SCR 307. 35  For a further explanation of this term in this context, see Cristie Ford, “Dogs and Tails: Remedies in Administrative Law” in Colleen Flood and Lorne Sossin (eds), Administrative Law in Context, Edmond Montgomery Publications (Toronto), 2008. 36  Probably most famously used in Reference re Language Rights under s.23 of the Manitoba Act, 1870 and s.133 of the Constitution Act, 1867 [1985] 1 SCR 721.  15  legislation or set aside a decision immediately perhaps gives the impression that courts do have the “last word” in constitutional and administrative matters in that country. Certainly, the concept of “dialogue” between the courts and the legislature, which is a popular theme for Canadian administrative and constitutional legal scholars 37 , and increasingly so for UK scholars 38 , does not appear to exist in Australia. The constitutional baseline set by Marbury v Madison  The US case of Marbury v Madison has been relied on in Australia as a basis for the existence of judicial review. In that case, Marshall CJ stated as follows: 39   It is emphatically the province and duty of the Judicial Department to say what the law is. Those who apply the rule to particular cases must, of necessity, expound and interpret that rule. If two laws conflict with each other, the Courts must decide on the operation of each … If, then, the Courts are to regard the Constitution, and the Constitution is superior to any ordinary act of the Legislature, the Constitution, and not such ordinary act, must govern the case to which they both apply.  That is, if a law is contrary to the constitution, it can and indeed must be overturned by the courts. The same principle can be applied to administrative decisions – if the decision is in some way contrary to the law, whether that be the constitution or the enabling legislation granting powers to the administrative decision-maker, it can be set aside and a new decision ordered.   37  See for example Rosalind Dixon, “The Supreme Court of Canada, Charter Dialogue and Deference”, (2009) 47 Osgoode Hall Law Journal 235. 38  See for example Eliza Kaczynska-Nay, “The Human Rights Act 1998: The UK ‘Dialogue’ Model of Human Rights Protection”, Australian National University National Europe Centre Briefing Paper Series, July 2009. 39  (1803) 5 US (1 Cranch) 137 at 177.  16  Marshall CJ identified a number of justifications for judicial review in his judgement. These include the written text of the constitution, which Marshall saw as ordinary law, although “supreme” over all other ordinary laws40; the constitutional role of judges, which entails an ability to enforce their interpretations of the constitution against other branches of government; the grant of jurisdiction to the court for “all cases arising under the Constitution”41; the fact that judges and the President were required to swear an oath which, in part, required them to uphold the Constitution 42 ; and the views of the framers.  Marbury v Madison, despite being a decision based on the construction of the US constitution, has been enthusiastically supported in Australia. In Attorney-General (Western Australia) v Marquet the High Court stated as follows 43 :  Unlike Britain in the nineteenth century, the constitutional norms which apply in this country are more complex than an unadorned Diceyan precept of parliamentary sovereignty. Those constitutional norms accord an essential place to the obligation of the judicial branch to assess the validity of legislative and executive acts against relevant constitutional requirements. As Fullagar J said, in Australian Communist Party v The Commonwealth 44, “in our system the principle of Marbury v Madison is accepted as axiomatic”. It is the courts, rather than the legislature itself, which have the function of finally deciding whether an Act is or is not within power.  Marbury v Madison has not had the same level of impact in other common law countries. This case has only been cited three times by the Supreme Court 45 . In Law Society of Upper Canada v Skapinker Estey J, writing for the court, noted that, in relation to the then-new Charter, “[t]he  40  Article VI, Clause 2 of the US Constitution. See also Mark Tushnet, “Marbury v Madison Around the World”, (2003-4) 71 Tennessee Law Journal 251 at 258. 41  Article III, Section 2 of the US Constitution. 42  Supra n39 at 180. 43  (2003) 217 CLR 545 at paragraph 66. 44  (1953) 83 CLR 1 at 262. 45  Search of the Canlii Supreme Court database ( undertaken on 10 December 2012  17  courts in the United States have had almost two hundred years experience at this task and it is of more than passing interest to those concerned with these new developments in Canada to study the experience of the United States courts”46. However, the Supreme Court clearly decided that it did not need the guidance of Marbury for very long, because that case has not been referred to since the 1986 decision of MacDonald v City of Montreal 47 . Instead, Canada has preferred to rely on “home grown” authority such as Crevier v Quebec (Attorney-General)48 as the basis for judicial review. A consequence of constitutional entrenchment  This brings us neatly to the final reason for judicial review, which is that it is enshrined in some jurisdictions’ constitutions. The best example of this is s.75 of the Constitution Act 1900 (Australia), which provides as follows:  In all matters: (i)  arising under any treaty; (ii)  affecting consuls or other representatives of other countries; (iii)  in which the Commonwealth, or a person suing or being sued on behalf of the Commonwealth, is a party; (iv)  between States, or between residents of different States, or between a State and a resident of another State; (v)  in which a writ of Mandamus or prohibition or an injunction is sought against an officer of the Commonwealth; the High Court shall have original jurisdiction.  The writ of certiorari was omitted from s.75(v) of the Constitution. However, the High Court has found on multiple occasions that it has the power to make an order of certiorari that is ancillary  46  [1984] 1 SCR 357 at paragraph 12 47  [1986] 1 SCR 460 at paragraph 187 48  [1981] 2 SCR 220 at 234.  18  to an order for mandamus or prohibition 49 . Also, ss.30 and 32 of the Australian Judiciary Act 1901 give the High Court the power to make such orders in specified circumstances. Section 75 has been hugely significant in Australian administrative law and will be discussed in more detail later in this thesis.  Whether a right to judicial review is constitutionally enshrined in Canada is an interesting question, but the prevailing view is that it is. Even though the jurisdiction of the Supreme Court is not prescribed by the Constitution, there is some separation between the executive and the judiciary. Section 96 does provide some recognition of the superior courts of the provinces, as follows:  The Governor General shall appoint the Judges of the Superior, District, and County Courts in each Province, except those of the Courts of Probate in Nova Scotia and New Brunswick.  The Supreme Court held as follows in Crevier v Quebec (Attorney-General) 50 :  In my opinion, where a provincial Legislature purports to insulate one of its statutory tribunals from any curial review of its adjudicative functions, the insulation encompassing jurisdiction, such provincial legislation must be struck down as unconstitutional by reason of having the effect of constituting the tribunal a s.96 court.  In other words, if the Canadian Parliament attempted to pass a law that rendered decisions of an administrative body immune from judicial review, the Supreme Court could well find that this would effectively confer judicial power on a non-judicial body. Crevier can therefore be  49  See for example Re Refugee Review Tribunal; Ex parte Aala (2000) 204 CLR 82. 50  Supra n48 at 234.  19  regarded as a Constitutional guarantee of a right to some kind of judicial review of administrative decisions.  While Crevier concerns provincial legislation, the Supreme Court appears to have gone further in its decision in MacMillan Bloedel v Simpson 51 , in which a law removing powers to punish for contempt from the superior courts when a young offender was involved was struck down. Lamer CJ in MacMillan cited Crevier as authority for the proposition that core jurisdiction could not be removed from superior courts. More recently, in Dunsmuir v New Brunswick the Supreme Court stated as follows 52 :  The inherent power of superior courts to review administrative action and ensure that it does not exceed its jurisdiction stems from the judicature provisions in ss. 96 to 101 of the Constitution Act, 1867: Crevier. As noted by Beetz J. in Syndicat national des employés de la commission scolaire régionale de l’Outaouais v. U.E.S., local 298, [1988] 2 S.C.R. 1048 (S.C.C.) at p. 1090, “[t]he role of the superior courts in maintaining the rule of law is so important that it is given constitutional protection”. In short, judicial review is constitutionally guaranteed in Canada, particularly with regard to the definition and enforcement of jurisdictional limits.  The meaning of “jurisdictional limits” will be discussed in more detail later in this thesis.  Finally, it could also be argued that s.7 of the Charter 53  confers a constitutional right of judicial review of administrative decisions, at least where the “life, liberty and security of the person” of the applicant are potentially affected. It is clear that s.7, like the rest of the Charter, can be used  51  [1995] 4 SCR 725. 52  Supra n7 at paragraph 31. 53  The Canadian Charter of Rights and Freedoms is set out in Part 1 of the The Constitution Act, 1982, being Schedule B to the Canada Act 1982 (UK), 1982, c 11.  20  to overturn an administrative decision 54  as well as finding an Act to be beyond the competence of the Federal or a Provincial parliament.  1.2 The “Administrative State”  The “administrative state” is a concept that is frequently referred to but curiously rarely defined in the literature, at least by lawyers. Indeed, Alan Cairns tells us that “the modern administrative state defies simple description”, and that “the attempt to pin down the contemporary administrative state is doomed to failure”55. Nevertheless, Seymour Wilson and Onkar Dwivedi make the following attempt 56 :  The administrative state denotes the phenomenon by which state institutions influence many aspects of the lives of citizens, especially those aspects which relate to the economic and social dimensions. It describes a system of governance through which public policies and programs, affecting all aspects of public life, are influenced by the actions of public officials.   The “administrative state” has been described as a fourth branch of government for which is there is no provision in the [US] constitution 57 , while others have decried the perceived ability of Parliament and / or Congress to pass very broad-brush legislation and then allow administrative bodies to fill in the detail and thereby effectively create law 58 . It is now impossible for the average citizen to avoid interaction with the state, and in many cases they simply could not  54  See for example Multani v Commission Scolaire Marguerite-Bourgeoys [2006] 1 SCR 256. 55  Alan Cairns, “The Past and Future of the Canadian Administrative State”, (1990) 40 University of Toronto Law Journal 319 at 322. 56  Onkar Dwivedi and Seymour Wilson, “Introduction” in Onkar Dwivedi (ed), The Administrative State in Canada: Essays in Honour of JE Hodgetts, University of Toronto Press, 1982 at 5. 57  See for example Richard A Epstein, “Why the Modern Administrative State is Inconsistent with the Rule of Law”, (2008) 3 New York University Journal of Law and Liberty 491. 58  Robert Moffit, “Reining in the Administrative State”, 16 February 2011.  21  survive without it. This is not to argue that such regulation is always and everywhere a bad thing, but there can be no denying the increasing pervasiveness of the state into the everyday life of the individual. In the Canadian context, the following passage from Cory J, writing for the court in Newfoundland Telephone Company v Newfoundland (Board of Commissioners of Public Utilities) captures the theme well 59 :  Administrative Boards play an increasingly important role in our society. They regulate many aspects of our life, from beginning to end. Hospital and medical boards regulate the methods and practice of the doctors that bring us into this world. Boards regulate the licensing and the operation of morticians who are concerned with our mortal remains. Marketing boards regulate the farm products we eat; transport boards regulate the means and flow of our travel; energy boards control the price and distribution of the forms of energy we use; planning boards and city councils regulate the location and types of buildings in which we live and work. In Canada, boards are a way of life. Boards and the functions they fulfill are legion.   The dependence by individuals on the state is another reason why it is now impossible to dispense with judicial review of administrative decisions altogether – the consequences to an individual of an unreasonable or even simply incorrect decision can be disastrous. The courts therefore have a crucial role to play in ensuring that individuals can obtain government services that are no longer a privilege for most people, but necessary for economic survival. However, the courts face competing public policy objectives in that the government, the elected representatives of the people, have decided to give the power to make the decision in question to an administrative decision-maker, and more pragmatically that government decision-makers need to be able to do their work without being constantly under constant close judicial scrutiny. It should be noted that judicial review has real impacts on administrative agencies. Some of this impact is beneficial. For example, after the decision in Baker v Canada (Minister for  59  [1992] 1 SCR 623 at 637.  22  Immigration and Citizenship) 60 , Citizenship and Immigration Canada (CIC) has had to produce much more carefully reasoned decisions than the offensive generalisations that formed the “reasons” in that case. On the other hand, the same processes have probably helped lead to the use of bland “boilerplate” or “standard” paragraphs, that are now common features of the reasons of administrative decision-makers. While courts have generally rejected the proposition that the use of standard paragraphs is either unreasonable or an error of law in itself 61 , it may be an unfortunate trend in administrative decision-making. Courts also rarely seem to have much appreciation of the financial impacts of their decisions 62 .  Another distinctive feature of the modern administrative state is, of course, the administrative tribunal. A tribunal is not a court, and is indeed often required by legislation not to act like one. For example, s.420(1) of the Migration Act 1958 (Australia) requires the Refugee Review Tribunal (RRT) to conduct itself in a manner that is “just, fair, economical, informal and quick”63, and s.420(2) provides that it is “is not bound by technicalities, legal forms or rules of evidence” and “must act according to substantial justice and the merits of the case”64. I do not intend to address the details of the operations of tribunals in this thesis, but it will suffice to note that tribunals are creatures of the executive, not the judiciary, despite the fact that their  60  Supra n135. 61  In Minister for Immigration and Citizenship v SZQHH [2012] FCAFC 45, the Full Federal Court of Australia rejected the argument that the use of standard paragraphs in a decision record refusing an applicant refugee status amounted to a reasonable apprehension of bias. 62  See Moore v British Columbia (Education) [2012] SCC 61. See also Singh v Minister of Employment and Immigration [1985] 1 SCR 177 and the comments of Peter Hogg QC on this decision at Hogg, Constitutional Law of Canada (Student edition 2009), Carswell, 2009 at Part 47.4(b), pp 1075-1076. 63  Many of these goals are obviously contradictory – see Minister for Immigration and Multicultural Affairs v Eshetu (1999) 197 CLR 611. 64  Eshetu (supra) found that this was merely an exhortatory provision and did not impose legal duties on the RRT. However, the Full Federal Court has since found in Minister for Immigration and Citizenship v Li [2012] FCAFC 74 that a failure to grant an adjournment in a Migration Review Tribunal (MRT) hearing was a breach of s.353 of the Migration Act 1958 (which is in similar terms to s.420) in the facts of that case, and amounted to a jurisdictional error.  23  procedures may mimic those of courts to a greater or lesser extent 65 . This may partly explain the apparent hostility or suspicion exhibited by courts towards tribunals in the years immediately after WWII – the tribunals not only seemed to be encroaching on the courts’ “territory”, but in some cases seemed as if they were (falsely) representing themselves as courts as well. This has led to what legal scholars describe as the inevitable tension between parliamentary and constitutional supremacy.  1.3 The tension between “Parliamentary supremacy” and the constitutional role of courts  In decided cases and literature on judicial review of administrative decisions, one frequently encounters appeals to the concepts of “Parliamentary supremacy” and “the Rule of Law”. Bastarache and LeBel JJ manage to appeal to both in the one paragraph in Dunsmuir v New Brunswick, by stating that “[i]n addition to the role judicial review plays in upholding the rule of law, it also performs an important constitutional function in maintaining legislative supremacy … In essence, the rule of law is maintained because the courts have the last word on jurisdiction, and legislative supremacy is assured because determining the applicable standard of review is accomplished by establishing legislative intent”66. It is now necessary to explain what these terms mean before proceeding further.    65  See for example on this point Ocean Port Hotel Ltd v British Columbia (General Manager, Liquor Control and Licensing Branch) [2001] 2 SCR 781. 66  Supra n7 at paragraph 30.  24  1.3.1 Parliamentary supremacy  The most fundamental legal reason for the courts to take the decisions of administrative decision- makers seriously is that such decision-makers only exercise power given to them by Acts of Parliament. That is, the creation of administrative tribunals and the proliferation of the administrative state is an exercise of the legislature, another arm of government. As noted above, the Supreme Court stated in Dunsmuir v New Brunswick that judicial review “performs an important constitutional function in maintaining legislative supremacy”67 – that is, the court recognises that it too is subject to the rule of law.  The basic concept of “Parliamentary supremacy” seems to be unexceptionable in the literature, but what precisely it means and how it is to be observed is another matter altogether. David Kinley reminds us that Dicey spoke of “Parliamentary sovereignty” rather than “Parliamentary supremacy”, and states that68:  To possess legislative sovereignty (or, as Dicey and others sometimes prefer, absolute legislative supremacy), a Parliament is subject to no legal limitations in its exercise of that power. A Parliament, on the other hand, which is said to possess legislative supremacy (that is something less than absolute legislative supremacy) is guaranteed only a superior claim against any other body claiming legislative competence, and not necessarily that in its exercise of legislative power it is not subject to any legal limitations.  The result of this distinction is that the Federal Parliaments of countries such as Canada and Australia, which have a written constitution and enumerated powers granted to the Federal  67  Ibid. 68  David Kinley, “Constitutional Brokerage in Australia: Constitutions and the Doctrines of Parliamentary Supremacy and the Rule of Law”, (1993-4) 22 Federal Law Review 194 at 197.  25  Parliament, can never be said to be “sovereign” or possess “absolute legislative supremacy”. It is more accurate to say that a law is valid until the highest court of the land decrees otherwise 69 . In federal systems such as Canada and Australia, there is also the point that federal and provincial or State legislation frequently interact with each other, and while federal legislation might prevail in the event of a direct inconsistency, neither level of government can be said to be “supreme” over the other. Despite these considerations, I will use the term “Parliamentary supremacy” in this thesis, as it is a term that is generally well understood and frequently used.  1.3.2 The rule of law defined  The concept of Parliamentary supremacy is, of course, only one aspect of the somewhat nebulous doctrine referred to as the “rule of law”. At its simplest, the rule of law is the principle that those that make the law are bound by it in the same way as everyone else. Peter Hogg and Cara Zwibel offer the following on AV Dicey’s concept of the rule of law70:  Dicey offered three meanings of the rule of law. First, the rule of law meant the supremacy of “regular law” over “arbitrary power”. This would be unobjectionable were it not for Dicey’s assumption that the discretionary powers of tribunals and officials were examples of “arbitrary power”, while the powers of judges simply reflected the “regular law”. Dicey’s second meaning of the rule of law was “equality before the law”, by which he meant that officials were bound by the same laws as ordinary subjects … The third meaning of the rule of law was that the constitution is a product of the common law, not of any constitutional instrument.   69  PH Lane goes so far to say that “the documents of Commonwealth Parliament do not carry their own imprimatur as do High Court decrees; its legislative Acts remain inchoate until they pass the scrutiny of the High Court whose judgments by contrast do not await legislative approval”. See Lane, “Judicial Review or Government by the High Court?” (1966) 5 Sydney Law Review 203 at 205. 70  Peter Hogg and Cara Zwibel, “The Rule of Law in the Supreme Court of Canada”, (2005) 55 University of Toronto Law Journal 715 at 715-6.  26  The same authors then offer their own interpretation of the concept 71 :   We propose three elements to the rule of law: (1) a body of laws that are publicly available, generally obeyed, and generally enforced; (2) the subjection of government to those laws (constitutionalism); and (3) an independent judiciary and legal profession to resolve disputes about those laws.  This is the central conflict in judicial review of administrative decisions. On one hand, there is an elected Parliament that has the constitutional power to pass laws. That Parliament has the power to create administrative decision-making bodies, sometimes in order to have access to experts in a particular field (particularly the case in economic and competition tribunals, or medical or legal disciplinary tribunals), and sometimes because the Minister could simply not personally make all the decisions that need to be made (particularly the case in social security and immigration tribunals). On the other, courts have the role of ensuring that the decision-makers’ enabling legislation is constitutionally valid, and that the decisions made must conform to both the constitution and the enabling legislation. The rule of law requires that the Parliament makes laws, the courts interpret and apply them, and neither can dictate to the other.  There are basically two possible responses to these principles:  1. Administrative decision-makers are given their powers by legislation. It is the task of the courts to ensure that the decision-maker exercises only those powers that they are specifically given. This means that the words of the enabling legislation must be scrutinised very carefully to ensure that the decision-maker is acting within his, her or its  71  Ibid at 718.  27  jurisdiction, and has asked himself, herself or itself the “correct question” in exercising the relevant powers.  2. By conferring a power on an administrative decision-maker, the Parliament has indicated its intention that a particular kind of matter be governed by decisions of the executive, and not (at least not directly) by the judiciary. Deference to administrative decisions “is rooted in part in a respect for governmental decisions to create administrative bodies with delegated powers”72. The role of the courts is to ascertain the intention of Parliament in conferring the power in question, and while decision-makers must act within the scope of that power, the courts should not interfere with the decision of the administrator unless he, she or it has (a) clearly acted outside the scope of the power they were given, and / or (b) has exercised that power in a manner that is in some sense unreasonable.  The first approach can be regarded as a strict jurisdictional approach to judicial review, and the second as a deference approach. The second approach can be seen as a means by which the courts balance their constitutional role in ensuring that administrative decision-makers act within their statutory powers – within their jurisdiction – while at the same time respecting the Parliament’s decision and intention to give certain powers to administrative decision-makers rather than the courts. Looking at it another way, the second approach leads to deference, while the first is a challenge to deference.    72  Mossop, supra n9 at 596.  28  1.3.3 Embodying the tension: privative clauses  The clearest clash between the principles of Parliamentary supremacy and the rule of law can be seen in the courts’ approach to privative clauses. At its simplest, a privative clause is a “clause in regulatory legislation prohibiting review of decisions by courts”73. In fact, there are two kinds of privative clause, and both frequently appear in the same legislative provision. These are the “finality clause”74, which protects an administrative decision from further internal review, and the ouster clause 75 , which purports to protect the decision from judicial review. Other varieties of privative clause may remove the remedial powers of the court, while still others may determine the kind of evidence that is final and conclusive.  An example of the finality and ouster clause being condensed into a single provision can be seen in s.474 of the Migration Act 1958 (Australia). Subsection 474(1) provides as follows:     73  Peter Bowal and Carlee Campbell, “Key Concepts: Administrative Law from A to Z”, (2007) 31 Law Now 31 at 34. 74  A good example of a finality clause in Canadian legislation can be found in ss.114 and 116 of the Labour Relations Act (Ontario) SO 1995, c 1. Subsection 114(1) provides as follows:  The Board has exclusive jurisdiction to exercise the powers conferred upon it by or under this Act and to determine all questions of fact or law that arise in any matter before it, and the action or decision of the Board thereon is final and conclusive for all purposes, but nevertheless the Board may at any time, if it considers it advisable to do so, reconsider any decision, order, direction, declaration or ruling made by it and vary or revoke any such decision, order, direction, declaration or ruling.  75  The Labour Relations Act (supra) also includes an ouster clause in s.116:  No decision, order, direction, declaration or ruling of the Board shall be questioned or reviewed in any court, and no order shall be made or process entered, or proceedings taken in any court, whether by way of injunction, declaratory judgment, certiorari, mandamus, prohibition, quo warranto, or otherwise, to question, review, prohibit or restrain the Board or any of its proceedings.  29  A privative clause decision: (a)  is final and conclusive; and (b)  must not be challenged, appealed against, reviewed, quashed or called in question in any court; and (c) is not subject to prohibition, mandamus, injunction, declaration or certiorari in any court on any account.  Evans et al identify three main reasons why privative clauses are included in legislation 76 . Firstly, applications to the courts would otherwise result in protracted delays, which could impact adversely on all parties. Secondly, there is the issue of expense, which is particularly an issue in labour arbitration cases, where the resources of the parties are often mismatched. One may have a large and wealthy employer taking action against marginalised employees, or a small business fighting a nation-wide union.  Thirdly, a tribunal may have been created for the very purpose of keeping a dispute out of the courts, which may be for one of the two reasons outlined above, or because members of the tribunal are appointed because of a specialist expertise lacking in the courts (which is especially the case in labour or economic tribunals). This concept of expertise is, as will be seen, a very important one in deciding the degree of deference to be shown to the determinations of an administrative decision-maker.  There is an obvious conflict to be resolved by a court when it is confronted with a privative clause. Evans et al comment as follows 77 :  A statutory direction that the decisions of a particular tribunal are not to be questioned or reviewed in any legal proceeding whatsoever challenges the pervasive assumption that it is ultimately the constitutional function of an independent judiciary to determine the rights of individuals according to law. In particular, it is ultimately for the courts, not administrative agencies, whose  76  John Evans, Hudson Janisch, David Mullan and RCB Risk, Administrative Law: Cases, Text and Materials, Edward Montgomery Publications (Toronto), 2003 at 703-4. 77  Ibid at 702-3.  30  members may not be lawyers or who may be appointed for a relatively short term, to interpret statutes and apply them to the facts of the individual case, and to ensure that administrative decisions are made on the basis of a procedure that meets minimum standards of fairness.  This is also a problem for the rule of law, as it is an attempt by the legislature (undoubtedly acting at the behest of the executive) to exclude the judiciary from its role as the final arbiter of the law. Mary Liston notes that “the risk to the accountability function of the rule of law was that these officials could behave as a law unto themselves, because they would be the sole judges of the substantive validity of their own acts”78. A number of Australian commentators, as will be seen, have taken the view that all privative clauses, or at least those that protect errors of law made by administrative decision-makers, are contrary to the rule of law.  As we have already seen, the Supreme Court in Crevier v Quebec (Attorney-General) 79  has found that it is not constitutionally possible for a privative clause to completely exclude judicial review. While the reasoning in Crevier is undoubtedly strained in places, it is difficult to disagree with the result. What is now clear is that Canadian courts will peruse a decision more carefully and are more likely to decide it on the “correctness” basis if there is no privative clause than if there is a privative clause. In Australia, where s.75 of the Constitution explicitly provides for a right to at least some judicial review of Commonwealth decisions, the same conclusion has been reached – originally in R v Hickman, ex parte Fox and Clinton80, a decision that was broadly sympathetic to privative clauses, and more recently in Plaintiff S157/2002 v  78  Mary Liston, “Governments in Miniature: The Rule of Law in the Administrative State” in Collen Flood and Lorne Sossin (eds), Administrative Law in Context, Edmond Montgomery Publications (Toronto), 2008 at 104. 79  Supra n43. 80  (1945) 70 CLR 598.  31  Commonwealth of Australia 81 , which upheld the constitutional validity of s.474 of the Migration Act, but then proceeded to render it of almost no effect. These cases will be discussed in more detail later in this thesis.  1.4 Summary  In this chapter I have defined key terms relevant to judicial review, and identified the crucial conflict that judicial review poses for the modern administrative state. On one hand, Parliament has made a positive decision to allocate decision-making power to a specified body or tribunal, and on the other the courts have a constitutional role in ensuring that such persons or bodies exercise only that power that they are given by Parliament. This conflict is particularly marked when the statute conferring that decision-making power includes a privative clause, which is the clearest possible indication that a court is not to interfere with the decisions of the administrative body.  However, as we have seen, courts will (and must) enforce the clear principle that a decision- maker may not exercise any power other than that which he, she or it is given by statute. That is, at the very minimum, errors going to the decision-maker’s jurisdiction cannot be tolerated by the courts. To that end, deference can be seen as a constitutional principle, whereby courts respect the will of the Parliament and ensure that the decision-maker acts within the scope of the power given to them.   81  (2003) 211 CLR 476.  32  In the next chapter, we will see how Canadian courts from WWII to at least 1975 at first took a very wide view of what constitutes “jurisdiction”, and how that position has changed since 1975 to a position where administrative decision-makers will be afforded deference to their views more often than not. From this discussion, we will be able to see what is meant by the Canadian term “substantive review”, and this will form a basis for comparing these principles to substantive review principles in the UK and New Zealand, and why the deference approach has been rejected in Australia.    33  CHAPTER TWO – SUBSTANTIVE REVIEW IN CANADA TO 1975  In this chapter, we will consider the law of substantive review and the principle of deference as it applies in Canada. I will argue, from considering a number of kinds of cases, that the deference principle has existed right from the start of Canadian administrative law, although it tended to be applied more in some areas than others until 1975. There appeared to be a reaction against affording deference to decision-makers in the period from WWII to 1975, in which courts gave little deference to interpretations not only of enabling statutes, but also legislation such as the Canadian Bill of Rights 82 . In interpreting the grant of statutory decision-making powers, courts have always been guided by the principle of legislative intent, although the way in which such intent has been divined has varied over time.  This chapter will then focus on two Canadian cases that took a very wide view of the jurisdiction of an administrative decision-maker, Metropolitan Life Insurance Co v International Union of Operating Engineers, Local 796 83  and Bell v Ontario Human Rights Commission 84 , as well as a UK case that was heavily influential in these decisions, Anisminic Pty Ltd v Foreign Compensation Commission 85 . Consideration of the issues raised by these cases will allow us to consider what the Supreme Court was attempting to prevent by moving to a deference-based approach from 1975.    82  SC 1960, c 44. 83  Supra n16. 84  Supra n17. 85  [1969] 2 AC 147.  34  2.1 Grounds of judicial review in Canada  Canadian administrative law recognises four grounds for judicial review 86 . These are constitutionality, especially in cases where an administrative decision is said to conflict with the Charter of Rights and Freedoms; procedural unfairness, which includes review on the grounds of actual or apprehended bias, or failure to provide a hearing; illegality, which involves an allegation that the decision-maker did not have the power to make the decision in question, and unreasonableness, meaning that the decision is alleged to be made without consideration of evidence, based on irrelevant considerations, made without examination of relevant considerations, and so on. In Canada, a decision does not have to be “so unreasonable that no reasonable decision-maker could have reached it”87 before being set aside on this ground, but a reviewing court will often have to give some deference to the decision-maker in coming to its decision. Cases such as those involving abuse of power or discretion could be regarded as one kind of unreasonableness, or perhaps as a fifth ground of review.  The term “substantive review” covers the third and fourth points made above. In short, substantive review is about looking at the substance of the decision, not just the procedures by which it was made. This will involve examination of the decision-maker’s enabling legislation and the reasons for the decision itself, in order to establish whether the decision-maker’s interpretation and application of his or her powers was reasonable. Often a court will be required to defer to some degree to the decision-maker’s interpretation of his or her powers in making an  86  Evans et al, supra n76 at 25-27. 87  Wednesbury, supra n14.  35  assessment of the reasonableness of a decision. Failure to do so could risk the court appearing to illegitimately “second-guess” the authorised decision-maker.  The best way to examine the meaning of the term “deference” in this context is to examine some Canadian cases that preceded the modern application of the deference period, and some of the more important decisions that have explained the meaning of and applied the term.  2.2 Early exemplars: Canadian Supreme Court jurisprudence prior to 1949   The common conception of the state of administrative law in Canada prior to the decision in CUPE is that the Supreme Court was markedly hostile to administrative decision-making, and had little or no time for the abilities in legal reasoning, or indeed even in fact-finding, of administrators. It is generally thought that prior to CUPE, most judges were subscribers to AV Dicey’s theory of constitutional and administrative law, which was that Parliament was supreme, and that the role of the courts was to ensure that administrative decision-makers acted strictly within the powers given to them by the legislature 88 . Dicey was particularly concerned with the exercise of discretionary powers by decision-makers, which he regarded as arbitrary and a usurpation of the role of both the legislature and the judiciary, and he called on judges to interpret legislation providing for such powers as narrowly as possible 89 . The result of this is that all administrative decisions were to be scrutinised very carefully by the judiciary to ensure that they did not step outside the statutory powers that they exercised. One could argue that in theory this is an unexceptional proposition, but in practice it often meant that courts would regularly  88  AV Dicey, Introduction to the Study of the Constitution, 10th ed. (London: Macmillan and Co, 1959). 89  For a good summary of Dicey’s conception of the rule of law, see Hogg and Zwibel, supra n70.  36  substitute their opinions for that of the delegated decision-maker. Some of the ways in which this was done will be discussed in more detail shortly.  2.2.1 Prince Edward Island (Commissioner of Public Lands) v Sulivan  As David Mullan has pointed out 90 , many of the very early decisions of the Supreme Court, while never using terms such as “deference” or “standard of review”91, actually exhibited a sophisticated understanding of the role and purpose of administrative decision-making. For example, the very first case in the Supreme Court that could be described as an administrative law case, decided in 1877, Prince Edward Island (Commissioner of Public Lands) v Sulivan 92 , involved a compulsory acquisition of land, an independent statutory office-holder, and a privative clause, all elements that would be familiar today. The key issue in the case was whether a landholder affected by an acquisition order could seek review of the decision at common law, where the statute in question provided for a statutory appeal within 30 days of the decision being made, and barred all other avenues of appeal.  The Supreme Court found that the privative clause was effective to remove all avenues of review other than that provided for by the statute. The applicant had not applied for review within the 30 day period, and “she must therefore abide by the consequences”93. Richards CJ stated that “I have not met with any case where special provision was made for the correction of the errors or  90  Mullan, “Deference to the Administrative Process”, supra n6. 91  In fact, the term “administrative law” itself was not used by the Supreme Court in an actual judgement until the 1975 case of Law Society of Upper Canada v French [1975] 2 SCR 767 at 771. See Mullan, ibid at 405. 92  Supra n23. 93  Ibid at 43.  37  omissions of the tribunal created by the Statute, and where the privative enactment was so strong and emphatic as it is in this Statute, when the Court has felt justified in setting aside the award of the inferior tribunal”94, a comment which could easily be taken from a 21st-century judgement (as will be seen).  Note the clear “deference” to statutory intention in this passage. Mullan comments on this judgement as follows 95 :  In sum, this very first “Administrative Law” case is one in which the Court was unanimous in its determination to give effect to legislative intention, to respect the administrative and remedial processes established by the legislation, and to have regard to public interest reasons for disregarding normally predominant private rights. In short, it is remarkably “modern” and comparable to what today would be seen as a “pragmatic and functional” approach to the scope or extent of judicial review.   2.2.2 Other pre-CUPE cases in which “deference” was shown  Certain categories of decision-makers are generally accorded deference because of their status in government, such as the Cabinet or a minister; the nature of the power they wield (such as prerogative or emergency powers); or a particular history. In this section, I will examine some of the more common situations in which deference was shown to administrative decision-makers prior to CUPE.    94  Ibid at 61. For a more detailed explanation of this case, see Mullan, “Deference to the Administrative Process”, supra n6 at 405-9. 95  Mullan, ibid at 410.  38 Labour tribunals  The Canadian (and other common law courts) have generally been reluctant to interfere with the decisions of labour tribunals. There are a number of reasons for the traditional display of deference to such bodies. Firstly, there could still exist a “hangover” from earlier days where courts were regarded as biased against organised labour. Anne Davies writes that 96 :  In labour law, there is a long history of scepticism about the role of the judiciary, dating back to the earliest history of the trade union movement. The starting point for many labour lawyers is that traditional judicial values – individualism, freedom of contract, and property rights – tend to align with the interests of employers, rather than with those of employees and trade unions. For example, there is an obvious contrast between the courts’ treatment of the contract of membership between individuals and trade unions, and the courts’ treatment of the contract of employment.  Davies also points to two other significant reasons for deference to labour tribunals – the grounds of “institutional competence” and expertise. The former ground recognises the fact that “judicial proceedings may not be well suited to resolving polycentric disputes in which more than two competing interests are at stake”97. In labour cases it is not necessarily simply the rights of the employer and the employees that must be considered – if a union claim is likely to flow through to other employees in the same sector, or the workforce generally, it may be necessary to consider the macroeconomic consequences of the decision, for example. The second is the related but separate consideration that members of labour tribunals are often appointed to their positions precisely because of their expertise in such matters, and their findings should be afforded significant weight.  96  Anne Davies, “Judicial Restraint in Labour Law”, (2009) 38 Industrial Law Journal 278. 97  Ibid at 290.  39   Another consideration when dealing with labour tribunals is that matters are frequently heard by a tripartite decision-making board, with one member appointed by unions, one appointed by employers, and the Chair agreed upon by the two others. The significance of this form of decision-making was addressed by the Supreme Court in Canadian Union of Public Employees v Ontario (Minister of Labour) 98 , in which Court set aside a decision of the Minister under the Hospital Disputes Labour Arbitration Act 99  to appoint a number of retired judges to a labour tribunal, on the basis that such appointments were “patently unreasonable”. Binnie J, writing for the majority, stated as follows 100 :  I conclude, therefore, that, although the s. 6(5) power is expressed in broad terms, the legislature intended the Minister, in making his selection, to have regard to relevant labour relations expertise as well as independence, impartiality and general acceptability within the labour relations community. By “general acceptability”, I do not mean that a particular candidate must be acceptable to all parties all the time, or to the parties to a particular HLDAA dispute. I mean only that the candidate has a track record in labour relations and is generally seen in the labour relations community as widely acceptable to both unions and management by reason of his or her independence, neutrality and proven expertise.  This decision has been criticised as showing a marked lack of deference to the Minister’s decision. Grant Huscroft, for example, has argued that Binnie J simply substituted his opinion for that of the Minister 101. Even accepting Huscroft’s criticism, CUPE v Ontario is still good evidence of the concern of the courts for the expertise and independence of labour adjudicators.   98   [2003] 1 SCR 539. 99  RSO 1990, c H.14. 100  Supra n98 at paragraph 111. 101  Grant Huscroft, “Judicial Review from CUPE to CUPE: Less is not always more” in Grant Huscroft and Michael Taggart (eds), Inside and Outside Canadian Administrative Law: Essays in Honour of David Mullan, University of Toronto Press, 2006.  40  Canadian courts have been consistent in their insistence on deference to the decisions of labour tribunals. In Toronto (City) v CUPE Local 79, LeBel J noted that “this court has repeatedly stressed the importance of judicial deference in the context of labour law”102, while Cory J noted in Toronto (City) Board of Education v Ontario Secondary School Teachers’ Federation, District 15 (Toronto) that the field of labour relations is “sensitive and volatile”, and that “it 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”103. In United Brotherhood of Carpenters and Joiners of America, Local 1985 v Graham Construction and Engineering Ltd Jackson JA was content to note that “given the historic reluctance to intervene in matters within the jurisdiction of a labour relations board, it is not a stretch to continue to say that the decisions of labour market boards are left to the near-exclusive determination of the decision-maker”104. It is also worth noting that Australia’s primary 20th century authority on the interpretation and effect of privative clauses concerned a decision of a coal industry labour arbitration board 105 . Wartime and national security issues  Another area where the courts have always been reluctant to interfere with the decisions of administrative decision-makers is in the field of wartime regulations 106 , most notably in  102  [2003] 3 SCR 77 at paragraph 68. 103  [1997] 1 SCR 487 at paragraph 35. 104  [2008] SKCA 67 at paragraph 57. 105  Hickman, supra n80. 106  See for example Re Gray (1918) 57 SCR 180 and Reference re Chemicals Regulations [1943] 1 SCR 1. The same can be said for the courts of most common law countries – for an Australian example, see Adelaide Company of Jehovah’s Witnesses Inc v Commonwealth (1943) 67 CLR 116. For a more recent example, see Lord Hoffmann’s  41  Reference re Persons of Japanese Race 107 , which upheld (in the main) the validity of regulations requiring the deportation of Canadian citizens of Japanese birth or indeed ethnicity. Perhaps more interestingly, Canada (Attorney General) v Noxzema Chemical Co of Canada 108  involved a challenge to a decision involving the rate of excise to be charged on the company’s products. Mullan describes the outcome of the case as follows 109 :  The Supreme Court, reversing the Exchequer Court of Canada, rejected the challenge. In so doing, the Court held that it was not its role to assess whether the Minister, acting through his officials, had been correct in his determination. Rather, given the wording of the relevant provision, this was a matter for the “judgment” of the Minister, and the Court could intervene only where he had failed to act “honestly or impartially” or (in the case of three of the five judges) failed to give the taxpayer the “opportunity of being heard”.   Again, as will be seen, this is a very 21 st  century approach to the exercise of a statutory power. However, the Supreme Court’s opinion on this matter changed dramatically from around 1949.  2.3 The denial of deference in the Supreme Court, 1949 – 1975  The real key to understanding the concepts of “deference” and the associated “standard of review” is that they can be seen as a reaction to the administrative law jurisprudence of the Supreme Court from 1949 to 1975. That period started with the abolition of appeals from the Supreme Court to the Judicial Committee of the Privy Council, a move that was made possible by the Imperial Statute of Westminster 1931 110 , but was not finally enacted for all matters until  comments at paragraph 62 of Secretary of State for the Home Department v Rehman [2003] 1 AC 153 (although he seemed to reverse himself almost completely in A v Secretary of State for the Home Department [2004] UKHL 56). 107  [1946] SCR 948. 108  [1942] SCR 178. 109  Mullan, “Deference to the Administrative Process”, supra n6 at 419-20. 110  See in particular s.2.  42  1949 111 . The abolition was purely prospective in nature, meaning that the final Canadian case in the Privy Council was not decided until 1959 112 . The overall effect of the Statute of Westminster was that Canada was no longer bound by British law, and that future decisions of the British courts would have no binding effect in Canada. However, old habits die hard, and Canadian courts continued to be heavily influenced by developments in the UK for some time thereafter, as will be seen very shortly. This period of intense judicial scrutiny of administrative decisions is generally taken to have come to an end with the Supreme Court’s decision in Service Employees’ International Union, Local No 333 v Nipawin District Staff Nurses Association et al 113 , which laid the groundwork for the seminal decision of CUPE.  Few academics or indeed judges have many kind words for the Supreme Court’s approach to judicial review of administrative decisions in this period. LeBel J of the Supreme Court, writing extrajudicially, has stated that:  Over the past 30 years, the Court has distanced itself from the Diceyan conceptions that dominated judicial minds pre-CUPE. That era was exemplified by judicial suspicion of administrative bodies and hoop-jumping to classify questions as jurisdictional so that courts could have the final say 114 .   111  Supreme Court Amendment Act, SC 1949 (2nd session), c 37, s.3. Appeals to the Privy Council in criminal matters had already been abolished by the Criminal Code Amendment Act, SC 1932-33, c 53, s.17. The validity of the Supreme Court Amendment Act was in fact upheld by the Judicial Committee of the Privy Council itself, in Attorney-General (Ontario) v Attorney-General (Canada) (Privy Council Appeals) [1947] AC 127. 112  Ponoka-Calmar Oils v Wakefield [1960] AC 18. For a good summary of the legal issues surrounding the severance of Canada’s ties with the Privy Council, see Harvey M Groberman, “Supremacy and Curial Deference: The Supreme Court of Canada’s Approach to Statutory Interpretation by Administrative Tribunals”, in materials prepared for the Law Society of British Columbia Continuing Legal Education seminar, Administrative Law conference 2009, held in Vancouver, BC, on October 29, 2009. 113  [1975] 1 SCR 382. 114  Justice Louis Le Bel, “Some Properly Deferential Thoughts on Deference”, (2008) 21 Canadian Journal of Administrative Law and Practice 1.  43  David Elliott has described the Supreme Court as becoming “increasingly interventionist” during the 1960s and 1970s 115, while David Mullan lamented that “what then makes the period from 1949 to 1975 (or thereabouts) so disappointing is that, at a time of rapid growth in the administrative state and the recognition of Administrative Law as an independent category of law, the Court rather than advancing on the theoretical front actually seemed to retreat”116 into a Diceyan-type jurisdictional analysis of administrative action.  2.4 Prelude to the Canadian decisions – Anisminic Pty Ltd v Foreign Compensation Commission  To fully understand the decision in Metropolitan Life, and perhaps some commentators’ vehement reaction to that judgement, it is necessary to first briefly consider the judgement of the House of Lords in Anisminic Pty Ltd v Foreign Compensation Commission 117 . Anisminic was the famous, or infamous, decision of the House of Lords which appeared to remove any distinction between jurisdictional error and error within jurisdiction.  It is not necessary to review the facts of the case in detail here, except that Anisminic Ltd sought compensation from the UK Foreign Compensation Commission after certain property that it owned in Egypt had been forcibly acquired by the Egyptian government, without compensation, in the wake of the Suez crisis. The Commission’s enabling legislation, the Foreign  115  David Elliott, “Khosa: Still Searching for that Star”, (2009) 33 Manitoba Law Journal 1 at 8. 116  Mullan, supra n6 at 430. 117  Supra n85.  44  Compensation Act 1950, provided in s.4(4) that “the determination by the Commission of any application made to them under this Act shall not be called in question in any court of law”.  Despite the existence of a privative clause, the House of Lords found that the Commission’s determination should be set aside. In short, the Foreign Compensation Act placed fewer barriers in the way to a successful claim under that Act for an “original owner” of property than for a successor in title. Anisminic claimed to be an original owner. Lord Reid said that inquiries that the Commission made about whether Anisminic had any successors in title were irrelevant, and in fact were outside the Commission’s jurisdiction. The key passage of the judgement can be found in Lord Reid’s comment in paragraph 22:  The Order requires the Commission to consider whether they are satisfied with regard to the prescribed matters. That is all they have to do. It cannot be for the Commission to determine the limits of its powers … [I]f they reach a wrong conclusion as to the width of their powers, the Court must be able to correct that – not because the tribunal has made an error of law, but because as a result of making an error of law they have dealt with and based their decision on a matter with which, on a true construction of their powers, they had no right to deal.   Lord Wilberforce further noted as follows at paragraph 143:  First, the cases in which a tribunal has been held to have passed outside its proper limits are not limited to those in which it had no power to enter upon its inquiry or its jurisdiction, or has not satisfied a condition precedent … A tribunal may quite properly validly enter upon its task and in the course of carrying it out may make a decision which is invalid – not merely erroneous. This may be described as “asking the wrong question” or “applying the wrong test” – expressions not wholly satisfactory since they do not, in themselves, distinguish between doing something which is not in the tribunal’s area and doing something wrong within that area – a crucial distinction which the Court has to make [emphases in original].    45  It is notable in Anisminic that the court did not consider whether, despite making seemingly irrelevant inquiries during the course of its deliberations, the FCC’s decision was reasonable or even correct. Mark Walters comments on the decision in Anisminic as follows 118 :  The ultimate effect of Lord Reid’s assertion in Anisminic – that a tribunal with jurisdiction in the “narrow and original sense”, that is, with authority to enter upon an inquiry, might lose that jurisdiction in the course of the inquiry by misconstruing the law and “asking the wrong question” – was to push the controversial idea of jurisdictional error to its breaking point. British and Canadian judges came to appreciate this point at about the same time, but responded differently. In 1979, Lord Denning concluded in Pearlman 119  that Anisminic had effectively abolished the distinction between jurisdictional and non-jurisdictional errors of law, so that all determinations of law by administrative tribunals were properly seen as jurisdictional and open to judicial correction, even in the face of a privative clause. The House of Lords would eventually agree and British judges would celebrate the fact that administrative law was “liberated” from the “esoteric” distinction between jurisdictional and non-jurisdictional errors of law 120 .  So if the decision in Anisminic, at least as expanded on by Pearlman, greatly simplifies judicial review of administrative decisions by abolishing the distinction between jurisdictional and non- jurisdictional errors, a distinction that is always difficult to draw or maintain, what is the objection to the decision? It appears to me that the prevailing view is that, once the House of Lords had found that the Commission had jurisdiction at the outset to inquire into the matter at hand, it should not have found that the Commission could somehow lose that jurisdiction at a later time. There is also the point by finding that any error of law goes to the decision-maker’s jurisdiction and makes the decision a legal nullity, the courts can simply ignore any privative clause, even the clearly and comprehensively worded privative clause that existed in this case.  118  Mark Walters, “Jurisdiction, Functionalism and Constitutionalism in Canadian Administrative Law”, in Forsyth et al (eds), Effective Judicial Review – A Cornerstone of Good Governance, Oxford University Press, 2010 at 303. 119  Pearlman v Governors of Harrow School [1979] QB 56. 120  O’Reilly v Mackman [1982] 2 AC 237, R v Hull University Visitor, ex parte Page [1993] AC 682.  46   It is interesting that in the Australian literature, many writers have praised the Anisminic approach, which was at least partly adopted in Australia by Craig v South Australia 121  and Plaintiff S157/2002 v Commonwealth of Australia 122 , and some have criticised the High Court for not simply accepting the Anisminic principle in its entirety 123 . Most writers have taken this approach mainly as a means to the end of defeating any privative clause in the Australian Migration Act 1958, which now exists in s.474 of that Act (although the High Court has rendered it largely ineffective) 124 . Such authors would presumably not share the Canadian criticism of Anisminic and Metropolitan Life, unless their real objection is to the immigration privative clause in particular. I will examine this Australian position further later in this thesis.  David Mullan also contends that “the adoption of a policy of deference in Canada has to be seen as a reaction against the interventionist posture of the House of Lords as particularly manifest in the judgment in Anisminic”125. However, the Canadian treatment of administrative tribunals still had to get worse, in the view of such writers, before it got better.    121  (1995) 184 CLR 163. 122  Supra n81. 123  See for example Kirby J (in obiter) in Re Minister for Immigration and Multicultural Affairs, ex parte Holland (2002) 185 ALR 504 at 509; The Hon Duncan Kerr MP, “Privative Clauses and the Courts: Why and How Australian Courts have Resisted Attempts to Remove the Citizen’s Right to Judicial Review of Unlawful Executive Action” (2005) 5 Queensland University of Technology Law and Justice Journal 195 at 214-5. 124  See for example Mary Crock, “Privative Clauses and The Rule of Law: The Place of Judicial Review Within The Construct of Australian Democracy” in Susan Kneebone (ed), Administrative Law and the Rule of Law: Still Part of the Same Package? (1999), published before s.474 came into effect, and Geoff Airo-Farulla and Steven White, “Separation of Powers, ‘Traditional’ Administration and Responsive Regulation”, (2004) 4 Macquarie Law Journal 57. These authors see privative clauses as contrary to the principle of separation of powers and the rule of law generally, and to Chapter III of the Australian Constitution in particular. 125  David Mullan, “Deference: Is It Useful Outside Canada?”, (2006) Acta Juridica 42 at 56.  47  2.5 “Complete lack of deference” cases in Canada  2.5.1 Metropolitan Life  Metropolitan Life Insurance Co v International Union of Operating Engineers, Local 796 126  is a case that has attracted considerable, and at times withering, criticism in academic literature. David Dyzenhaus and Evan Fox-Decent describe it as the Supreme Court’s “most reactionary” decision on judicial review of administrative decisions 127 . David Elliott states that the Supreme Court’s “increasingly interventionist” approach “culminated” with the decision in Metropolitan Life, which he describes as a “terse judgement”128.  David Mullan states that the Supreme Court was particularly suspicious of administrative tribunals from WWII to around 1975, and that Metropolitan Life was the “high-water mark” of this approach129. That nautical metaphor is reversed by Sheila Wildeman, who refers to Metropolitan Life as the “low watermark of deference” in Canada130. So what, exactly, was so wrong with this decision?  The issue in Metropolitan Life was a claim by the union to be certified as the bargaining agent for certain employees of the company for the purposes of the Ontario Labour Relations Act 131 . The central provisions of that Act for the purposes of the case were as follows:   126  Supra n16. 127  David Dyzenhaus and Evan Fox-Decent, “Rethinking the Process / Substance Distinction: Baker v Canada”, (2001) 51 University of Toronto Law Journal 193 at 201. 128  Elliott, supra n115 at 8. 129  Mullan, “Deference – Is It Useful Outside Canada?”, supra n125 at 48. 130  Sheila Wildeman, “A Fine Romance? The Modern Standards of Review in Theory and Practice” in Colleen Flood and Lorne Sossin (eds), Administrative Law in Context, Edmond Montgomery Toronto, 2008 at 244. 131  RSO 1960, c 202.  48  7(1) Upon an application for certification, the Board shall ascertain the number of employees in the bargaining unit at the time the application was made and the number of employees in the union who were members of the trade union at such time as is determined under clause j of subsection 2 of section 77.  (2)  If the Board is satisfied that not less than 45 per cent and not more than 55 per cent of the employees in the bargaining unit are members of the trade union, the Board shall, and if the Board is satisfied that more than 55 per cent of such employees are members of the trade union, the Board may direct that a representation vote be taken.  (3)  If on the taking of a representation vote more than 50 per cent of the ballots of all those eligible to vote are cast in favour of the trade union, and in other cases, if the Board is satisfied that more than 55 per cent of the employees in the bargaining unit are members of the trade union, the Board shall certify the trade union as the bargaining agent of the employees in the bargaining unit.   It can be seen that 7(3) of that Act required the Board to certify the union as the bargaining agent if satisfied that more than 55% of the employees in the “bargaining unit” were members. Further, a strongly worded privative clause was included by way of s.79, which gave the Board exclusive jurisdiction to exercise the powers conferred on it by the Act and to determine all questions of fact or law that arise in any matter before it, and provided that the action or decision of the Board thereon is final and conclusive for all purposes. It appears that both parties agreed in this case that the “bargaining unit” consisted of 30 employees, and the union claimed that 21 of those employees were its members, thereby meeting the requirement of s.7(3).  The company opposed the certification, arguing that the union’s own constitution prevented it from representing the employees in question. The Ontario Labour Relations Board rejected the company’s submission and granted the certification, a decision that was upheld by both the Ontario Supreme Court and Court of Appeal.  49   The Supreme Court upheld the company’s appeal. Cartwright CJ, writing for the court, stated as follows 132 :  If the Board had addressed itself to the question whether fifty-five per cent of the employees were members of the union within the meaning of s. 7(3) of the Act its decision could not have been interfered with by the Court although it appeared that the Board in reaching it had erred in fact or in law or in both … But it is clear from the reasons of the Board read as a whole and particularly from the excerpts therefrom which I have quoted above that the Board did not perform the task imposed upon it by s. 7.   This is a very clear application of the “preliminary question” or “asking the wrong question” test referred to in Anisminic. That is, the Supreme Court found that the Board had not asked the question that its enabling statute required it to ask, and that it had therefore addressed itself to an issue that lay outside its jurisdiction. Note that this approach takes a very wide interpretation of what matters lie within a tribunal’s “jurisdiction”, and therefore gives a court considerable leeway in setting such a decision aside despite the existence of a privative clause. The reasoning is that an error within jurisdiction may well be protected by the clause, but if the decision-maker does not even make a decision that falls within their statutory jurisdiction, they have fallen outside their enabling legislation altogether and the privative clause is of no assistance to them. Whether a decision that is made without jurisdiction is reasonable is not an issue, as it is simply illegal and therefore a nullity.  The Court continued by finding that what the Board had done was not to consider whether the employees in question were actually members of the union, but whether the union considered  132  Metropolitan Life, supra n16 at 434-5.  50  them as members. According to the court, these were two entirely different matters. Only by considering the union’s own constitution could the Board properly decide whether the employees were in fact “members of the union” for the purposes of s.7 of the Act. Cartwright J summed the matter up as follows 133 :  In proceeding in this manner the Board has failed to deal with the question remitted to it (i.e. whether the employees in question were members of the union at the relevant date) and instead has decided a question which was not remitted to it (i.e. whether in regard to those employees there has been fulfillment of the conditions stated above).   Cartwright CJ went on to find that the Board had “stepped outside its jurisdiction” and that the decision was therefore not protected by the privative clause 134, and that “the reason that the certification cannot stand is not that the Board has erred in fact or in law in determining what employees were members of the union within the meaning of s. 7 of the act, but rather that it has refused to put that question to itself, has instead put another question which was not remitted to it and has thereby stepped outside its jurisdiction in the manner indicated above”135. That is, the Board failed to answer a “preliminary question” that it must answer before it could legitimately claim to have jurisdiction over the matter in question.  To sum up, Evans et al set out their objections to the Metropolitan Life decision as follows 136 :  The approach to the definition of jurisdictional review in Metropolitan Life is open to two major criticisms. First, denying that administrative agencies have the authority to decide questions of law conclusively in effect nullifies statutory preclusive clauses. This is difficult to reconcile with the constitutional principle of legislative supremacy, subject to the limits imposed by the Constitution Acts 1867  133  Ibid at 435. 134  Ibid. 135  Ibid at 436. 136  Supra n76 at 707.  51  – 1982. And to the extent that judicial review of administrative agencies for jurisdictional error has a constitutional base in section 96 137 , it deprives legislatures of an important power in a modern administrative state: the power to allocate decision-making responsibilities of courts and agencies in a way that recognises that the effective delivery of a legislative program may require restricting judicial intervention in the administrative process.  Secondly, the reasoning in Metropolitan Life is based on some widely discredited views about the interpretation of statutes. In particular, it assumes that statutory language has a determinate meaning, and that the courts are uniquely qualified to divine its “correct” interpretation. Moreover, on this view, since the interpretation of legislation is a legal question, the specialist expertise of the agency is irrelevant to the exercise.   Note that Evans et al do not see nullification of privative clauses by courts as necessarily a good thing. This approach is, as we shall see later, in sharp contrast to the approach taken by most Australian commentators.  2.5.2 Bell v Ontario Human Rights Commission  Bell is another case that has attracted considerable criticism in the Canadian literature. Dyzenhaus and Fox-Decent state that Bell “rivals Metropolitan Life in its determination to cut the administrative state down to Diceyan size”138. However, the really vituperative comments seem to be reserved for the earlier decision.  The facts in Bell were reasonably straightforward. A Mr McKay, who was Jamaican by birth, alleged that he had been refused rental accommodation by Mr Bell because of his race, and filed a complaint with the Ontario Human Rights Commission (OHRC). The OHRC was unable to resolve the dispute by mediation. On its recommendation, the Minister for Labour constituted a  137  Referring to the Constitution Act 1867. 138  Supra n127 at 209.  52  Board of Inquiry, as was permitted by s.13(1) of the Ontario Human Rights Code 1961-2 (“the Code”). At the start of the inquiry, the Board was asked to find that it had no jurisdiction in the matter. The issue was s.3 of the Code, which provided as follows:  No person, directly or indirectly, alone or with another, by himself or by the interposition of another, shall, (a)  deny to any person or class of persons occupancy of any commercial unit or any self-contained dwelling unit; or (b)  discriminate against any person or class of persons with respect to any term or condition of occupancy of any commercial unit or any self- contained dwelling unit, because of the race, creed, colour, nationality, ancestry or place of origin of such person or class of persons.   Mr Bell claimed that the premises that he had for rent were not a “self-contained dwelling unit”, but were rather simply rooms in his house. The premises therefore did not fall within s.3 of the Code and the Board therefore had no jurisdiction. The Board refused to make such a finding at the outset and instead stated that this would be a matter for consideration during the hearing. Mr Bell then sought a writ of prohibition against the Board hearing the matter. The trial judge agreed with Mr Bell, and the matter ultimately proceeded to the Supreme Court 139 .  The Supreme Court agreed that the premises in question were not a “self-contained dwelling unit”, and found that the Board lacked jurisdiction. Martland J, writing for the majority, discussed the purpose of the Code as follows 140 :   139  It is notable here that the courts were prepared to intervene and prevent the OHRC from hearing the matter, where it might have been thought that intervention before the OHRC even commenced hearings would be premature. This part of the decision was criticised by the Supreme Court in Halifax (Regional Municipality) v Nova Scotia (Human Rights Commission) [2012] SCC 10 where Cromwell J, writing for the Court, stated that “[i]n my view, Bell should no longer be followed and courts should exercise great restraint in intervening at this early stage of the process” (paragraph 17). 140  Bell, supra n17 at 768.  53  As has already been pointed out, the purpose of the Act is to seek to obtain equality of treatment without regard to race, creed, colour, nationality, ancestry, or place of origin, and, to assist in achieving that objective, machinery has been created for the investigation of complaints of discrimination on those grounds.  However, Martland J then goes on to note that the Act does not permit the Board to simply investigate any claim of racial discrimination 141 :  [T]he Act is specifically limited by its terms to dealing with such discrimination when it occurs in relation to defined fields of operation. It states, in terms, that it does not interfere with free expression of opinion on any subject. It does not prevent a householder from refusing to employ a domestic servant because of his antipathy to the race, colour or creed of a person seeking such employment. Similarly, it does not prevent the owner of a house containing dwelling units which are not self-contained from refusing to lease such accommodation to anyone.   In deciding to grant the writ of prohibition, Martland J noted as follows 142 :  The Minister’s authority to appoint a board of inquiry on the recommendation of the Commission arises, and arises only, “if the Commission is unable to effect a settlement of the matter complained of”, i.e., a complaint of discrimination “contrary to this Act”. The task of a board of inquiry appointed by the Minister is to investigate “the matter”, i.e., to determine whether, in a matter within the purview of the Act, there has been discrimination.  The result was that because the discrimination complained of was not of a kind that was “contrary to [the] Act”, the Board had no jurisdiction, and was prohibited from hearing the matter. Whether the Board’s interpretation of its own enabling statute was reasonable or not was not even considered.    141  Ibid. 142  Ibid at 768-9.  54  2.6 Conclusion: the turning of the tide  In this chapter, we have seen that Canadian courts have historically deferred to administrative decision-makers, at least in some areas of administrative law, right from the time of Federation. The reaction against deference really began after WWII, as Canada was pulled along in the wake of UK decisions to that effect. Even after the passage of the Statute of Westminister, the Supreme Court of Canada continued to be heavily influenced by developments in the UK, most particularly the Anisminic decision. The deference approach that developed from 1975 was therefore not an entirely new development in Canadian law (although some of the terminology was). We will now turn to consider how and why the tide turned on these issues, and provide a clearer definition of what the terms “substantive review” and “deference” really mean.  55  CHAPTER THREE – THE RETURN OF DEFERENCE IN CANADIAN SUBSTANTIVE REVIEW  As has already been indicated, there was a major shift (or shift back) in the Supreme Court’s approach to judicial review of administrative decisions in the late 1970s. This chapter will discuss the two Supreme Court cases that did most to explain what “deference” is and how it works, and the related issue of the “standard of review” to be applied to a particular decision. It will be seen in this chapter that the Supreme Court focuses heavily on the intention of the Parliament and the expertise of the decision-maker in determining the degree of deference to be given to that decision-maker. I will argue at the conclusion of this chapter that deference is the appropriate balance to be struck when considering the competing constitutional principles of Parliamentary supremacy and the exercise of judicial power in review of administrative decisions.  3.1 Canadian Union of Public Employees, Local 963 v New Brunswick Liquor Corp  All Canadian commentators agree that CUPE is one of the most signification cases in Canadian administrative law. It was one of two exceptionally important judgements handed down during 1979, the other being Nicholson v Haldimand-Norfolk (Regional) Board of Commissioners of Police 143 , which today is still one of the leading procedural fairness cases in Canada. Dyzenhaus and Fox-Decent describe the judgement in CUPE as “the beginning of the Supreme Court’s  143  [1979] 1 SCR 311.   56  rejection of the Diceyan paradigm”144, while Audrey Macklin states that CUPE “is to the standard of review what Nicholson is to procedure and Baker 145  is to discretion: a judgement that shifts the legal landscape onto new terrain”146. L’Heureux-Dubé J, writing extrajudically, commented as follows 147 :  The Supreme Court’s decision in [CUPE] has been described as “one of the most influential judgments in modern Canadian administrative law”148. In the wake of CUPE, it could no longer be assumed that an administrative tribunal’s interpretation of its statute would be subject to correction on judicial review simply because the reviewing judge disagreed with the board’s interpretation. The Supreme Court, it seemed, had sent a message to lower courts to show some deference to specialised bodies charged with administering legislation.  The facts in CUPE were fairly straightforward, and the legislation involved in the case anything but. The union went on strike in 1979, and on 22 August 1979 made a complaint to the Public Service Labour Relations Board of New Brunswick (“the Board”) that the corporation was replacing striking staff with management personnel. The corporation in turn complained that the union was picketing their premises. Both of these actions were said to be contrary to s.102(3) of the Public Service Labour Relations Act (NB) 149 , which provided as follows:  Where subsection (1) and subsection (2) are complied with employees may strike and during the continuance of the strike (a)  the employer shall not replace the striking employees or fill their position with any other employee, and (b)  no employee shall picket, parade or in any manner demonstrate in or near any place of business of the employer.  144  Supra 127 at 200. 145  Baker, supra n60. 146  Audrey Macklin, “Standard of Review: The Pragmatic and Functional Test” in Colleen Flood and Lorne Sossin (eds), Administrative Law in Context, Edmond Montgomery (Toronto) 2008 at 203. 147  Madame Justice Claire L’Heureux-Dubé, “The ‘Ebb’ and ‘Flow’ of Administrative Law on the ‘General Question of Law’” in Michael Taggart (ed), The Province of Administrative Law, Oxford: Hart Publishing, 1997 at 308-9. 148  National Corn Growers’ Association v Canada (Import Tribunal) [1990] 2 SCR 1324 at 1331. 149  RSNB 1973, c P-25.  57   Dickson J, writing for the Supreme Court, described this section as follows 150 :  On one point there can be little doubt – section 102(3)(a) is very badly drafted. It bristles with ambiguities. Mr. Justice Limerick of the New Brunswick Appeal Division, in the course of his reasons in the present litigation, said: “Four possible interpretations immediately come to mind”.  This statement, made just two pages into the judgement, turns out to be crucial in the final decision. The Act also included a privative clause at s.101 as follows:  (1)  Except as provided in this Act, every order, award, direction, decision, declaration, or ruling of the Board, the Arbitration Tribunal or an adjudicator is final and shall not be questioned or reviewed in any court.  (2)  No order shall be made or process entered, and no proceedings shall be taken in any court, whether by way of injunction, certiorari, prohibition, quo warranto, or otherwise, to question, review, prohibit or restrain the Board, the Arbitration Tribunal or an adjudicator in any of its or his proceedings.  The Supreme Court allowed the union’s appeal and restored the decision of the Board. The crucial passage in the judgement can be found at paragraph 29:  The question of what is and is not jurisdictional is often very difficult to determine. The courts, in my view, should not be alert to brand as jurisdictional, and therefore subject to broader curial review, that which may be doubtfully so. Upon a careful reading of the Act, the Board’s decision, and the judgments in the Court of Appeal, however, I find it difficult to brand as “patently unreasonable” the interpretation given to s.102(3)(a) by the Board in this case. At a minimum, the Board’s interpretation would seem at least as reasonable as the alternative interpretations suggested in the Court of Appeal. Certainly the Board cannot be said to have so misinterpreted the provision in question as to “embark on an inquiry or answer a question not remitted to it”.   150  Supra n5 at 299.  58  Some of the most important features of the judgement are as follows. Firstly, instead of a court focusing its energy on attempting to find its way around a privative clause, as in Metropolitan Life, Dickson J examined the reason for the existence of privative clauses in the first place, and emphasised the “legislative choice to confer certain tasks on administrative actors, the specialised expertise and accumulated experience of administrative bodies, and the virtues of judicial restraint”151. Privative clauses were therefore not to be regarded as some kind of affront to the rule of law and to be circumvented at all costs 152 .  Secondly, Dickson J found that the interpretation of s.102(3) “would seem to lie logically at the heart of the specialised jurisdiction confined to the Board”153. A court should only regard the Board’s interpretation of its own legislation as a jurisdictional error when that interpretation is “so patently unreasonable that its construction cannot rationally be supported by the relevant legislation”.154 That is, in this case, the “standard of review” to be applied by the courts was one of “patent unreasonableness”. The differing standards of review available on judicial review of administrative decisions will be discussed in more detail in the section dealing with the decision in Dunsmuir v New Brunswick 155 .   151  Ibid at 204. 152  It should be noted that only two years after CUPE was decided, the Supreme Court made it clear that a privative clause can never completely remove the jurisdiction of the courts on judicial review – Crevier, supra n48. The subject of an administrative decision will therefore never be without recourse to review of a truly unreasonable decision. 153  Supra n5 at paragraph 15. 154  Ibid at paragraph 16. 155  Supra n7.  59  David Dyzenhaus states that an important result of CUPE is that judges are no longer to have an interpretive monopoly insofar as law is concerned 156 . Tribunals are no longer to be condescendingly regarded as “inferior tribunals”, but as “specialised bodies that possess a legislative mandate to apply their expertise and experience to matters that they may be better suited to address than the ordinary court”157. If more than one reasonable interpretation of an enabling statute is possible, why should courts have such a monopoly? This was indeed something new for Canadian law, in that it gives administrative decision-makers perhaps not an equal footing with courts, but gives them a measure of respect for their decision-making capacity that was lacking in Metropolitan Life and Bell.  Thirdly, in CUPE, as has already been noted, Dickson J noted that s.102(3) of the Act was very badly drafted, meaning that there was no one clearly “right” interpretation. None of the various interpretations of that subsection given by the lower courts, the Board or the parties was patently unreasonable, so the courts should let the decision of the Board, the body tasked by statute with interpreting the Act, stand.  Dickson J seemed prepared to take the view that “good faith” on the part of decision-makers should be presumed in the absence of evidence to the contrary (such as existed in Roncarelli v Duplessis 158 , where there was a very clear abuse of power by Premier Duplessis). A decision will be reasonable if it is rational, in the sense that if it is a matter on which “reasonable minds might differ”159, and deference will be shown if the decision-maker is interpreting his or her enabling legislation (or “home statute” as it has become known), has  156  David Dyzenhaus, “Constituting the Rule of Law: Fundamental Values in Administrative Law”, (2002) 27 Queens Law Journal 445 at 456. 157  Macklin, supra n146 at 203. 158  [1959] SCR 122. 159  See for example SZMDS, supra n4 at paragraph 131.  60  expertise in the relevant field (two factors which are clearly related), and is protected by a privative clause. Finally, by referring to “patent unreasonableness”, Dickson J sets the scene for the later “pragmatic and functional test” and “standard of review” analyses that would determine whether or not deference should be shown in a particular situation.  It is notable that later courts went on to find that it is not just in cases of poor drafting that there can be more than one reasonable interpretation of a statute. How often is there only one possible interpretation of a statute? Canadian courts have been willing to find that there are often multiple reasonable interpretations of a statutory provision, and thereby allow the interpretation of the administrative decision-maker to stand 160 .  One weakness in the judgement may be that while Dickson J was justifiably critical of the “preliminary or collateral question” doctrine that was applied in Metropolitan Life, he did not clearly specify an alternative for determining what truly is a question of jurisdiction and what is not 161 . He did specify that “courts, in my view, should not be alert to brand as jurisdictional, and therefore subject to broader curial review, that which may be doubtfully so”162, and also noted that “jurisdiction is typically to be determined at the outset of the inquiry”163, which could be taken as a refutation of the concept that a tribunal could lose jurisdiction in the course of its inquiry.   160  See for example National Corn Growers, supra n148, and Domtar Inc v Quebec (Commission d’Appel en Matiere de Lesions Professionnelles) [1993] 2 SCR 756. 161  This was still a live issue nearly thirty years later in Dunsmuir, supra n7, which will be discussed later in this thesis. 162  Supra n5 at paragraph 10. 163  Ibid at paragraph 9.  61  The last point illustrates that is not correct to say that there have been no academic criticisms of CUPE, but they are mainly of the “CUPE did not go far enough” rather than the “CUPE was wrong” variety. CUPE could have been an even better reasoned and even more influential decision if the Supreme Court had conceded what was really the obvious, that is, that it was making new law, or perhaps returning the law to its pre-WWII state. Instead Dickson J seemed reluctant to simply state that Metropolitan Life was wrongly decided, perhaps out of deference to his predecessors on the Court. For example, Wade MacLauchlan notes that Dickson J made “great efforts to demonstrate consistency with Metropolitan Life and other decisions whose main point was to find a path around privative clauses ”164 , while David Mullan has written that Dickson J “failed to repudiate” the earlier case165, clearly implying that it is a case that needs clear repudiation.  Finally, David Mullan has elsewhere identified the following principles of administrative law as having their roots in the CUPE decision 166 . Firstly, deference may be necessary for tribunals not protected by any privative clause 167 . Secondly, at least limited deference to agencies must be given even in cases that came by way of statutory appeal (that is, an appeal to a court permitted by the enabling legislation itself) as opposed to judicial review 168 . Thirdly, CUPE has resulted in the extension of deference to discretionary decisions, at least in some situations, as made clear by  164  Wade MacLauchlan, “Transforming Administrative Law: The Didactic Role of the Supreme Court of Canada”, (2001) 80 Canadian Bar Review 281 at 285-6. 165   David Mullan, “A Blast from the Past: A Surreptitious Resurgence of Metropolitan Life?”, (1992) 5 Administrative Law Reports (2d) 97 at 101. 166  David Mullan, “A View from North America: A Canadian Perspective on Three Troubling Issues”, (2004) 17 Canadian Journal of Administrative Law and Practice 167 at 171-2 167  United Brotherhood of Carpenters and Joiners of America, Local 579 v Bradco Construction Ltd [1993] 2 SCR 316. 168  Pezim v British Columbia (Superintendent of Brokers) [1994] 2 SCR 557; Canada (Director of Investigation and Research) v Southam Inc [1997] 1 SCR 748.  62  Baker v Canada (Minister for Immigration and Citizenship) 169 . This case is also authority for the principle that deference can be afforded to procedural choices of agencies that have the power to make their own procedures and have some expertise in the matter.  Finally, CUPE led to a (temporary at least) disappearance of “jurisdictional error” discussions170. It should be noted that since Mullan’s article was written, the jurisdictional error concept may have been given new life by Dunsmuir v New Brunswick 171  and subsequent cases, and this issue will be discussed more fully shortly.  3.2 The four standards of review in Canadian administrative law  We have seen that in CUPE, the Supreme Court recognised (or recognised once more) that administrative decision-makers have a role conferred by Parliament and expertise in their field, and that their interpretations of their own enabling legislation (including matters of jurisdiction), while never definitive, should at least be given “weight” in judicial review. By finding that the interpretation given to s.102 by the Board was not “patently unreasonable”, Dickson J at least implicitly created the concept of the “standard of review” of an administrative decision, a matter to which we will now turn.  But how much deference should be given to administrative decision-makers? Were the courts to defer only to an administrative decision-maker whose decisions were protected by a privative  169  Supra n60. 170  Pushpanathan v Canada (Minister of Citizenship and Immigration) [1998] 1 SCR 982. 171  Supra n7.  63  clause? How much, if any, consideration should be given to the expertise of the decision-maker? Finally, to what should the court defer – the decision itself or the reasons given for that decision? The way that Canadian courts have answered these questions is through the development and application of standards of review.  Since 1979, four “standards of review” have existed in Canadian administrative law. Until 1997, there were two standards – “patent unreasonableness” and correctness. In 1997, the Supreme Court introduced a third, intermediate standard of “reasonableness simpliciter”172. However, Dunsmuir v New Brunswick 173  collapsed the two reasonableness standards into one, that of “reasonableness”174.  The law surrounding the issue of the choice of the standard of review is almost a thesis topic in itself, and must by necessity be discussed briefly here. However, it is clear from CUPE that Dickson J did not intend that all administrative decisions were to be reviewed on the basis of “patent unreasonableness”175. Questions of jurisdiction, it would seem, were still to be assessed on a basis of “correctness”, and CUPE simply said that courts should not be overly willing to classify provisions in the agency’s enabling legislation as going to jurisdiction when this  172  Canada (Director of Investigation and Research) v Southam Inc [1997] 1 SCR 748. 173  Supra n7 at paragraph 45. 174  It is not clear whether the post-Dunsmuir standard of “reasonableness” is the same thing as the “reasonable simpliciter” standard. Audrey Macklin states that in Dunsmuir the court “reverted” to a two-standard system – supra n146 at 227. Piper Henderson on the other hand states that Dunsmuir “substitutes the old ‘reasonableness simpliciter’ and ‘patent unreasonableness’ standards with a single ‘reasonableness’ standard”, suggesting that the new standard is different from either of the two previous ones. See Henderson, “Supreme Court of Canada’s New ‘Reasonableness’ Standard of Review Applied in Recent Education Cases”, (2008) 18 Education and Law Journal 179 at 180. 175  CUPE, supra n5 at paragraph 16.  64  classification was “doubtful”176. When, then, should a tribunal’s decision attract deference and when should it not?  This question has, not surprisingly, never been fully satisfactorily resolved. It is notable that some of the early post-CUPE cases have attracted accusations of backsliding from the breakthroughs made in that case 177 . The Supreme Court has never fully disassociated itself from the concept of jurisdictional review, although it went close in Pushpanathan v Canada (Minister of Citizenship and Immigration) 178, where Bastarache J stated that “it should be understood that a question which ‘goes to jurisdiction’ is simply descriptive of a provision for which the proper standard of review is correctness”. In one of the major post-Dunsmuir cases, Alberta (Information and Privacy Commissioner) v Alberta Teachers’ Association, a majority of the court stated that “it may be that the time has come to reconsider whether, for purposes of judicial review, the category of true questions of jurisdiction exists and is necessary to identifying the appropriate standard of review”179, which suggests that the court may be preparing to finally jettison the concept once and for all.  3.2.1 Flaws in the (pre-Dunsmuir) reasonableness standards  In this section, I will identify two conceptual flaws that existed in the “standard of review” jurisprudence prior to Dunsmuir. The most important of these was the introduction of a third,  176  Ibid at paragraph 10. 177   For a good summary of Supreme Court decisions from 1979 to 1987 on the deference issue, see David Mullan, “The Supreme Court of Canada and Jurisdictional Error: Compromising New Brunswick Liquor?”, (1987) 1 Canadian Journal of Administrative Law and Practice 71. 178  Pushpanathan, supra n170 at paragraph 28. 179  [2011] SCC 61 at paragraph 34.  65  intermediate “reasonableness simpliciter” standard of review, instead of a “sliding scale” of reasonableness. I will argue in this thesis that the “sliding scale” is a preferable form of reasonableness review to any attempt to differentiate between “patent unreasonableness” and “reasonableness simpliciter”. Secondly, the Canadian courts’ prohibition on “reweighing” matters before the decision-maker conflicts with the concept of reasonableness in administrative decision-making. If a decision-maker gives unreasonable weight to a relevant factor before him or her, surely that decision is itself unreasonable, but Canadian courts have to date not agreed with this proposition. A “sliding scale” of review?  The “reasonableness simpliciter” standard of review was, in my opinion, probably the biggest error in Canadian administrative law, at least post-CUPE. When is a decision “unreasonable” but not “patently unreasonable”? One case that made a noble, if most likely ultimately unsuccessful, attempt to distinguish the two standards is Law Society of New Brunswick v Ryan 180 . The Supreme Court explained the difference between the two standards of review at paragraphs 52 and 53 as follows:  [52] 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”181 … A decision that is patently unreasonable is so flawed that no amount of curial deference can justify letting it stand.  180  [2003] 1 SCR 247. 181  Referring to Canada (Attorney General) v Public Service Alliance of Canada [1993] 1 SCR 941, at pp 963-64, per Cory J; Centre Communautaire Juridique de l’Estrie v Sherbrooke (City) [1996] 3 SCR 84, at paras 9-12, per Gonthier J.  66   [53] 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.   The problem that remains is that even on the Ryan analysis, a decision that is unreasonable in some sense would still be allowed to stand simply because that unreasonableness does not leap out immediately at the reader. It is not hard to think of situations where a decision that suffers from unreasonableness that is only detectable at a later date is worse than one that has immediate and obvious defects.  The other important feature of Ryan is that it rejects the notion that there is a “spectrum” or “continuum” of reasonableness. That is, it was argued in Ryan that there are more or less infinite gradations of reasonableness depending on the issue in question. Iaccobucci J, writing for the Court, rejected this proposition, stating that “If it is inappropriate to add a fourth standard to the three already identified, it would be even more problematic to create an infinite number of standards in practice by imagining that reasonableness can float along a spectrum of deference such that it is sometimes quite close to correctness and sometimes quite close to patent unreasonableness”182, and that the standards of review might be “arranged along a gradient of deference but it was never meant to suggest an infinite number of possible standards”183. Iaccobucci J also stated as follows 184 :   182  Supra n180 at paragraph 44. 183  Ibid at paragraph 45. 184  Ibid at paragraph 46.  67  Judicial review of administrative action on a standard of reasonableness involves deferential self-discipline. A court will often be forced to accept that a decision is reasonable even if it is unlikely that the court would have reasoned or decided as the tribunal did 185 . If the standard of reasonableness could “float” this would remove the discipline involved in judicial review: courts could hold that decisions were unreasonable by adjusting the standard towards correctness instead of explaining why the decision was not supported by any reasons that can bear a somewhat probing examination.  This proposition seems to defeat the purpose of “reasonableness”. A decision to refuse an applicant refugee status, for example, is a much more serious matter than conditions imposed on a broadcasting licence, for example. If a court cannot consider the impact on the applicant of the decision and decide what is “reasonable” accordingly, why bother with a reasonableness standard at all?  Iaccobucci J was correct to require “judicial discipline” of judges, and to point out that mere disagreement with an administrative decision is not sufficient grounds to overturn that decision. However, how would a sliding scale of review exacerbate a problem that could exist regardless? There does not seem to be any real link between the two matters. “Weight”  In Suresh v Canada (Minister of Citizenship and Immigration) 186 , the Supreme Court took the view that courts are prohibited from “reweighing” factors considered by the administrative decision-maker 187 . A unanimous court stated as follows at paragraph 29:   185  Southam, supra n172 at paragraphs 78-80. 186  [2002] 1 SCR 3. 187  In Khosa, supra n29, Binnie J, writing for the majority of the Supreme Court, also declined to “reweigh” the evidence before the Immigration Appeals Division, but he did not argue that there is a blanket prohibition on the courts doing so – see paragraphs 65-67.  68  The first question is what standard should be adopted with respect to the Minister’s decision that a refugee constitutes a danger to the security of Canada. We agree with Robertson JA that the reviewing court should adopt a deferential approach to this question and should set aside the Minister’s discretionary decision if it is patently unreasonable in the sense that it was made arbitrarily or in bad faith, it cannot be supported on the evidence, or the Minister failed to consider the appropriate factors. The court should not reweigh the factors or interfere merely because it would have come to a different conclusion.  This prohibition on “reweighing” seems to contradict Baker188, where L’Heureux-Dubé J, writng for herself and Gonthier, Cory, McLachlin, Iacobucci, Bastarache and Binnie JJ, stated as follows at paragraph 75:  The principles discussed above indicate that, for the exercise of the discretion to fall within the standard of reasonableness, the decision-maker should consider children’s best interests as an important factor, give them substantial weight, and be alert, alive and sensitive to them. That is not to say that children’s best interests must always outweigh other considerations, or that there will not be other reasons for denying an H&C claim even when children’s interests are given this consideration. However, where the interests of children are minimized, in a manner inconsistent with Canada’s humanitarian and compassionate tradition and the Minister’s guidelines, the decision will be unreasonable.  It is very difficult to see how a decision can be “reasonable” in any substantive sense if a crucial factor requiring consideration is given only cursory treatment by a decision-maker. David Dyzenhaus notes that “Suresh largely pays lip service to Baker on review of substance, since it diminishes review of substance to a checklist of factors which the Minister must take into account”189. Dyzenhaus has also written that “Weight is … just a metaphor for a proper inquiry into the balance of reasons”190. Even Australian courts, which (as will be seen) insist on a strict  188  Baker, supra n160 at paragraph 75. 189  Dyzenhaus, “Constituting the Rule of Law”, supra n156 at 506. See also David Mullan, “Establishing the Standard of Review: The Struggle for Complexity?”, (2004) 17 Canadian Journal of Administrative Law and Practice 59” at 67-8. 190  David Dyzenhaus, “The Role of (Administrative) Law in International Law”, (2004) 68 Journal of Law and Contemporary Problems 127 at 135.  69  separation between “merits review” and “judicial review”, have concluded that “failure to give adequate weight to a matter of great importance” amounts to an error of law191.  This is an issue that seems to be under reconsideration by the Supreme Court. In Doré v Barreau du Québec 192  Abella J, writing for the court, found that at least where “Charter values” are in question, it will be appropriate for a court to examine whether an administrative decision-maker has given appropriate weight to competing considerations. It remains to be seen what “Charter values” actually are – is this merely another way of describing the rights protected by the Charter, or something else? If the latter, what are these values and how will they be formulated by the courts?  It is also significant that Suresh was ultimately successful in his case on procedural and Charter grounds. One of the grounds on which the deportation order was set aside was that Citizenship and Immigration Canada (CIC) had failed to appreciate that deportation of a refugee to a place where he or she faces a “substantial” risk of torture is a violation of s.7 of the Charter, and had not given this issue sufficient consideration in making its decision. Having found that the Minister of Immigration’s refusal to provide the applicant with the details of the case against him was a breach of s.7 of the Charter, the Court then found as follows 193 :  Despite the legitimate purpose of s. 53(1)(b) of the Immigration Act in striking a balance between the need to fulfil Canada’s commitments with respect to refugees and the maintenance of the safety and good order of Canadian society, the lack of basic procedural protections provided to Suresh cannot be justified by s. 1 in our  191  Minister for Aboriginal Affairs v Peko-Wallsend Ltd (1986) 162 CLR 24 at 42. 192  Supra n21 at paragraphs 47 and 57. 193  Supra n186 at paragraph 128.  70  view. Valid objectives do not, without more, suffice to justify limitations on rights. The limitations must be connected to the objective and be proportional.  While a detailed examination of s.1 is beyond the scope of this thesis, it should be noted that the leading case on s.1, R v Oakes, expressly states that a party invoking s.1 must show that the means chosen to address a legitimate objective are “reasonable and demonstrably justified”, which is said to involve a “form of proportionality test”194. A breach of constitutionally or legislatively conferred rights can only be upheld if the breach is somehow proportional to the objective it serves. Given the decision in Doré v Barreau du Québec 195 , which appears to mandate a “proportionality” approach for review of administrative decisions impacting on Charter rights, it is unlikely that the substantive review elements of Suresh would be decided the same way today. It will be seen in Chapter 4 of this thesis that the UK has already substantially implemented a proportionality approach where fundamental rights are at stake.  Another indication that the Suresh prohibition on “reweighing” may be on the way out came in the recent decision of Moore v British Columbia (Education) 196 . Moore was decided under the BC Human Rights Code 197, but the decision’s approach to the relative weight of considerations may have some implications for judicial review in Canada. Abella J, writing for the court, found that the decision of the North Vancouver School District to close a diagnostic centre for children with learning disabilities (dyslexia, in the case of the applicant) breached s.8 of the Code. Abella J stated as follows at paragraphs 46 and 47:   194  R v Oakes [1986] 1 SCR 103 at paragraph 70. 195  Supra n21. 196  Supra n62. 197  RSBC 1996, c 210  71   46 … In brief, the Tribunal found that when the decision to close the Diagnostic Centre was made, the District’s motivations were exclusively financial, and it had failed to consider the consequences or plan for alternate accommodations.  47. This failure was crucial in light of the expert evidence that intensive supports were needed generally to remedy Jeffrey’s learning disability, and that he had not received the support he needed in the public school  Abella J noted at paragraphs 51 and 52 that the school district, when faced with financial difficulties, had decided to close the diagnostic centre, but keep open other facilities such as the Vancouver “outdoor school”. She stated at paragraph 52 that “[t]he failure to consider financial alternatives completely undermines what is, in essence, the District’s argument, namely that it was justified in providing no meaningful access to an education for Jeffrey because it had no economic choice”. This reads very much like an argument that the school Board had not given sufficient weight to other measures that were open to it. If this line of reasoning is imported into judicial review, the Suresh prohibition on “reweighing” will be no more. It could also have the effect of giving the courts the power to rewrite the budgets of government authorities, which would definitely be more problematic.  3.2.2 Judicial criticism of the “three standards” approach  As long ago as 1997, courts were also expressing their confusion with the three-standard approach. In Miller v Workers’ Compensation Commission (Newfoundland), Barry J noted as follows 198 :   198  1997 CanLII 10862 (Supreme Court of Newfoundland and Labrador Trial Division) at paragraph 27.  72  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.   The first real judicial attack in the Supreme Court on the three-standard system came in Toronto (City) v Canadian Union of Public Employees, Local 79 199 . LeBel J concurred in the result in this case, but was critical of the application of both the patent unreasonableness and reasonableness simpliciter standards. LeBel J noted as follows at paragraphs 106 and 107:  [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”.  [107] Under both patent unreasonableness and reasonableness simpliciter, mere disagreement with the adjudicator’s decision is insufficient to warrant intervention … 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”. 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” (Ryan200). 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)?   Sheila Wildeman 201  comments on LeBel’s “cri de couer” as follows:  199  Supra n102. 200  Supra n180 at paragraph 55. 201  Supra  n130 at 263.  73   The idea of letting unreasonable or irrational decisions stand, whether because they are not irrational enough, or because they require some work to discover, is in conflict with both the principle of Parliamentary supremacy (read in the light of the long-standing common law principle that the legislature should not be understood to intend irrationality), and with the “animating principle that, in a society governed by the rule of law, power is not to be exercised arbitrarily or capriciously”202.   Further criticism of the three-standard system came in Council of Canadians With Disabilities v VIA Rail Canada Inc 203 , in which a 5-4 majority agreed that the reasonableness standards needed reformulation, but were not prepared to go the extra step and make any substantial change.  3.2.3 Limits on the principle of deference  We have seen from the above discussion that once CUPE had enunciated the principle of deference to administrative decision-makers, the Supreme Court appeared to attempt to limit it in some ways. The idea that courts should not “reweigh” matters before administrators appears to contradict the entire notion of “reasonableness review”, and the refusal by Ryan, amongst other cases, to recognise a “spectrum” of reasonableness could also have the effect of emptying the “reasonableness simpliciter” standard of much of its content. The attempt to distinguish between that standard and “patent unreasonableness” always seemed doomed, and it was not surprising that by 2008 the Supreme Court was ready for another attempt to reformulate the law of substantive review.    202  Toronto (City) v CUPE, supra 102 at paragraphs 125 and 128. 203  [2007] SCC 15.  74  3.3 A new high water mark? Dunsmuir v New Brunswick  The most significant restatement by the Supreme Court of when deference should be given to an administrative decision-maker and when it should not came in the 2008 decision of Dunsmuir v New Brunswick 204 . There can be no doubt that Dunsmuir is a significant decision, but academic opinion is divided on whether it is a revolutionary as it was first supposed to be. Susan Gratton notes that “in Dunsmuir the Court embarked on a mission to simplify the entire system of judicial review, dispensing with the patent unreasonableness standard and replacing the pragmatic and functional test with a new standard of review analysis”205. Dustin Kennall is a fan of the decision, noting that “Dunsmuir abandoned the prolix, unpredictable ‘pragmatic and functional’ approach to determining the appropriate standard of review in favour of a more circumscribed analysis”206.  David Gruber is somewhat more cautious, stating that “the Court in Dunsmuir attempted ostensibly to simplify the system of judicial review and to develop a coherent, workable, and principled framework for judicial review by dispensing with the patent unreasonable standard and replacing the pragmatic and functional test with a new standard of review analysis”207 (my emphasis). Finally, David Mullan, possibly giving his position away with the title of his article, is somewhat sceptical about whether Dunsmuir will change a great deal, stating that “the determination of how to approach review of any particular decision has not necessarily become  204  Supra n7. 205  Susan Gratton, “Standing at the Divide: The Relationship Between Administrative Law and the Charter Post- Multani”, (2008) 53 McGill Law Journal 477 at 487. 206  Dustin Kennall, “De-Regulating the Regulatory Compact: The Legacy of Dunsmuir and the Jurisdictional Question Doctrine”, (2011) 24 Canadian Journal of Administrative Law and Practice 115. 207  David Gruber, “Judicial Review Advocacy in the Post-Dunsmuir Era”, (2009) 22 Canadian Journal of Administrative Law and Practice 303.  75  any easier, or for that matter, less likely to minimise the extent to which counsel and lower courts have to concern themselves with the appropriate posture of the courts”208, although he does concede that “the initial choice between two standards (as opposed to amongst three) may make life simpler”209. What is clear is that Dunsmuir has done away with the artificial distinction between patent unreasonableness and reasonableness simpliciter, and that can only be a good thing for administrative law in Canada.  I will approach this case by firstly considering the facts and the judicial background. I will then move on to discuss each of the three judgements given in that case, comparing and contrasting them as required. In doing so, I will focus on the impact this case has had on the meanings of “deference”, “reasonableness” and “correctness”, and the situations in which the standards of review should be applied.  3.3.1 Facts  David Dunsmuir was an employee of the New Brunswick Court of Queen’s Bench, who was dismissed from that employment after three reprimands. Dunsmuir was dismissed under “at pleasure” provisions of his contract, and the letter of termination explicitly stated that he was not being dismissed for cause. Dunsmuir sought review of his termination under the Public Sector Labour Relations Act 210 , arguing that despite the wording of the letter, he had in fact been  208  David Mullan, “Dunsmuir v New Brunswick, Standard of Review and Procedural Fairness for Public Servants: Let’s Try Again!”, (2008) 21 Canadian Journal of Administrative Law and Practice 117 at 135. 209  Ibid. 210  Supra n149.  76  dismissed for cause, and therefore had the right to certain procedural protections available under the Civil Service Act 211 . In particular, s.97(2.1) of the former Act stated as follows:  Where an adjudicator determines that an employee has been discharged or otherwise disciplined by the employer for cause … the adjudicator may substitute such other penalty for the discharge or discipline as to the adjudicator seems just and reasonable in all the circumstances.  An arbitrator appointed for the purposes of Dunsmuir’s complaint found that the effect of s.97(2.1) was that he could find that the dismissal was for cause, despite the express statement in the letter to the contrary. He therefore ordered Dunsmuir’s reinstatement because the procedural requirements of the Civil Service Act had not been followed. New Brunswick sought judicial review of the decision. The arbitrator’s decision was quashed by the Court of Queen’s Bench212, and this decision was upheld by the New Brunswick Court of Appeal 213 . Dunsmuir appealed against the latter decision to the Supreme Court.  3.3.2 Majority judgement  The most immediately notable feature of the Supreme Court’s judgement was that the two “reasonableness” standards should be collapsed into one. Bastarache and LeBel JJ, writing for themselves, McLachlin CJ and Fish and Abella JJ, stated as follows 214 :  As discussed by LeBel J at length in Toronto (City) v CUPE 215 , notwithstanding the increased clarity that Ryan 216  brought to the issue and the theoretical  211  SNB 1984, c C-5.1. 212  [2005] NBQB 270. 213  [2006] NBCA 27. 214  Supra n7 at paragraph 41. 215  Supra n92. 216  Law Society of New Brunswick v Ryan, supra n170.  77  differences between the standards of patent unreasonableness and reasonableness simpliciter, a review of the cases reveals that any actual difference between them in terms of their operation appears to be illusory.  At paragraph 45 217  Bastarache and LeBel JJ state as follows:  We therefore conclude that the two variants of reasonableness review should be collapsed into a single form of “reasonableness” review. The result is a system of judicial review comprising two standards – correctness and reasonableness. But the revised system cannot be expected to be simpler and more workable unless the concepts it employs are clearly defined.  Bastarache and LeBel JJ then define the terms “reasonableness” and “correctness”. The former is defined as follows 218 :  Reasonableness is a deferential standard animated by the principle that underlies the development of the two previous standards of reasonableness: certain questions that come before administrative tribunals do not lend themselves to one specific, particular result. Instead, they may give rise to a number of possible, reasonable conclusions. Tribunals have a margin of appreciation within the range of acceptable and rational solutions. A court conducting a review for reasonableness inquires into the qualities that make a decision reasonable, referring both to the process of articulating the reasons and to outcomes. In judicial review, reasonableness is concerned mostly with the existence of justification, transparency and intelligibility within the decision-making process. But it is also concerned with whether the decision falls within a range of possible, acceptable outcomes which are defensible in respect of the facts and law.  The last sentence of this paragraph appears to mean, in my opinion, that it is possible for there to be more than one reasonable outcome in an administrative proceeding, and courts should be wary of simply substituting their view for that of the decision-maker. Bastarache and LeBel JJ emphasise that “the move towards a single reasonableness standard does not pave the way for a  217  Supra n7 at paragraph 45. 218  Ibid at paragraph 47.  78  more intrusive review by courts and does not represent a return to pre-Southam formalism”219. In other words, the pre-1975 kinds of reasoning are not to be resuscitated.  The term “deference” is defined as follows220:  Deference is both an attitude of the court and a requirement of the law of judicial review. It does not mean that courts are subservient to the determinations of decision makers, or that courts must show blind reverence to their interpretations, or that they may be content to pay lip service to the concept of reasonableness review while in fact imposing their own view. Rather, deference imports respect for the decision-making process of adjudicative bodies with regard to both the facts and the law. The notion of deference “is rooted in part in a respect for governmental decisions to create administrative bodies with delegated powers” (Canada (Attorney General) v Mossop, [1993] 1 SCR 554, at p 596, per L’Heureux-Dubé J, dissenting). We agree with David Dyzenhaus where he states that the concept of “deference as respect” requires of the courts “not submission but a respectful attention to the reasons offered or which could be offered in support of a decision”221.  It is important to note that deference is defined as “respect” for the decision-making ability of the tribunal whose decision is under review, and a willingness to at least consider whether its reasoning is plausible, on matters of both fact and law (at least the latter being anathema to Australian judges and commentators). That is, the courts are not to regard themselves as the ultimate paragons of wisdom on all matters, although there may be times when the decision- maker’s reasons or decision is simply unreasonable, and the courts must intervene.   219  Ibid at paragraph 48. This line of reasoning has been reinforced by Celgene Corp v Canada (Attorney General) [2011] 1 SCR 3, which presented a golden opportunity for the court to return to a pre-CUPE analysis of jurisdictional error if it had so wished. It did not take the opportunity, and reinforced the principle that jurisdictional errors will be rare. 220  Ibid. 221  David Dyzenhaus, “The Politics of Deference: Judicial Review and Democracy”, in Michael Taggart, ed, The Province of Administrative Law, Oxford: Hart Publishing, 1997 at 286.  79  This definition does leave some matters unresolved. What is meant by “reasons which could be offered in support of a decision”? Just how much “weight” should the views of the decision- maker carry? Should the weight vary according to the nature of the decision under review? If so, does that bring us back to a “sliding scale” of reasonableness? Cases since Dunsmuir have continued to reject the “sliding scale” approach, despite the best efforts of Binnie and Deschamps JJ 222 , and I would argue that this approach is misguided and empties the concept of “reasonableness” of much of its content.  “Correctness” is defined as follows223:  As important as it is that courts have a proper understanding of reasonableness review as a deferential standard, it is also without question that the standard of correctness must be maintained in respect of jurisdictional and some other questions of law. This promotes just decisions and avoids inconsistent and unauthorized application of law. When applying the correctness standard, a reviewing court will not show deference to the decision maker’s reasoning process; it will rather undertake its own analysis of the question. The analysis will bring the court to decide whether it agrees with the determination of the decision maker; if not, the court will substitute its own view and provide the correct answer. From the outset, the court must ask whether the tribunal’s decision was correct.  Note that “jurisdictional [issues] and some other questions of law” are still be to be reviewed on a correctness basis. Most significantly, “[a]dministrative bodies must also be correct in their determinations of true questions of jurisdiction or vires”224. Statements to this effect have caused  222  Alberta Teachers, supra n179 at paragraphs 85 and 86. 223  Supra n7 at paragraph 50. 224  Ibid at paragraph 59.  80  some concern amongst Canadian commentators, as will be seen shortly. Bastarache and LeBel JJ sum up at paragraph 64 225 :  The [standard of review] analysis must be contextual. As mentioned above, it is dependent on the application of a number of relevant factors, including: (1) the presence or absence of a privative clause; (2) the purpose of the tribunal as determined by interpretation of enabling legislation; (3) the nature of the question at issue, and; (4) the expertise of the tribunal. In many cases, it will not be necessary to consider all of the factors, as some of them may be determinative in the application of the reasonableness standard in a specific case.  This list of factors is very similar to that given in the 1998 case of Pushpanathan v Canada (Minister of Citizenship and Immigration) 226 . That is, despite the majority discarding the use of the term “pragmatic and functional review”227, which was the term of choice in that case, it would appear that the Pushpanathan factors will inform the choice of standard of review under the new “standard of review analysis”.  Finally, Bastarache and LeBel JJ go on to find that the appropriate standard of review of the arbitrator’s decision was reasonableness, and that the arbitrator has acted unreasonably in his interpretation of s.97(2.1) of the Public Sector Labour Relations Act and his decision to reinstate Dunsmuir to his position. The decisions of the lower courts were therefore upheld 228 .     225  Ibid at paragraph 64. 226  Supra n170 at paragraphs 29-38. 227  Supra n7 at paragraph 63. 228  The majority also found that dismissal of public servants is a matter to be dealt with by the terms of their contract of employment, if one exists, and not by general principles of administrative law. This aspect of the decision is outside the scope of this thesis.  81  3.3.3 Just moving traffic? The judgement of Binnie J  Binnie J wrote a concurring judgement, which he noted in particular that the arbitrator “was right to be conscious of the impact of his decision on the appellant 229 , but he stretched the law too far in coming to his rescue”230. Binnie J therefore agreed that the arbitrator had acted unreasonably and that the lower courts’ rulings were correct.  Overall, Binnie J regarded the majority as taking an overly formalistic approach to the determination of the standard of review. At paragraph 154, Binnie J notes that “[t]he problem is that courts have lately felt obliged to devote too much time to multi-part threshold tests instead of focussing on the who, what, why and wherefor of the litigant’s complaint on its merits”. Note again the clear statement that the merits of the decision are in question. The “merits”, or the particular circumstances of the applicant in the case at bar, is obviously an important issue for Binnie J, as can be seen at paragraph 138:  In our recent jurisprudence, the “nature of the question” before the decision- maker has been considered as one of a number of elements to be considered in choosing amongst the various standards of review. At this point, however, I believe it plays a more important role in terms of substantive review. It helps to define the range of reasonable outcomes within which the administrator is authorized to choose.  Binnie J here is clearly attempting to provide an answer as to how to decide whether a particular administrative decision falls within the range of “possible, acceptable outcomes” to which the  229  This statement in itself can be seen as requiring a form of “continuum” of reasonableness, in that Binnie J is here emphasising that courts must consider the impact of the decision on the individual in question. 230  Supra n7 at paragraph 157.  82  majority refers at paragraph 47. One can do this by examining the nature of the issue that is before the decision-maker, and then comparing the issue in question to the outcome arrived at.  Binnie J also makes a number of more specific complaints about the majority approach. Firstly, the majority judgement does not seem concerned with bodies other than administrative tribunals 231 . In other words, Binnie J states that one factor that the majority has omitted from their standard of review analysis is the nature of the decision-maker itself. In particular, what level of deference should be given to bodies that create policy, rather than simply adjudicate disputes? A number of commentators, including David Mullan 232 , have expressed agreement with Binnie J on this point.  This leads neatly to the second criticism made of the majority by Binnie J – that it is no longer possible to argue that there is no “sliding scale” of reasonableness, despite the rejection of this argument in Ryan 233 . Binnie J states as follows at paragraph 149:  [A] single “reasonableness” standard will now necessarily incorporate both the degree of deference formerly reflected in the distinction between patent unreasonableness and reasonableness simpliciter, and an assessment of the range of options reasonably open to the decision maker in the circumstances, in light of the reasons given for the decision. Any reappraisal of our approach to judicial review should, I think, explicitly recognize these different dimensions to the “reasonableness” standard.  Binnie J adds as follows at paragraph 150:   231  Ibid at paragraph 121. 232  David Mullan, Recent Developments in Administrative Law (21 April 2008). at 11 233  Ryan, supra n180.  83  I agree with my colleagues that “reasonableness” depends on the context234. It must be calibrated to fit the circumstances. A driving speed that is “reasonable” when motoring along a four-lane interprovincial highway is not “reasonable” when driving along an inner city street. The standard (“reasonableness”) stays the same, but the reasonableness assessment will vary with the relevant circumstances.  Binnie J is even clearer on the point at paragraph 152:  In Law Society of New Brunswick v Ryan 235 , for example, the Court rejected the argument that “it is sometimes appropriate to apply the reasonableness standard more deferentially and sometimes less deferentially depending on the circumstances” (para 43). It seems to me that collapsing everything beyond “correctness” into a single “reasonableness” standard will require a reviewing court to do exactly that.  If the “context” of the decision is a relevant factor in determining the standard of review, as conceded by Bastarache and LeBel JJ, it is difficult to see how it can be argued that there is no “continuum” or “spectrum” of reasonableness. In my view, this is the most significant flaw in the majority judgement, and Binnie J should be preferred on this point.  Thirdly, at paragraph 148, Binnie J picks up the fact that Bastarache and LeBel JJ referred to the term “irrationality” at paragraph 46. Binnie J notes as follows:  When, then, should a decision be deemed “unreasonable”? My colleagues suggest a test of irrationality (para 46), but the editors of de Smith point out that “many decisions which fall foul of [unreasonableness] have been coldly rational” (de Smith, Woolf & Jowell: Judicial Review of Administrative Action (5th ed. 1995), at para 13-003). A decision meeting this description by this Court is CUPE v Ontario (Minister of Labour) [2003] 1 SCR 539, where the Minister’s appointment of retired judges with little experience in labour matters to chair “interest” arbitrations (as opposed to “grievance” arbitrations) between hospitals and hospital workers was “coldly rational” in terms of the Minister’s own agenda,  234  See in particular Dunsmuir, supra n7, at paragraphs 64 and 74. 235  Supra n180.  84  but was held by a majority of this Court to be patently unreasonable in terms of the history, object and purpose of the authorizing legislation. He had not used the appointment power for the purposes for which the legislature had conferred it.  On one hand, Binnie J is correct to point out the mixed terminology in the majority’s judgement. On the other hand, a “coldly rational” decision could still be unreasonable in the Dunsmuir sense, especially in the sense that the decision-maker failed to take all relevant factors into account. For example, one way to defend the decision in CUPE v Ontario (Minister of Labour) 236  is that the Minister failed to consider the fact that historically, appointments made to labour tribunals tended to be individuals who had particular backgrounds or expertise, which was lacking in the retired judges appointed by the Minister in that case. While CUPE v Ontario is one of the less convincing decisions made by Binnie J, it does illustrate that a decision can be strictly logical and rational, and yet unreasonable. Such a decision may well be reasonable but not proportional, and the difference between the two concepts will be discussed in more detail in the next chapter of this thesis.  Finally, Binnie J notes that the presence or absence of a privative clause should be a highly persuasive, although not determinative, factor in deciding the standard of review, as can be seen at paragraph 143:  The existence of a privative clause is currently subsumed within the “pragmatic and functional” test as one factor amongst others to be considered in determining the appropriate standard of review, where it supports the choice of the patent unreasonableness standard. A single standard of “reasonableness” cannot mean that the degree of deference is unaffected by the existence of a suitably worded privative clause … A system of judicial review based on the rule of law ought not to treat a privative clause as conclusive, but it is more than just another “factor” in the hopper of pragmatism and functionality. Its existence should presumptively  236  Supra n98.  85  foreclose judicial review on the basis of outcome on substantive grounds unless the applicant can show that the clause, properly interpreted, permits it or there is some legal reason why it cannot be given effect.  While Bastarache and LeBel JJ put the presence or absence of a privative clause as the first factor in their standard of review analysis 237 , it is not clear whether these factors are intended to be ordered in any way. Binnie J clearly states that they should be considered in a hierarchy, with the privative clause as the primary factor. It is hard to disagree with the idea that a legislative indication that courts should be slow to intervene in a particular field should be anything other than the first thing that a court should look for.  3.3.4 The “just the facts” approach: Deschamps J  Deschamps J also wrote a concurring judgement, on her own behalf and for Charron and Rothstein JJ. Her approach is summed up in the first paragraph of her judgement 238 :  The law of judicial review of administrative action not only requires repairs, it needs to be cleared of superfluous discussions and processes. This area of the law can be simplified by examining the substance of the work courts are called upon to do when reviewing any case, whether it be in the context of administrative or of appellate review. Any review starts with the identification of the questions at issue as questions of law, questions of fact or questions of mixed fact and law. Very little else needs to be done in order to determine whether deference needs to be shown to an administrative body.  Deschamps J here seems to be saying that in the end, all that really needs to be done to determine the standard of review is to examine the question that was before the decision-maker. In her  237  Supra n7 at paragraph 64. 238  Ibid at paragraph 158.  86  view, findings of fact and mixed fact and law always attract deference 239 , while determinations of law will attract deference or not depending on the existence or not of a privative clause, any statutory appeal, and whether the decision-maker is interpreting a matter of law within its “core expertise”240.  Deschamps J would have found the appropriate standard of review to be correctness, on the basis that the Public Sector Labour Relations Act did not apply to Dunsmuir, and that the issue was therefore really the common law interpretation of his contract 241 . Deschamps J found that the adjudicator had only considered the legislation and not common law rules of interpretation of contracts, and that his decision was therefore incorrect 242 .  Deschamps J would also have confined the entire Dunsmuir judgement to adjudicative tribunals 243 . While it is true that the majority’s analysis is best applicable to adjudicative decisions and not formulations of broad policy, it is important to note that it is not only administrative tribunals that make such determinations. Ministerial delegates may also make adjudicative decisions, although it is frequently the case that such decisions are reviewed by tribunals before they reach the court (such as the review of Citizenship and Immigration Canada decisions by the Immigration Appeal Board). Delegates will also be at least as expert in their field as tribunal members in many cases, as they will see more applications.   239  Ibid at paragraphs 161 and 164 respectively. 240  Ibid at paragraphs 162 to 163. 241  Ibid at paragraph 168. 242  Ibid at paragraphs 169 to 171. 243  Ibid at paragraph 165.  87  Finally, Deschamps J also agrees with Binnie J that “reasonableness” can only be viewed as a sliding scale of sorts. At paragraph 167 Deschamps J comments as follows:  The problem with the definitions resides in attempts by the courts to enclose the concept of reasonableness in a formula fitting all cases. No matter how this Court defines this concept, any context considered by a reviewing court will, more often than not, look more like a rainbow than a black and white situation. One cannot change this reality.  I have already argued that this is the only way that any unitary standard of reasonableness can be interpreted, and I therefore agree with Deschamps J on this point.  3.4 Dunsmuir reviewed  While Dunsmuir may not have solved all of Canada’s administrative law problems, it is a beneficial decision. The distinction between the two reasonableness standards was always untenable and needed to be abolished. It also sets out some important insights as to what deference is and when it should be applied.  3.4.1 What Dunsmuir has achieved  In my view, each of the judgements has its strengths and weaknesses, but overall the judgement of Binnie J is the best constructed. The main strength of the majority judgement is that Bastarache and LeBel JJ provide the clearest discussion of what deference actually is. By focusing on deference as “respectful attention”, Bastarache and LeBel JJ make it clear that deference is not the same as obedience or obsequiousness, but merely an acknowledgement of  88  the expertise of the decision-maker. The final decision on the interpretation of the decision- maker’s enabling legislation will remain with the court, despite the fact that it may decide not to interfere with the interpretation of the decision-maker. Read in this way, the violent Australian reaction against any notion of deference becomes even more difficult to understand, particularly when one considers that deference to the decision-maker is merely one aspect of the whole concept of substantive review.  3.4.2 The “spectrum of reasonableness”  Binnie J recognises (as does Deschamps J) that a single standard of reasonableness can only ever amount to a “spectrum”. By referring to the “context” of an administrative decision, Bastarache and LeBel JJ seem to implicitly concede this point, but appear reluctant to overrule earlier authority such as Ryan, in the same way that Dickson J was reluctant to expressly overrule Metropolitan Life in CUPE.  The most common objection to a “sliding scale” of review is the argument that was made by Iaccobucci J in Law Society of New Brunswick v Ryan, namely that “[i]f the standard of reasonableness could ‘float’ this would remove the discipline involved in judicial review: courts could hold that decisions were unreasonable by adjusting the standard towards correctness instead of explaining why the decision was not supported by any reasons that can bear a somewhat probing examination”244. Judicial review in and of itself has elsewhere been described  244  Supra n180 at paragraph 46.  89  as a “discipline”245, which suggests that judges always and everywhere are meant to exercise some kind of restraint in their dealings with administrative decisions.  In my view, this argument is unfounded for a number of reasons. Firstly, the concept of deference itself serves as a “brake” on judicial activism, especially when one considers that Dunsmuir is expressly not an administrative law “ground zero”. Previous decisions on when to defer and when not to defer remain good law. Second, if a court is really determined to intervene it will find a way to regard a decision as either unreasonable or incorrect, depending on the standard of review, in any event. Thirdly, a sliding scale is simply implicit in the notion of reasonableness – for example, refusing a person a visitor visa on the basis of a minor criminal conviction may well be reasonable, but refusing a person protection as a refugee for the same reason would be entirely unreasonable. There is simply no way to escape the idea that what is reasonable will depend on the circumstances of the case.  Binnie J himself seemed to recant from his views on the “sliding scale” somewhat in Canada (Citizenship and Immigration) v Khosa 246 , where he wrote the leading judgement, but this particular issue was conspicuous by its absence. Binnie J did revive the “sliding scale” once more, however, in his concurring reasons in Alberta (Information and Privacy Commissioner) v Alberta Teachers’ Association247. Writing for himself and Deschamps J, Binnie J stated that “it is clear that ‘the range of acceptable and rational solutions’ is context specific and varies with the  245  Thomas Poole, “The Reformation of English Administrative Law”, (2009) 68 Cambridge Law Journal 142 at 142. 246  Supra n29 at paragraph 61. 247  Supra n179.  90  circumstances including the nature of the issue under review”248, and that in some cases there may be a greater or lesser “intensity of scrutiny”249 or “aggressiveness”250 within the reasonableness standard. With the retirement of Binnie J from the bench, and the imminent retirement of Deschamps J 251 , it remains to be seen whether any other judge will be prepared to run with this argument in the future.  3.4.3 “True questions of jurisdiction or vires”  One of the major criticisms of Dunsmuir is that it could be seen as resuscitating the Metropolitan Life type approach to jurisdictional errors. David Mullan, for example, argues that paragraph 59 of Dunsmuir, which states that “[a]dministrative bodies must also be correct in their determinations of true questions of jurisdiction or vires”, resurrects the possibility that courts could again have the opportunity to class most matters as jurisdictional in nature 252 . In my opinion, Mullan’s concerns were unfounded then and have been demonstrated to be so since. The issue is not so much whether some questions are jurisdictional in nature, but whether the expertise of the decision-maker will be respected when considering his, her or its jurisdiction. That is, while a decision-maker certainly cannot be permitted to conclusively determine his or her own jurisdiction, their views on whether they have the jurisdiction to make a certain decision should be respected.   248  Ibid at paragraph 85. 249  Ibid. 250  Ibid at paragraph 86. 251  “Supreme Court Justice retires, giving Harper chance to appoint majority”, The Globe and Mail, 18 May 2012. 252  Mullan, “Let’s Try Again”, supra n208 at 129-30.  91  While Dunsmuir certainly left open some ambiguities, most particularly on the issue of when a “true question of jurisdiction or vires” exists, cases since Dunsmuir have kept up a generally deferential approach to administrative decision-making, and have not resuscitated the much- criticised Metropolitan Life kind of reasoning. While a detailed examination the post-Dunsmuir cases is beyond the scope of this thesis, cases since Dunsmuir have, if anything, widened the kinds of decisions in which deference will be shown, up to and including constitutional (or at least Charter) issues 253 . In particular, in Alberta (Information and Privacy Commissioner) v Alberta Teachers’ Association Rothstein J, writing for the majority, found that Dunsmuir established a presumption that reasonableness is the applicable standard of review, unless the question in dispute relates to either constitutional law 254 , a question of central importance to the legal system as a whole that it outside the expertise of the decision-maker, the question relates to jurisdictional lines between two or more competing tribunals, or that the question is one of “true jurisdiction or vires”255. Rothstein J also went so far as to say that “it may be that the time has come to reconsider whether, for purposes of judicial review, the category of true questions of jurisdiction exists and is necessary to identifying the appropriate standard of review”256. While the court in that case did not go so far as to actually abolish the concept of “questions of jurisdiction”, the fact that the issue was even raised should reassure commentators that there is not going to be any return to Metropolitan Life in the near future.    253  Doré, supra n21. 254  On this point now see Doré, supra n21. 255  Supra n179 at paragraphs 54 and 57-61. See also John Evans, “Standards of Review in Administrative Law”, paper prepared for the 2012 British Columbia Administrative Law Conference at 3. 256  Supra n179 at paragraph 42.  92  3.4.4 To what should courts defer?  One other issue that may have been left unresolved by Dunsmuir is to what courts are supposed to defer. Firstly, is there a hierarchy or just an enumerated list of factors? Bastarache and LeBel JJ did not put their “standard of review analysis” factors in any particular order, while Binnie J stated that the presence or absence of a privative clause is the primary (although not determinative) consideration 257 . Secondly, are courts deferring to a privative clause (if any), the expertise of the decision-maker, or the reasons for the decision?  In my opinion, if the key issue is legislative intent, a privative clause is the clearest possible form of statutory indication that courts should defer to the decision-maker, and should be given primacy in the manner suggested by Binnie J. I would also argue that the expertise of the decision-maker is also part of legislative intent – there would be little point in creating a tribunal or delegating one or more decision-makers with the power to make a particular administrative decision if they were not going to be made by persons with some expertise in the field. Indeed, in Nor-Man Regional Health Authority Inc v Manitoba Association of Health Care Professionals Fish J noted that labour arbitrators “benefit from institutional expertise … even if they lack personal expertise”258. Similarly, Rothstein J stated in Alberta (Information and Privacy Commissioner) v Alberta Teachers’ Association that “through the creation of administrative  257  It remains an unresolved question as to whether the existence of a statutory appeal should be taken as inviting review by the courts, and whether this should lead to courts being more willing to impose a correctness standard. Then one faces the situation of a statutory appeal accompanied by a privative clause – should this situation result in the correctness or reasonableness standard being applied? 258  [2011] SCC 59 at paragraph 53  93  tribunals, legislatures confer decision-making authority on certain matters to decision-makers who are assumed to have specialised expertise with the assigned subject matter”259.  David Mullan has noted that “what was once … the most important factor among those constituting the former ‘pragmatic and functional’ criteria for determining the appropriate standard of review has now in itself become in effect a presumption”260. While Mullan has elsewhere, with some justification, argued that “at least some sectors of the administrative process as a reward or retirement home for political hacks”261, expertise on the part of decision- makers should be assumed. A decision-maker who regularly makes incorrect or unreasonable decisions may well demonstrate a lack of expertise, but the decisions should be reviewed on their merits, not those of the individual decision-maker.  The question of reasons is a particularly interesting one, especially when one considers what Bastarache and LeBel might have meant when they referred to “reasons offered or which could be offered in support of a decision”262. This remark is still somewhat cryptic, but the Supreme Court in Newfoundland and Labrador Nurses’ Union v Newfoundland and Labrador (Treasury Board) did make it clear that Dunsmuir does not require a separate analysis of the reasons for the decision and the result – instead, “the reasons must be read together with the outcome and serve the purpose of showing whether the result falls within a range of possible outcomes”263. That is, the purpose of reasons is to “allow the reviewing court to understand why the tribunal made its  259  Supra n179 at paragraph 1 260  David Mullan, “Developments in Judicial Review of Administrative Action – 2011-12”, paper prepared for the 2012 British Columbia Administrative Law Conference at 11 261  Mullan, “Deference: Is It Useful Outside Canada?”, supra n125 at 52. 262  Supra n7 at paragraph 48. 263  [2011] SCC 62 at paragraph 14.  94  decision and permit it to determine whether the conclusion is within the range of acceptable outcomes”264.  A similar statement was made in Halifax (Regional Municipality) v Canada (Public Works and Government Services), where Cromwell J, writing for the court, stated that “[r]easonableness review is concerned both with the transparency and intelligibility of the reasons given for a decision and with the outcome of the decision-making process”265. That is, reasons will be assessed to determine whether they explain why decision in question was made, and if the outcome falls within the “possible, acceptable outcomes” they will be sufficient. How one is to determine whether a particular outcome is possible and acceptable is something that the Supreme Court may still need to explain another day.  One final issue to be raised here is that even if a court upholds the decision of an administrative decision-maker, it may still be showing a marked lack of deference if it proceeds to rewrite the reasons of the decision-maker, on the basis that there were other reasons that “could have been given” in support of the decision. If the court in Baker266 had rewritten the reasons given by Mr Lorenz to remove the inflammatory language and cultural stereotyping, and upheld the decision on the basis of the “new and improved” reasons, would that have been an example of “deference” or not? This slightly condescending attitude towards the reasons given by administrative decision-makers can also be seen in the Australian cases, discussed in Chapter 6 of this thesis.    264  Ibid at paragraph 16. 265  Supra n15 at paragraph 44. 266  Supra n60.  95  3.5 “Substantive review” and “deference” in Canada defined  “Substantive review”, from the preceding discussion, can be defined as a form of judicial review whereby the reviewing court examines the substance of the decision under review, not merely the procedures by which it was made. In some cases, as most recently articulated by Dunsmuir, a court will take the view that there is only one correct answer to a particular question, or that there is only one correct decision that can be made, in a particular case. In this situation, the court will determine what that correct answer is. Not surprisingly, this “standard of review” is known as correctness.  However, in most cases there will be more than one correct decision possible – that is, there will be a range of “possible, acceptable outcomes” that the decision-maker could reach. In that case the decision-maker will “defer” to the opinion of the decision-maker, which does not mean that their opinion must be accepted. Instead, deference is a judicial acknowledgement of the fact that Parliament had decided that a particular decision-making power should be exercised by a particular person or body, and that the person or body appointed to make the decision usually has some particular expertise in their field. Determining the area of expertise of the person or body in question is often a difficult task, and has not, in my opinion been carried out in a consistent manner. In particular, there is a peculiar unwillingness by the Supreme Court to defer to the opinion of human rights tribunals, the justification usually being that these tribunals are not expert in interpreting international conventions such as the International Covenant on Civil and  96  Political Rights, and the importance of the decisions to the individuals in question 267 . On the other hand, the expertise of labour relations boards is rarely questioned 268 . Nevertheless, it is difficult to argue with the principle that the “democratic credentials” and expertise of decision- makers should not at least be taken into account when reviewing their decisions.  It is important to note, therefore, that deference is only one part of the “package” of substantive review. As I will argue later, Australian commentators who have vehemently rejected any concept of deference in Australian administrative law appear to have taken the term out of context, and looked at it in isolation rather than as one part of a coherent system of judicial review. Having defined what the terms “substantive review” and “deference” mean, we can now turn to examine how these concepts, even if not these terms, are applied in other common law jurisdictions, and how their application is similar and different to the Canadian approach.   267  As just one example see Mossop, supra n9. More recent decisions have begun to give some deference to human rights tribunals, so far only in making of costs orders – see for example Canadian Human Rights Commission v Canada (Attorney-General) [2011] SCC 53 and Smith v Alliance Pipeline [2011] 1 SCR 160 – but at least a start has been made. 268  Davies, supra n96.  97  CHAPTER FOUR DEFERENCE IN THE UNITED KINGDOM  As would have already become apparent in this thesis, the history of administrative law in the UK is long and detailed. However, UK law has evolved significantly since the seminal 1985 decision of Council of Civil Service Unions v Minister for the Civil Service 269  (popularly known as the “GCHQ case”), and this brief examination of the UK law of judicial review will focus on the changes in the law since this judgement 270 . It is my contention that since the GCHQ case, UK administrative law has been moving towards a system similar to that of Canada’s, albeit with different terminology, in which courts impose one standard of review for administrative decisions that impact on rights protected by the European Convention on Human Rights (ECHR), and a less stringent standard for other decisions.  Thomas Poole has summed the current position up as follows 271 :  We are witnessing a reconfiguration in the law of judicial review, intimations of which can be found on the surface of both the cases and the commentary. The language of Wednesbury unreasonableness and ultra vires increasingly gives way to talk about rights, proportionality and deference. This semantic recasting of judicial review reflects deeper mutations that go to the very heart of the discipline.   The UK position is somewhat complicated by the fact that the Human Rights Act 1998 272  has incorporated the ECHR into UK law. The result has been that substantive review of  269  [1984] 3 All ER 935. 270  It is convenient to point out now that British courts and commentators have occasionally used the term “super- Wednesbury” in referring to the variable standard of reasonableness review. I will not use this term in this thesis because, as Philip Joseph (“The Demise of Ultra Vires – Judicial Review in New Zealand Courts”, [2001] Public Law 354 at 359) has pointed out, the term has not been used consistently – sometimes it refers to a high degree of deference and sometimes to the lowest degree. It is therefore best to avoid it altogether. 271  Poole, supra n245 at 142.  98  administrative decision-making in the UK is expressed to be on different bases depending on the kind of law in question. When considering EU laws applicable in the UK, or UK laws expressly implementing EU laws in Britain (such as the Human Rights Act), British courts have undertaken a form of proportionality review common to European legal systems, which involves using a concept of deference similar to Canada. In cases not involving any form of EU law, British courts have moved to a “sliding scale” of reasonableness, which in some ways resembles the different Canadian standards of review, but in other ways resembles the “sliding scale” championed by Binnie J in Dunsmuir v New Brunswick 273  and Alberta (Information and Privacy Commissioner) v Alberta Teachers’ Association274. The question that must be answered is whether there is in reality any difference between the two forms of review.  4.1 Reasonableness and irrationality – the GCHQ case  The GCHQ case is notable for its restatement of the grounds of judicial review in the UK. In this case, the government attempted to introduce a policy whereby staff of the General Communications Headquarters, a crucial inter-governmental communications agency (and probably spy agency), were no longer permitted to be members of a trade union. The Council of Civil Service Unions (CCSU) sought judicial review of the decision, arguing that the union had a legitimate expectation 275  that it would be consulted before any such decision was made, and that no such consultation had occurred.  272  Supra n20. 273  Supra n7. 274  Supra n179. 275  A detailed explanation of the concept of legitimate expectations is beyond the scope of this thesis, but it should be noted that in the UK, a legitimate expectation can be used to protect substantive rights – R v North and East  99   The House of Lords found that despite the lack of any statutory requirement to consult, the union would in fact generally have a legitimate expectation that it would be consulted before any decision adverse to its interests was made. However, no such requirement existed when national security issues were at stake, and this was one of those situations 276 . The CCSU therefore lost its case, but did succeed in creating a legal duty to consult in most cases.  For the purposes of this thesis, however, the key part of the judgement can be found in the judgement of Lord Diplock. His Lordship stated as follows 277 :  … [O]ne can conveniently classify under three heads the grounds on which administrative action is subject to control by judicial review. The first ground I would call “illegality”, the second “irrationality” and the third “procedural impropriety” …  By “illegality” as a ground for judicial review I mean that the decision-maker must understand correctly the law that regulates his decision-making power and must give effect to it. Whether he has or not is par excellence a justiciable question to be decided, in the event of dispute, by those persons, the judges, by whom the judicial power of the state is exercisable.  By “irrationality” I mean what can by now be succinctly referred to as “Wednesbury unreasonableness” (see Associated Provincial Picture Houses Ltd v Wednesbury Corp [1948] 1 KB 223). It applies to a decision which is so outrageous in its defiance of logic or of accepted moral standards that no sensible person who had applied his mind to the question to be decided could have arrived at it …  I have described the third head as “procedural impropriety” rather than failure to observe basic rules of natural justice or failure to act with procedural fairness towards the person who will be affected by the decision. This is because susceptibility to judicial review under this head covers also failure by an  Devon Health Authority, ex parte Coughlan [2001] QB 213. However, in Canada only procedural expectations are protected – Baker, supra n135. 276  Supra n269 at 944 – 945. 277  Ibid at 950-1.  100  administrative tribunal to observe procedural rules that are expressly laid down in the legislative instrument by which its jurisdiction is conferred, even where such failure does not involve any denial of natural justice.  It should be noted that Wednesbury actually envisaged two kinds of unreasonableness – the use of a power for an improper purpose, of the kind exhibited in Canada by Roncarelli v Duplessis 278, or a decision that is “so absurd that no sensible person could ever dream that it lay within the powers of the authority”279. It is the latter meaning of the term that has popularly become known as “Wednesbury unreasonableness”.  Paul Craig explains the unreasonableness and irrationality concept as follows 280 :  The special interpretation of the term “reasonableness” was warranted by the constitutional position of the courts. They should not intervene simply because they believed that a different way of exercising discretionary power would be more reasonable than that chosen by the public body … Hence the controls over the substantive ends that can be pursued by an administrative authority are expressed in terms of relevancy, purpose or unreasonableness in its substantive sense. By phrasing control in these terms, the courts preserve the impression that they are thereby only fulfilling the legislative will. They will not dictate which result should be reached, but they will impose limits on which ends cannot be pursued.   The use of the word “impression” suggests that the author does not believe that the courts always do simply “fulfil the legislative will”, but this text could be taken directly from any Canadian administrative law textbook 281 .  278  Supra n158. 279  Supra n14 at 229, per Lord Greene MR. Paul Craig refers to this as “substantive unreasonableness” – Craig, Administrative Law (6 th  ed), Sweet and Maxwell, 2008 at paragraph 17-002 (p 532). 280  Craig, ibid, at paragraph 17-002 (p 533). 281  It is also notable that Diplock LJ in the GCHQ case also predicted the possible introduction of a proportionality ground of judicial review. His Lordship stated at 950 as follows:   101   “Wednesbury unreasonableness” is not, therefore, a “man on the Clapham omnibus” test, but an issue going to the constitutional division of responsibilities between the courts and legislature. It is not sufficient that the “reasonable person” would regard a decision as unreasonable, and instead it must be so unreasonable that it could not be an exercise of the power that was intended by the Parliament.  4.2 Varying the Wednesbury principle – “anxious scrutiny”282  Despite the GCHQ case equating “irrationality” with “Wednesbury unreasonableness”, it was only two years later that we can see the first hint of a “sliding scale” of reasonableness. One possible reason for this is that the British courts essentially made a compromise between the traditional unreasonableness test and the “proportionality” test commonly employed in continental Europe, which was becoming more important for British law through the EU. Julian Rivers, employing a good deal of deference-type terminology, explains the concept as follows 283 :  The idea of variable judicial restraint is straightforward. We have seen that restraint operates to preserve to non-judicial bodies a range of necessary/efficient options … A large degree of restraint means that the court will be very unwilling to question the view of the primary decision-maker that what is necessary to achieve a certain level of public interest is also balanced. A moderate degree of  That is not to say that further development on a case by case basis may not in course of time add further grounds. I have in mind particularly the possible adoption in the future of the principle of “proportionality” which is recognised in the administrative law of several of our fellow members of the European Economic Community; but to dispose of the instant case the three already well- established heads that I have mentioned will suffice.  I will return to the meaning of “proportionality” later in this thesis.  282  Referring to Budgaycay v Secretary of State for the Home Department [1987] AC 514 at 531. 283  Julian Rivers, “Proportionality and Variable Intensity of Review”, (2006) 65 Cambridge Law Journal 174 at 203.  102  restraint means that the court will want to check that the costs and gains are indeed roughly commensurable. A small degree of restraint will reduce the set of necessary decisions to a minimum; the court will need to be convinced itself that the decision, rule or policy in question, even though necessary, really is the best way of optimising the relevant rights and interests.  Substitute the terms “patent unreasonableness”, “reasonableness simpliciter” and “correctness” for Rivers’ degrees of restraint, and we can see that he takes the view that UK courts can apply a degree of “deference” in the same way as the Canadian courts after Southam284 and  before Dunsmuir. Note also his reference to a “range of necessary / efficient options”, which could also have been taken directly from Dunsmuir 285 .  4.2.1 Application of the “anxious scrutiny” principle  In the 1987 case of Budgaycay v Secretary of State for the Home Department, the House of Lords was concerned with a deportation order issued against the applicant. The case was argued on the Wednesbury unreasonableness ground, but the House of Lords, allowing the application, stated as follows 286 :  [T]he most fundamental of human rights is the individual’s right to life and when an administrative decision under challenge is said to be one which may put the applicant’s life at risk, the basis of the decision must surely call for the most anxious scrutiny.   In other words, the more important the right at stake, the more carefully scrutinised an administrative decision will be. It is unlikely that Lord Greene in Wednesbury had a decision of  284  Supra n172 285  Supra n7 at paragraph 47. 286  Supra n282 at 531.  103  the kind considered in Budgaycay in mind when he gave his famous example of the dismissal of a teacher because of her red hair as an example of the kind of outrageous decision that would be regarded as “absurd”287. A decision will be more likely to be found to be outrageous and unsupportable when a fundamental right is impacted. Note that this kind of language was picked up in Canadian law at least as early as Baker v Canada (Minister of Citizenship and Immigration), in which L’Huereux-Dubé J, writing for the majority, stated that “a reasonable exercise of the power conferred by the section requires close attention to the interests and needs of children”288, and that “[c]hildren’s rights, and attention to their interests, are central humanitarian and compassionate values in Canadian society”289.  The “anxious scrutiny” terminology was called upon in two cases in the 1990s, both of which predated the Human Rights Act. In R v Secretary of State for the Home Department, ex parte Brind 290 , the House of Lords considered directives made under the Broadcasting Act 1981 preventing broadcasting of statements by persons representing groups that had been proscribed as terrorist organisations. Lord Bridge noted that there was not (at that time) any bill of rights under domestic UK law, but went on to state as follows 291 :  This surely does not mean that in deciding whether the Secretary of State, in the exercise of his discretion, could reasonably impose the restriction he has imposed on the broadcasting organisations, we are not perfectly entitled to start from the premise that any restriction on the right to freedom of expression requires to be justified, and nothing less than an important competing public interest will be sufficient to justify it.   287  Supra n14 at 229. 288  Supra n60 at paragraph 67. 289  Ibid. 290  [1991] 1 AC 696. 291  Ibid at 748-9.  104   While Brind did not go quite as far as some other cases which found there was really no difference between the common law and the ECHR 292 , it is obvious that the judgement in that case stands for the proposition that where fundamental rights are involved, the courts will not wait for a “red-haired teachers” type of situation before intervening.  In R v Ministry of Defence, ex parte Smith 293  a challenge was brought against the then-existing policy of discharging known homosexuals from the armed forces. Quoting Budgaycay, the House of Lords found that “the more substantial the interference with human rights, the more the court will require by way of justification before it is satisfied that the decision is reasonable”294, but was not prepared to find that the decision was unreasonable, given that it impacted on matters of military discipline and potentially national security. It is also noteworthy that in Smith, Sir Thomas Bingham LJ, in appeal to expertise that would be familiar to Canadian administrative lawyers, noted as follows 295 :  The greater the policy content of a decision, and the more remote the subject matter of a decision from ordinary judicial experience, the more hesitant the court must necessarily be in holding a decision to be irrational. That is good law and, like most good law, common sense. Where decisions of a policy-laden, esoteric or security-based nature are in issue even greater caution than normal must be shown in applying the test, but the test itself is sufficiently flexible to cover all situations.   The applicants in Smith later took their case to the European Court of Justice, and more will be said of the result later.  292  See for example Attorney-General v Guardian Newspapers (the “Spycatcher case”) [1990] 1 AC 109 and Derbyshire County Council v Times Newspapers Ltd [1993] AC 534. 293  [1996] QB 517. 294  Ibid at 554. 295  Ibid at 556.  105   One can see in this judgement a link, also made in Canadian law, between reasonableness (or “rationality” to use the term currently in use in the UK), justification and reasons. A decision will be reasonable if it can be justified in all the circumstances, based on an examination of the written reasons for that decision. In Newfoundland and Labrador Nurses’ Union v Newfoundland and Labrador (Treasury Board) the Supreme Court has now made clear that “reasons must be sufficient to permit the parties to understand why the tribunal made the decision and to enable judicial review of that decision”296. Reasons are sufficient if they “allow the reviewing court to understand why the tribunal made its decision and permit it to determine whether the conclusion is within the range of acceptable outcomes”297  A good example of the application of the irrationality principle can be seen in the Cambridge Health Authority cases. R v Cambridge District Health Authority, ex parte B 298  was concerned with a decision of the health authority to cease cancer treatments for a 10-year-old girl, after her treatments to date had been unsuccessful. Laws J, in the court of Queen’s Bench, stated that “the law requires that where a public body enjoys a discretion whose exercise may infringe [a basic liberty], it is not to be permitted to perpetrate any such infringement unless it can show a substantial objective justification on public interest grounds”299, and found that since no such justification had been made out – note again the significance of the adequacy of reasons for the decision – the decision was “irrational”.   296  Supra n263 at paragraph 11. 297  Ibid at paragraph 18. 298  [1995] 1 FLR 1055. 299  Ibid at 1060.  106  On appeal, the Court of Appeal found that courts were “not the arbiters of the merits in such cases”300, and could not express any views on the likelihood of success of further treatment. Sir Thomas Bingham MR stated as follows 301 :  I have no doubt that in a perfect world any treatment which a patient, or a patient’s family, sought would be provided if doctors were willing to give it, no matter how much it cost, particularly when a life was potentially at stake. It would however, in my view, be shutting one’s eyes to the real world if the court were to proceed on the basis that we do live in such a world. It is common knowledge that health authorities of all kinds are constantly pressed to make ends meet … Difficult and agonising judgments have to be made as to how a limited budget is best allocated to the maximum advantage of the maximum number of patients. That is not a judgment which the court can make.   This is probably the kind of outcome that Binnie J had in mind when he spoke of “coldly rational” decisions in Dunsmuir302. In a similar judgement, Lord Woolf MR held in R v Lord Saville of Newdigate, ex parte A that the term “irrational” may not really be appropriate, given that the decision-maker “could be the most rational of persons”303, which sounds rather like damning with faint praise. Despite all this, it is certainly difficult to argue that the Health Authority’s decision, given the myriad considerations it had to take into account, was irrational, and Ian Turner for one has defended it 304 . This case remains a good illustration of the limits of the rationality approach to judicial review, and demonstrates the greater degree of deference given to a decision-maker using an “irrationality” approach to judicial review (at least in UK law), than in using a “variable reasonableness” approach.   300  Craig, supra n279, at paragraph 19-001 (pp 613-4). 301  [1995] 2 All ER 129 at 135-6. 302  Supra n7 at paragraph 148. 303  [1994] 4 All ER 860 at paragraph 33. 304  Ian Turner, “Judicial Review, Irrationality, and the Legitimacy of Merits-Review”, (2008) 29 Liverpool Law Review 309 at 313.  107  4.2.2 Cases where “anxious scrutiny” is not necessary  Turner has also identified a number of situations where courts will be reluctant to find that a decision of an administrator is “irrational”305. The first is where matters relating to raising and spending public revenue are involved. For example, in Nottinghamshire County Council v Secretary of State for the Environment the council challenged the Secretary’s decision in relation to awarding of environmental improvement grants. In this instance, Lord Scarman defined unreasonableness as “a pattern of perversity or absurdity of such proportions that the guidance could not have been framed by a bona fide exercise of political judgement”306. This wording goes close to saying that when considering matters where the courts lack “expertise, information or accountability”, in the words of Nicholas Blake in commenting on the case307, a decision will only be set aside if it is so perverse that an inference of bad faith can be drawn.  Another is where the exercise of wide discretionary powers is in issue. R v Chief Constable of Sussex, ex parte International Trader’s Ferry Ltd308 involved a decision of the Chief Constable to provide police protection only two days per week for trucks attempting to run a blockade by animal rights protesters at the port of Shoreham. Livestock trucks would be turned back on other days. The House of Lords found this decision was not irrational because of the wide discretionary powers given to the Chief Constable, and it deferred to his expertise 309 :   305  Ian Turner, “Judicial Review, Irrationality and the Limits of Intervention by the Courts”, [2010] 21 Kings Law Journal 311 at 322-3. 306  Nottinghamshire County Council v Secretary of State for the Environment [1986] 2 AC 240 at 248. 307  Nicholas Blake, “Importing Proportionality: Clarification or Confusion”, [2002] European Human Rights Law Review 19 at 26. 308  [1999] 2 AC 418. 309  Ibid at 430.  108  The courts have long made it clear that, though they will readily review the way in which decisions are reached, they will respect the margin of appreciation or discretion which a Chief Constable has. He knows through his officers the local situation, the availability of officers and his financial resources, the other demands on the police in the area at different times … Where the use of limited resources has to be decided, the undesirability of the court stepping in too quickly has been made very clear by the courts.   Again, this kind of reasoning would look familiar to Canadian lawyers, in that the court recognised that the Chief Constable had expertise in matters relating to policing and police resources that the court did not, and declined to interfere with his decision as a result.  UK courts have also shown deference to decision-makers in areas that would be familiar to Canadian lawyers – wide discretionary powers and expertise. One might argue why wide discretionary powers should attract the degree of deference they do, because of the impact that such decisions can have on the persons affected by them. In Canada, Doré v Barreau du Québec 310  has simultaneously recognised that decision-makers who regularly make decisions that impact on Charter values may have some expertise in the interpretation of the Charter, and that review of administrative decisions affecting such values will require a form of proportionality analysis. UK and Canadian courts also face the problem of determining what a decision-maker is expert in, if anything. For example, is an officer who regularly makes decisions on visa applications expert in international conventions that regularly impact on whether such a visa should be granted, or only in the enabling legislation itself?      310  Supra n21.  109  4.2.3 Use of proportionality in pre-Human Rights Act cases?    There has also been the occasional case where a court has apparently used a proportionality analysis in reaching its decision prior to the coming into effect of the Human Rights Act, and even when no ECHR right is in issue. In R v Coventry City Council, ex parte Phoenix Aviation 311 , another animal rights-related case, Simon Brown LJ considered whether a decision by the Council to severely restrict exports of live animals out of its airport, after a number of disruptions caused by animal rights protesters, had been lawful. His Lordship stated that “the council’s resolution was … disproportionate to the security risk presented at the time”312 and found it to be unlawful. Ian Turner writes, and it is hard to disagree with him, that “these words suggest that the judge may have engaged in a greater review of the merits of the council’s decision than normal, because he was employing a different test of review to irrationality, namely proportionality”313.  Despite this apparent use of the proportionality test in a non-fundamental rights setting, this approach has, as will be seen, still not caught on with the British judiciary. The question of whether there is really any difference between the expanded “irrationality” basis of judicial review used in cases other than those involving the Human Rights Act 1998, and the “proportionality” approach, will be discussed at the end of this chapter.    311  [1995] 3 All ER 37 (QB). 312  Ibid at 63. 313  Turner, “Limits of Intervention by the Courts”, supra n305 at 313.  110  4.2.4 Indicia of deference – common features and key differences  From the discussion above, we can see a number of common features in when and how Canadian and UK courts defer to administrative decision-makers. Firstly, “fundamental rights” are given greater weight and sanction greater intensity of review. That is, the greater the impact of the decision on the applicant, the greater the scrutiny given to the decision. Secondly, courts in both countries recognise the necessity of justification to show deference, by focusing on the adequacy or otherwise for the reasons for the decision. A related consideration is that the justification for the decision should be in the public interest. The decision-maker in R v Cambridge District Health Authority, ex parte B 314 , for example, was shown deference because of the polycentric nature of the decision in question (not just the applicant’s health care was in issue) and the evidence before the decision-maker.   Thirdly, courts emphasise the importance of relative expertise to show deference. That is, the decision-maker must be expert is something that the courts are not (or at least the court is no more expert that the decision-maker) before deference can be given. The courts must also demonstrate their expertise by attending to justifications, and also proper balancing of considerations required, taking into account legislative intent (something at which the courts ought to be expert).  Fourthly, exercise of a discretionary power involves a “margin of appreciation”, but it is not unlimited. This is what deference is all about – respect for the democratic credentials and  314  Supra n298.  111  expertise of the decision-maker, but not subservience to him, her or it. Both Canadian and UK courts frequently look for a range of plausible, acceptable, reasonable or efficient outcomes, and determine whether the decision in question falls within those parameters.  One issue where Canadian and UK courts differ is whether reasonableness amounts to a spectrum or continuum. As we have seen, despite the efforts of Binnie and Deschamps JJ, the other members of the Supreme Court still refuse to admit what should really be obvious, that a concept of reasonableness only makes sense if it is considered as a sliding scale depending on the issues in question (although Doré v Barreau du Québec 315  may have changed this approach for cases in which the Charter is in question). The UK, by using the “variable reasonableness” approach in non-HRA cases and proportionality in HRA cases, has rejected this approach, and it is to the issue of proportionality that we will now turn.  4.3 The Human Rights Act 1998  The Human Rights Act has been a major influence on the development of British administrative law, and requires brief examination. Section 1 of the Act defines the term “convention rights” in terms of a number of rights set out in the EHCR and a number of protocols, and the term “convention” is defined in s.21 as “the Convention for the Protection of Human Rights and Fundamental Freedoms, agreed by the Council of Europe at Rome on 4th November 1950 as it has effect for the time being in relation to the United Kingdom”. The Human Rights Act therefore incorporates the ECHR, at least in part, into domestic British law.  315  Supra n21.  112   Unlike the Canadian Charter, the Human Rights Act does not permit a court to invalidate primary legislation, but a court can issue a “declaration of incompatibility” under s.4 of the Act. The most important provision for the purposes of this thesis is s.6:  (1) It is unlawful for a public authority to act in a way which is incompatible with a Convention right.  (2)  Subsection (1) does not apply to an act if— (a)  as the result of one or more provisions of primary legislation, the authority could not have acted differently; or (b)  in the case of one or more provisions of, or made under, primary legislation which cannot be read or given effect in a way which is compatible with the Convention rights, the authority was acting so as to give effect to or enforce those provisions.  (3)  In this section “public authority” includes— (a)  a court or tribunal, and (b)  any person certain of whose functions are functions of a public nature 316 , but does not include either House of Parliament or a person exercising functions in connection with proceedings in Parliament.  A number of commentators have argued that the Human Rights Act has transformed British administrative law from a focus on procedure and rationality on the part of the decision-maker to a focus on the rights of the person affected. Thomas Poole writes as follows 317 :  For the era we are now entering is marked by a much more direct and frequent recourse to arguments about rights – especially but not exclusively those of the European Convention on Human Rights (“ECHR”) … While there had been an increase in rights talk in cases like Bugdaycay 318 , Witham 319  and Smith 320  only the  316  The term “functions of a public nature” is not defined in the Human Rights Act (supra n20), and the term “tribunal” would most likely have the same meaning today as it does in the Tribunals, Courts and Enforcement Act 2007 (2007 c 15). The meaning of these terms will be discussed as necessary. 317  Poole, supra n245 at 145. 318  Supra n282. 319  R v Lord Chancellor, ex parte Witham [1998] QB 575. 320  Supra n293.  113  introduction of the HRA facilitated the kind of deep, structural change we have seen since.   Canada has gone through the same process with the Charter – only the introduction of the Charter has caused a definitive shift from a “jurisdictional analysis” approach of the kind used in Metropolitan Life 321 , to a rights-based approach that is particularly obvious in Doré v Barreau du Québec 322 . Cheryl Saunders, writing on a number of common law systems, including the UK, makes a similar point 323 :  Many of the old technical limitations on the scope of judicial review are going, if they have not already gone. The concept of justiciability is dwindling and the range of justiciable questions is correspondingly wider; the identity of the decision-maker has become less important in determining standards of review; the requirements for standing have been significantly relaxed, to the extent that they remain a factor at all; the simplification of the grounds of review has effectively expanded them; and the dividing line between jurisdictional or non-jurisdictional error is blurred or abandoned. The once key distinction between legality and merits, always precarious in the face of claims about the unreasonableness of executive action is eroding, partly through pressure on the concept of reasonableness itself but most particularly as the principle of proportionality gains ground. There is a range of explanations for these developments. Most obviously, they are linked with the rise of rights, encompassing … the reconceptualisation of administrative justice itself in human rights terms.   That is, the idea of “rights-based” administrative law jurisprudence is not something unique to the UK. Rights-based administrative law is modern administrative law and it is happening in all liberal-democratic countries to varying extents, with implications for all common law countries.   321  Supra n16. 322  Supra n21. 323  Cheryl Saunders, “Apples, Oranges and Comparative Administrative Law”, [2006] Acta Juridica 423 at 442.  114  4.4 Irrationality after the Human Rights Act  4.4.1 Definition of “rationality”  There does not appear to have been any comprehensive restatement of the rationality principle since the Human Rights Act came into effect. That is, the rationality ground of review is still a “so unreasonable that no reasonable person could come to it” ground, although the nature of the right impacted on will be a consideration in determining when a decision is taken to be unreasonable. However, the UK courts, in pursuing rationality review, have not taken the Canadian approach that there are clear and discrete “standards of review” – instead, there is a spectrum or continuum of reasonableness. In R (Mahmood) v Secretary of State for the Home Department 324  Laws LJ referred to the “anxious scrutiny” test and stated at paragraph 19 as follows:  … that approach and the basic Wednesbury rule are by no means hermetically sealed the one from the other. There is, rather, what may be called a sliding scale of review; the graver the impact of the decision in question upon the individual affected by it, the more substantial the justification that will be required.  At least in cases where there are no unqualified rights involved, the UK courts appear to take the view that a decision will not be irrational if there is some evidence to support it. For example, in R v Johns ex parte Derby City Council the High Court found that the decision to exclude Mr and Mrs Johns as foster carers was not irrational in the GCHQ sense, on the basis that there was considered to be “a body of opinion which considers that a child or young person who is homosexual or is doubtful about his or her sexual orientation may experience isolation and fear  324  [2001] 1 WLR 840.  115  of discovery if their carer is antipathetic to or disapproves of homosexuality or same-sex relationships”325 and that “material also indicates that there is support in the literature for the view that those who hide their sexual orientation or find it difficult to ‘come out’ may have more health problems and in particular mental health problems”326. The fact that other evidence may have supported the Johns’ capacity to be foster carers was neither here nor there.  4.4.2 Moves towards a single test of judicial review?  Despite (as we shall see) the acceptance of the proportionality approach in cases concerning the Human Rights Act and other EU laws applicable in the UK, British courts have continued to apply the irrationality test to other matters of substantive review. There have been a number of cases in which courts have suggested that the end of the irrationality approach is nigh, or even desirable, but there has not yet been any definitive move to do away with the doctrine altogether. For example, in R (Association of British Civilian Internees – Far East Region) v Secretary of State for Defence 327 , Dyson LJ noted that the application of an irrationality test will often (although not always) yield the same result as a proportionality analysis, and that courts had previously argued that proportionality should be accepted as the approach for all matters of substantive review 328. However, his Lordship then added that “it is not for this court to perform its [irrationality’s] burial rites”329. The House of Lords refused leave to appeal against this decision, leaving the irrationality ground intact and a clearly annoyed Michael Fordham QC  325  [2011] EWCA 375 (Admin) at paragraph 106. 326  Ibid. 327  [2003] QB 1397. 328  Ibid at paragraphs 33 and 34. In particular Dyson LJ referred to the speech of Lord Slynn in R (Alconbury Ltd) v Secretary of State for the Environment, Transport and the Regions [2001] UKHL 23, and the speech of Lord Cooke at paragraph 32 of Daly, supra n12. 329  Ibid at paragraph 35.  116  commenting that “the Law Lords, failing to take the hint, refused permission to appeal, and put the development of the common law back a decade”330.  Another early attempt to expand proportionality review to all administrative decisions, R (on the application of Prolife Alliance) v British Broadcasting Corporation 331 , involved a refusal by the BBC to screen a party political broadcast on behalf of the Alliance. The broadcast would have involved images of aborted foetuses and abortion techniques, and the BBC decided that such images would be too graphic and disturbing for its audience. Lord Walker stated as follows at paragraph 144:  The Wednesbury test, for all its defects, had the advantage of simplicity, and it might be thought unsatisfactory that it must now be replaced (when human rights are in play) by a much more complex and contextually sensitive approach. But the scope and reach of the Human Rights Act is so extensive that there is no alternative. It might be a mistake, at this stage in the bedding-down of the Human Rights Act, for your Lordships’ House to go too far in attempting any comprehensive statement of principle. But it is clear that any simple “one size fits all” formulation of the test would be impossible.  In other words, while Wednesbury had to be extended to cover a variable scale of review, in a fashion very similar to that proposed by Binnie J in Dunsmuir 332  and Alberta (Information and Privacy Commissioner) v Alberta Teachers’ Association333, the Lords were not prepared to move to a (then) little-tested proportionality regime for all administrative decisions.  Similarly, in Doherty v Birmingham City Council 334 , a case which raised both common law and Human Rights Act issues, the House of Lords again found that a universal “proportionality” test for review of  330  Michael Fordham, “Wednesbury”, (2007) Judicial Review 66 at paragraph 6. 331  [2004] 1 AC 185. 332  Supra n7 at paragraphs 150-152. 333  Supra n179 at paragraphs 85 and 86. 334  [2009] 1 AC 367.  117  all administrative decisions in the UK should not be introduced. This was despite the comment by Lord Walker that human rights “must be woven into the fabric of public law”335 and a number of observations by Lord Mance. At paragraph 135 Lord Mance states as follows:  The difference in approach between the grounds of conventional or domestic judicial review and review for compatibility with Human Rights Convention rights should not however be exaggerated and can be seen to have narrowed, with “the ‘Wednesbury’ test … moving closer to proportionality [so that] in some cases it is not possible to see any daylight between the two tests” (ABCIFER336, para 34). The common law has been increasingly ready to identify certain basic rights in respect of which “the most anxious” scrutiny is appropriate … Even so, as the subsequent history of ex parte Smith demonstrates, the result may not always achieve the degree of protection for Convention rights which the Strasbourg Court requires: Smith and Grady v United Kingdom (1999) 29 EHRR 493. So there remains room in another case to reconsider how far conventional or domestic judicial review and Convention review can be further assimilated, and in particular whether proportionality has a role in conventional judicial review.  There are three key points in this paragraph. Firstly, the difference between Wednesbury and proportionality should not be exaggerated, and there is often no “daylight” between the tests, especially where “anxious scrutiny” is involved. Secondly, Wednesbury unreasonableness and proportionality remain distinct tests, and the former may not necessarily provide the same level of protection from administrative action. Finally, future cases may require further convergence of the tests. Like Lord Diplock in the GCHQ case, the court is unwilling to rule out the possibility of a universal standard of proportionality in the future, but is not yet prepared to take this step. One reason for not taking this step immediately appears to be that the interpretation of the Human Rights Act is not yet “bedded down” – once there is a clear and comprehensive jurisprudence on the use of the proportionality test, the time may then come to extend it to all judicial review applications in the UK.  335  Ibid at paragraph 109. 336  British Civilian Internees, supra n327.  118   A similar kind of analysis can be seen in Canadian cases dealing with review of administrative decisions on Charter grounds. In Slaight Communications Inc v Davidson 337  Lamer J found that the Charter applies to administrative decision-making and applied the Oakes test 338 , but stated that “[t]he precise relationship between the traditional standard of administrative law review of patent unreasonableness and the new constitutional standard of review will be worked out in future cases”339. In Multani v Commission Scolaire Marguerite-Bourgeoys340 the Supreme Court took the view that the Oakes analysis was to be applied to all administrative law cases involving the Charter, and that Charter findings of administrative bodies will always be reviewed on a correctness basis. However, only six years later the Supreme Court has now appeared to come to the view that “legitimacy”341 has been given to administrative decision-making and that decision- makers can now be trusted to consider Charter issues in the course of their deliberations, without a correctness standard of review always hanging over their heads. This is yet more evidence of a convergence between UK and Canadian administrative law.  4.4.3 The role of deference  The British courts have shown some move towards the Canadian deference and reasonableness approach in recent years, when applying the “irrationality” approach to substantive review. In R v Chief Constable of Sussex ex parte International Trader’s Ferry Ltd, Lord Cooke stated that a  337  [1989] 1 SCR 1038. 338  R v Oakes, supra n194. 339  Supra n337 at 1049. 340  Supra n54. 341  Doré, supra n21 at paragraph 35.  119  test of “unreasonableness” or “irrationality” should be rephrased so that the question becomes “was the decision one that a reasonable authority could have reached?”342, which is obviously getting close to the Dunsmuir reasonableness standard, arrived at nine years later. Even more interestingly, despite some suspicion about the term “deference” from a number of British judges 343 , in a dissenting judgement in International Transport Roth GmbH v Secretary of State for the Home Department 344 , Laws LJ expressly sought to introduce a doctrine of deference into UK law, specifying a list of relevant factors that would determine how much deference was given. Anne Davies sums up the relevant factors as follows 345 :  First, more deference should be shown to Parliament than to the executive. Second, more deference should be shown in cases involving qualified rights than in cases involving unqualified rights. Third, more deference should be shown to decisions that had the seal of democratic approval than to those which did not, and fourth, more deference should be shown where the decision-maker had expertise in the relevant subject matter.  Davies suggests replacing the term “deference” with “judicial restraint”, believing that this term has less of a servile overtone than the Canadian term of choice 346 . Writing specifically about labour law, Davies notes that courts have deferred to labour tribunals on the basis that they frequently make polycentric decisions, the expertise of the tribunal and the use to which that expertise is put, and the “democratic accountability” issue, referring to the fact that the  342  Supra n308 at 452. 343  See for example Prolife Alliance, supra n331, in which Lord Hoffmann stated at paragraph 75 that “although the word ‘deference’ is now very popular in describing the relationship between the judicial and the other branches of government, I do not think that its overtones of servility, or perhaps gracious concession, are appropriate to describe what is happening”. 344  [2002] All ER (D) 325 at paragraphs 81 – 87. 345  Davies, supra n96 at 285. 346  Ibid at 278. As will be seen in the next chapter of this thesis, some New Zealand judges have expressed similar concerns.  120  Parliament has created specialised labour law bodies, whose role should be respected by the courts 347 . None of this reasoning would be out of place in Canada.  The terms “deference” and “margin of appreciation” are, however, used more frequently in cases where a proportionality analysis is undertaken. We will now consider these cases.  4.5 Proportionality  4.5.1 Origins of the principle  Proportionality is a form of judicial review that began in continental Europe, and has been “transplanted” into the UK as a result of the Human Rights Act, and other legislation that has incorporated European law into the UK. Margit Cohn explains the origins of proportionality as follows 348 :  The principle of proportionality (Verhaltnismaβigkeitsgrundsatz) is central to German public law … The principle is now applied as an independent and perhaps the most important and extensive umbrella ground for examining the validity of administrative actions … In its current form, the formula created by German courts comprises three subtests or limbs. First, the measure must be suitable for the achievement of the aim pursued. Secondly, no other milder means could have been employed to achieve that aim (a “necessity” test). Finally, under a proportionality stricto sensu test, a type of cost-benefit analysis is required; for the measure to be upheld, the benefit at large must outweigh the injury to the implicated individual    347  Ibid at 290-1. 348  Margit Cohn, “Legal Transplant Chronicles: The Evolution of Unreasonableness and Proportionality Review of Administration in the United Kingdom”, (2010) 58 American Journal of Comparative Law 583 at 608-9.  121  That is, under a proportionality analysis, the court must effectively determine whether the decision was justified in terms of its objectives. The question could almost be rephrased as “are the objectives justifiable, and do the ends justify the means”?  The prompt for the introduction of the proportionality principle into UK law, at least where the Human Rights Act is concerned, was the decision of the European Court of Human Rights in Smith and Grady v United Kingdom 349 . Having been unsuccessful before the UK courts, the applicants from R v Ministry of Defence ex parte Smith 350  took their case to the European courts and were successful. The European court found that Smith’s and Grady’s rights under Article 8 of the ECHR had been infringed, and that although Article 8 is a “qualified right”, the Ministry of Defence could not justify the breach. Most relevantly for the purposes of this thesis, the Court found that the UK reasonableness test, even applying the “anxious scrutiny” test, was insufficient, and stated as follows 351 :  The threshold at which the High Court and the Court of Appeal could find the Ministry of Defence policy irrational was placed so high that it effectively excluded any consideration by the domestic courts of the question of whether the interference with the applicants’ rights answered a pressing social need or was proportionate to the national security and public order aims pursued, principles which lie at the heart of the Court’s analysis of complaints under Article 8 of the Convention.   In other words, when considering rights provided for by the ECHR, “irrationality” is too high a standard for a court to have to reach. Only a proportionality approach is sufficient.   349  (1999) 29 EHRR 493. 350  Supra n293. 351  Supra n349 at 543.  122  4.5.2 Differences between the irrationality and proportionality approaches  The difference between the irrationality and proportionality approaches is usually explained as the latter requiring an additional step in analysis. In De Freitas v Permanent Secretary of Ministry of Agriculture, Fisheries, Lands and Housing, Lord Clyde stated that a court, in applying a proportionality analysis, needed to consider the following three issues 352 :  (i) the legislative objective is sufficiently important to justify limiting a fundamental right; (ii) the measures designed to meet the legislative objective are rationally connected to it; and (iii) the means used to impair the right or freedom are no more than is necessary to accomplish the objective.  The contrast with the irrationality ground was more clearly expounded on in R v Secretary of State for the Home Department, ex parte Daly 353 , which is now generally regarded as the leading case on proportionality review, and which will be considered further later in this chapter. Lord Steyn stated as follows at paragraph 27:  Most cases would be decided in the same way whichever approach is adopted. But the intensity of review is somewhat greater under the proportionality approach … I would mention three concrete differences without suggesting that my statement is exhaustive. First, the doctrine of proportionality may require the reviewing court to assess the balance which the decision maker has struck, not merely whether it is within the range of rational or reasonable decisions. Secondly, the proportionality test may go further than the traditional grounds of review inasmuch as it may require attention to be directed to the relative weight accorded to interests and considerations. Thirdly, even the heightened scrutiny test developed in R v Ministry of Defence, ex p Smith [1996] QB 517, 554 is not necessarily appropriate to the protection of human rights.   352  [1999] 1 AC 69 at 80. 353  Supra n12.  123  Ian Turner comments that “[t]he proportionality test is clearly a searching method of review and projects courts more into the role of the primary decision-maker than the irrationality test, which simply requires the decision maker to remain within an area of reasonable response”354. Turner’s analysis sounds very much like the “possible, acceptable outcomes” language of Dunsmuir355. It looks more and more as if the British rationality approach can be equated with Canadian post- Dunsmuir reasonableness, and proportionality as something akin to analysis under s.1 of the Charter. Post- Doré, Canada may now have the same approach – proportionality, a principle that always informed the Oakes 356  test, when Charter rights are in issue, and either reasonableness or correctness analysis when they are not.  The exact difference between proportionality and a “variable unreasonableness” analysis is not particularly clear, and indeed in Daly Lord Steyn admitted that “[m]ost cases would be decided in the same way whichever approach is adopted”357. Lord Steyn added as follows358:  This [the shift to proportionality analysis] does not mean that there has been a shift to merits review. On the contrary … the respective roles of judges and administrators are fundamentally distinct and will remain so … Laws LJ rightly emphasised in Mahmood 359 , at p 847, para 18, “that the intensity of review in a public law case will depend on the subject matter in hand”. That is so even in cases involving Convention rights. In law context is everything.  It is notable that Lord Steyn denies that courts engage in “merits review”, of which more will be said later. In response to Lord Steyn, Lord Cooke reflects on the difference between proportionality and unreasonableness in paragraph 32:  354  Turner, “Legitimacy of Merits Review”, supra n304 at 314. 355  Supra n7 at paragraph 47. 356  Supra n194 at paragraph 70. 357  Supra n12 at paragraph 27. 358  Ibid at paragraph 28. 359  Supra n324.  124   Lord Steyn illuminates the distinctions between “traditional” … standards of judicial review and higher standards under the European Convention or the common law of human rights. As he indicates, often the results are the same. But the view that the standards are substantially the same appears to have received its quietus in Smith and Grady v United Kingdom (1999) 29 EHRR 493 and Lustig- Prean and Beckett v United Kingdom (1999) 29 EHRR 548. And I think that the day will come when it will be more widely recognised that Associated Provincial Picture Houses Ltd v Wednesbury Corporation [1948] 1 KB 223 was an unfortunately retrogressive decision in English administrative law, insofar as it suggested that there are degrees of unreasonableness and that only a very extreme degree can bring an administrative decision within the legitimate scope of judicial invalidation. The depth of judicial review and the deference due to administrative discretion vary with the subject matter. It may well be, however, that the law can never be satisfied in any administrative field merely by a finding that the decision under review is not capricious or absurd.  Lord Cooke’s criticism of the concept of “degrees of unreasonableness” is very redolent of the merging of the two standards of Canadian unreasonableness in Dunsmuir at paragraphs 41-46 of that judgement. Both jurisdictions now refer to a single standard of reasonableness, although so for only the UK courts have been prepared to admit that reasonableness is a “spectrum”.  Bradley Selway QC (later Selway J of the Federal Court of Australia) comments on the Mahmood and Daly judgements as follows 360 :  [I]n Mahmood Laws LJ identified two possible bases for judicial review: “Upon the question, ‘What is the correct standard of review in a case such as this?’, there are at least in theory [two possible common law approaches]. The first is the conventional Wednesbury position ... On this model the court makes no judgment of its own as to the relative weight to be attached to this or that factor taken into account in the decision-making process; it is concerned only to see that everything relevant and nothing irrelevant has been considered, and that a rational mind has been brought to bear by the Secretary of State in reaching the decision. The second approach recognises that a fundamental right, here family life, is engaged in the case; and in consequence the court will insist that that fact be respected by  360  Bradley Selway, “The Principle Behind Common Law Judicial Review of Administrative Action – The Search Continues”, (2002) 30 Federal Law Review 217 at 223-4.  125  the decision-maker, who is accordingly required to demonstrate either that his proposed action does not in truth interfere with the right, or if it does, that there exist considerations which may reasonably be accepted as amounting to a substantial objective justification for the interference.” In that case Laws LJ adopted the second approach.  The Mahmood analysis was accepted in Daly. In consequence of this new broad approach the House of Lords has now acknowledged that the principle of proportionality is recognised and applied by the English common law.  Note Selway’s reference to the existence of two possible “standards of review”, and also his comment that simple irrationality review does not involve a consideration of the “weight” given by decision-makers to relevant factors. Again, the resemblance to Canadian principles of judicial review is marked.  4.5.3 “Merits review”  British courts remain insistent that they do not undertake “merits review” of administrative decisions. The exact difference between merits and judicial review is not always – perhaps never – clear, but the former Australian Solicitor-General, David Bennett QC, has defined the terms as follows 361 :  A merits review body will “stand in the shoes” of the primary decision-maker, and will make a fresh decision based upon all the evidence available to it. The object of merits review is to ensure that the “correct or preferable”362 decision is made on the material before the review body. The object of judicial review, on the other hand, is to ensure that the decision made by the primary decision-maker was properly made within the legal limits of the relevant power.    361  David Bennett, “Balancing Merits Review and Judicial Review”, (2003) 53 Administrative Review 3 at 7. 362  Referring to Drake v Minister for Immigration and Ethnic Affairs (1979) 2 ALD 60 at 68.  126  “Merits review”, on this definition, is a very wide term, ranging from a formal internal review of a decision, to a quasi-judicial hearing before a formally constituted tribunal such as the Canadian Human Rights Tribunal, but not including proceedings before a court. More will be said of the distinction, if indeed any really exists, between merits and judicial review later in this chapter of this thesis, and in the chapter devoted to Australian judicial review.  One illustration of the difference between merits and judicial review, as explained by the British courts, will suffice. Huang v Secretary of State for the Home Department 363  may be an atypical case, because it dealt with a statutory appeal rather than true judicial review, but is nonetheless worthy of consideration. In this case, Mrs Huang, a failed applicant for humanitarian stay in the UK, appealed against the Home Department’s decision to an “adjudicator”, as permitted by s.65 of the Immigration and Asylum Act 1999. That Act permitted a further appeal to the Court of Appeal from the adjudicator’s findings on a question of law. Lord Bingham, writing for the House of Lords, found that the adjudicator, by focusing on whether there was an error in the original decision, did not fulfil their role. His Lordship stated that 364 :  It remains the case that the judge is not the primary decision-maker … The appellate immigration authority, deciding an appeal under section 65, is not reviewing the decision of another decision-maker. It is deciding whether or not it is unlawful to refuse leave to enter or remain, and it is doing so on the basis of up to date facts.  His Lordship added as follows at paragraph 15:  The first task of the appellate immigration authority is to establish the relevant facts. These may well have changed since the original decision was made. In any  363  [2007] 2 AC 167. 364  Ibid at paragraph 13.  127  event, particularly where the applicant has not been interviewed, the authority will be much better placed to investigate the facts, test the evidence, assess the sincerity of the applicant’s evidence and the genuineness of his or her concerns and evaluate the nature and strength of the family bond in the particular case.   That is, the appellate authority had acted in too “judicial” a manner in this case, and should have considered Mrs Huang’s case de novo rather than simply examining the primary decision- maker’s decision for any errors. It is the court that is prohibited from engaging in “merits review”.  4.5.4 Deference and the “margin of appreciation”  In the European legal system, courts have granted what is known as a “margin of appreciation” to administrative decision-makers when reviewing their decisions. This concept has become known as the “area of discretion” under EU law. Margit Cohn explains this concept as a recognition by the EU courts that member States will be more attuned to what is happening “on the ground” in their own country, and therefore often being in a better position to explain whether a breach of a qualified right is necessary and proportionate than an international judge 365 . While the House of Lords in R v DPP ex parte Kebilene 366  and the Prolife Alliance case 367  rejected any direct transplant of the European concept into UK law, a domestic concept of deference has taken root nevertheless. Proportionality is generally seen as a more searching kind of analysis than mere unreasonableness (even variable unreasonableness) review, and it “may require the reviewing court to assess the balance which the decision maker has struck, not merely  365  Cohn, supra n348 at 612-3. 366  [2000] 2 AC 326. 367  Supra n331.  128  whether it is within the range of rational or reasonable decisions”368. Further analysis is beyond the scope of this thesis, but it would appear that, if anything, UK administrative law goes further than Canadian law in its willingness to examine the “merits” of an administrative decision.  Some cases have expressly stated that the difference between merits review and judicial review is that the latter affords a degree of deference to the decision-maker that the former does not. For example, the House of Lords stated in Tweed v Parades Commission for Northern Ireland as follows 369 :  In addressing the critical question in any proportionality case as to whether the interference with the right in question is objectively justified, it is the court’s recognition of what has been called variously the margin of discretion, or the discretionary area of judgment, or the deference or latitude due to administrative decision-makers, which stops the challenge from being a merits review. The extent of this margin will depend, as the cases show, on a variety of considerations and, with it, the intensity of review appropriate in the particular case.   The House of Lords had earlier made a similar statement in A v Secretary of State for the Home Department, where Lord Nicholls stated as follows 370 :  The duty of the courts is to check that legislation and ministerial decisions do not overlook the human rights of persons adversely affected. In enacting legislation and reaching decisions Parliament and ministers must give due weight to fundamental rights and freedoms. For their part, when carrying out their assigned task the courts will accord to Parliament and ministers, as the primary decision- makers, an appropriate degree of latitude. The latitude will vary according to the subject matter under consideration, the importance of the human right in question, and the extent of the encroachment upon that right. The courts will intervene only when it is apparent that, in balancing the various considerations involved, the  368  Supra n12 at paragraph 27. 369  [2007] 1 AC 650 at paragraph 55. This case also stated at paragraph 36 that the word “latitude” is preferable to “deference”. 370  Supra n106 at paragraph 80.  129  primary decision-maker must have given insufficient weight to the human rights factor.   That is, the stated difference between pure merits review and judicial review on the proportionality ground is that a judge may not simply substitute their decision for that of the primary decision-maker, where an administrative tribunal can and indeed sometimes must (such as in Huang 371 ). Instead, some degree of deference must be given to the primary decision-maker.  4.5.5 Deference illustrated – the Denbigh High School case  The British approach to deference can be encapsulated by examining two cases. In R (on the application of Begum) v Headteacher and Governors of Denbigh High School 372 , the substantive issue was whether the school’s uniform policy breached the student’s Article 9(1) rights when it refused her permission to wear a particular form of Islamic dress known as a “jilbab” (other forms of Islamic dress were permitted). The House of Lords took the view that Parliament had left such decisions to schools, and that those schools were the “experts” in what was acceptable or required in their local area. Lord Bingham stated as follows at paragraph 33:  The respondent criticised the school for permitting the headscarf while refusing to permit the jilbab, for refusing permission to wear the jilbab when some other schools permitted it and for adhering to their own view of what Islamic dress required. None of these criticisms can in my opinion be sustained. The headscarf was permitted in 1993, following detailed consideration of the uniform policy, in response to requests by several girls. There was no evidence that this was opposed. But there was no pressure at any time, save by the respondent, to wear the jilbab, and that has been opposed. Different schools have different uniform policies, no doubt influenced by the composition of their pupil bodies and a range of other matters. Each school has to decide what uniform, if any, will best serve  371  Supra n363. 372  [2007] 1 AC 100.  130  its wider educational purposes. The school did not reject the respondent’s request out of hand: it took advice, and was told that its existing policy conformed with the requirements of mainstream Muslim opinion.   Baroness Hale gave a broadly concurring opinion at paragraphs 97-8. In her view, the school’s dress code was “devised to meet the social conditions prevailing in the area at that time and was a proportionate response to the need to balance social cohesion and religious diversity”373. In other words, the school had a particular expertise and exercised its discretion in a reasonable and proportionate manner.  Lord Hoffmann commented at paragraph 64 of the judgement that “a domestic court should accept the decision of Parliament to allow individual schools to make their own decisions about uniforms”. This is, to use Anne Davies’ terminology374, a “democratic accountability” argument, involving respect for a democratically elected Parliament’s ability to grant decision-making power to others. It would be less than startling to Canadian lawyers. His Lordship also stated as follows at paragraph 66:  What matters is the result: was the right to manifest a religious belief restricted in a way which is not justified under article 9.2? The fact that the decision-maker is allowed an area of judgment in imposing requirements which may have the effect of restricting the right does not entitle a court to say that a justifiable and proportionate restriction should be struck down because the decision-maker did not approach the question in the structured way in which a judge might have done. Head teachers and governors cannot be expected to make such decisions with textbooks on human rights law at their elbows.    373  Craig, supra n279 at 18-045 (p597). 374  Davies, supra n96 at 290-1.  131  Taken together, these judgements illustrate the point that the school was both empowered by the Parliament to make the sort of decisions that it did, and had a better “on the ground” knowledge of prevailing conditions than the court. Its decision was therefore worthy of deference.  There is a marked resemblance between Denbigh and the Canadian decision in Multani v Commission Scolaire Marguerite-Bourgeoys 375 . Both cases involved an argument that a school authority’s dress requirements violated human rights standards, in Multani s.2(a) of the Charter. The argument in Multani was that the Commission’s refusal to allow a Sikh boy to carry a kirpan was an infringement of his freedom of religion, and the Supreme Court found that such an absolute prohibition was an infringement of s.2(a) (although some restrictions on how the kirpan could be carried were permissible).  The difference in the outcomes of the two cases might be explained on the basis that Denbigh High School was prepared to allow Begum to wear other forms of Islamic dress, and took advice from an Imam on whether its policy was acceptable to the Moslem community, while the Commission seemed determined to ban the kirpan altogether (albeit for the legitimate reason of ensuring safety in its schools). However, both courts balanced the rights of the school authority against those of the students, and came to a conclusion accordingly. Multani may have since been overruled at least in part by Doré v Barreau du Québec 376 , which now requires Canadian courts to defer to bodies making decisions on Charter values, at least where that body has some expertise in the issue. Whether the result would have been any different is another matter.   375  Supra n54. 376  Supra n21.  132  4.6 Merits review, “variable” review and proportionality – is there really any difference?  As we have already seen, the House of Lords in Daly stated that the acceptance of proportionality review for Human Rights Act issues “does not mean that there has been a shift to merits review”377. But is this in fact the case? Surely any kind of review on the basis of unreasonableness, or even patent unreasonableness, is a form of merits review. In any such case, the court is examining the substance of the decision and determining whether it meets a minimum level of reasonableness. This is so regardless of the degree of deference to be given to the primary decision-maker. It is notable that Canadian courts do not seem compelled to deal with this issue – discussions of the differences between merits and judicial review are notably absent from the leading Canadian Supreme Court cases, and Canadian judicial review is all the better for it.  The view that there is no real difference between merits and proportionality review has also been taken by a number of commentators. Bradley Selway has noted that “the new English approach clearly permits merit review subject only to whatever forbearance the judge, as a matter of policy, is prepared to give”378. Michael Taggart has made similar comments379:  First, by definition, proportionality review is much closer to merits review than variable intensity unreasonableness review, notwithstanding British denials. Secondly, the proportionality methodology is best powered by a list of enumerated rights otherwise it loses much of its much-admired analytic and structuring qualities. It applies in the UK, Canada and New Zealand in the context  377  Daly, supra n12 at paragraph 28. 378  Selway, supra n360 at 224. 379  Michael Taggart, “Australian Exceptionalism in Judicial Review”, (2008) 36 Federal Law Review 1 at 25.  133  of bills or charters of rights, as it will inevitably in those Australian jurisdictions with statutory bills of rights.  This is not to say that the abandonment of the distinction between “law” and “merits” is a bad thing. Indeed, I will argue at the end of this thesis that the distinction should be formally abandoned in both the UK, where really it has been abandoned in all but name, and in Australia, where review of the merits is seen as something other than an exercise of “judicial power”. Ian Turner has also argued that there is no real distinction between proportionality, or indeed “variable intensity irrationality” review, but comes to the conclusion that the courts would be best off returning to the Wednesbury / GCHQ kind of analysis as a result 380 :  Those individuals who support, for whatever reason, a greater degree of intensity of review by the courts over administrative decision-making arguably do not wish to witness a complete merger of the judicial and executive functions: there must still be a threshold over which a supervisory court is not permitted to step. But is the identification of such a threshold – the judicial substitution of executive judgement, for example – achievable in reality? The author of this study does not believe so … the courts should revert to more orthodox principles of public law.  The problem with Turner’s analysis is that even the orthodox Wednesbury approach is really a form of merits review, with a greater degree of deference given to the decision-maker than the “anxious scrutiny” or proportionality approaches. I do, however, agree with his assertion that it is simply impossible to identify a bright line delineating reasonableness, proportionality and merits review. To my way of thinking, they are all forms of review of the substance, or the merits, of the decision, with one the reasonableness standard of review providing a greater degree of deference to the administrative decision-maker than the proportionality. And then when we  380  Turner, “Limits of Intervention by the Courts”, supra n305 at 331.  134  consider the fact that reasonableness is itself a variegated standard in UK law, how can there be any real distinction between proportionality and a high level of “anxious scrutiny”?  A number of commentators have argued that, while it is important to distinguish between judicial and merits review, the difference, at least when undertaking a proportionality analysis, is really only one of degree, that degree being the degree of deference given to the decision-maker. Mark Aronson has commented as follows 381 :  Judicial review’s professed indifference to the substantive merits of the impugned decision is not always convincing, and not ultimately reconcilable with some of the grounds of review. (Review for “reasonableness, eg, clearly involves an examination of the impugned decision’s merits, albeit from a perspective of a large degree of deference.) But even though the difference between judicial review and merits review may at places be only one of degree, it is important to maintain that difference. Judicial deference to the views and actions of the primary decision maker is in one sense the essence of judicial review’s technique. That difference is underpinned by a political sense of the court’s secondary role in relation to the primary decision-maker, and by the practical sense of the latter institutional competence in the substantive issues relative to that of the court.   The last sentence of this quote in particular would seem very familiar to Canadians. The former Chief Justice of the High Court of Australia, Sir Anthony Mason, has commented in a similar vein 382 :  One aspect of proportionality as applied by English courts is the tendency to offer a margin of appreciation to the executive in its weighing of the competing claims of the individual and the public interest … It is the existence of this margin of appreciation accorded to the decision-maker that distinguishes proportionality from merits review. There is preserved an area of residual discretion to the decision-maker so that proportionality does not lead to the court deciding whether  381  Mark Aronson, “Responses to Privatisation and Outsourcing” in Michael Taggart (ed), The Province of Administrative Law, Oxford: Hart Publishing, 1997 at 47-8. 382  Sir Anthony Mason, “Lecture 3 – Australian Administrative Law Compared with Overseas Models of Administrative Law”, (2001) 31 Australian Institute of Administrative Law Forum 45 at 51-2.  135  the impugned decision is correct. Thus, in Daly 383  Lord Steyn felt able to say that the application of the proportionality standard “does not mean that there has been a shift to merits review”.  Julian Rivers also argues that a properly formulated standard of deference will also prevent a court from engaging in merits review 384 :  This [a proportionality approach] does not mean that the court increasingly displaces the executive and the legislature in matters of factual expertise and policy-choice. Rather, it means that the more serious a limitation of rights is, the more evidence the court will require that the factual basis of the limitation has been correctly established, and the more argument it will require that alternative, less intrusive, policy-choices are, all things considered, less desirable.  Again, can this distinction really be maintained? Does the existence of a level of deference, or a “margin of appreciation”, somehow transform merits review into judicial review? Again, I would argue that it does not. Regardless of whether any deference is given or not, the court is reviewing the merits of the decision. The only issue is whether it decides that a decision is sufficiently unreasonable or disproportionate to warrant its being set aside.  Some commentators have argued that British courts would be better off implementing a principle of due deference. The foremost proponent of this concept, Murray Hunt, has defined this term as one of deferring where deference is earned by the decision-maker, by providing comprehensive justification for their decision by way of written reasons, and demonstrating why these reasons should be upheld by the court 385 . This is a similar approach to David Dyzenhaus, who has defined deference as “respectful attention to the reasons offered or which could be offered in  383  Supra n12 at paragraph 28. 384  Supra n283 at 205. 385  Murray Hunt, “Sovereignty’s Blight: Why Contemporary Public Law Needs the Concept of Due Deference” in Nicholas Bamforth and Peter Leyland (eds) Public Law in a Multi-Layered Constitution, Hart, 2003.  136  support of a decision”386, a passage which was, as we have seen, citied with approval in Dunsmuir 387 .  It can be seen from Baroness Hale’s judgement in Denbigh that this appears to be the way that UK law is headed anyway, and it is not clear that there is any real difference between Hunt’s concept of due deference and the kind of deference currently afforded by UK courts to administrative decision-makers at present.  Finally, some commentators have argued that the entire concept of deference is irrelevant or undesirable. TRS Allan is the foremost of these writers in the UK context, and critiqued Hunt’s “due deference” concept as follows388:  A doctrine of judicial deference, I shall argue, is either empty or pernicious. It is empty if it purports to implement a separation of powers between the courts and other branches of government; that separation is independently secured by the proper application of legal principles defining the scope of individual rights or the limits of public powers. A doctrine of deference is pernicious if, forsaking the separation of powers, correctly conceived, it permits the abdication of judicial responsibility in favour of reliance on the good faith or good sense or special expertise of public officials, whose judgments about the implications of rights in specific cases may well be wrong. In its latter manifestation, judicial deference amounts to the abandonment of impartiality between citizen and state: in acceding to the supposedly superior wisdom of the public agency (or of Parliament), the court is co-opted into the executive (or the legislature), leaving the claimant without any independent means of redress for an arguable violation of rights.  Instead, in Allan’s view, courts should simply focus on the nature of the question in issue389, and whether the decision has impacted unjustifiably on a person’s rights – “the variety of context and circumstance, pertinent to the permissibility of rights restriction, will always outstrip the  386  Dyzenhaus, “Politics of Deference”, supra n221 at 286. 387  Supra n7 at paragraph 48. 388  TRS Allan, “Human Rights and Judicial Review: A Critique of ‘Due Deference’”, (2006) 65 Cambridge Law Journal 671 at 675-6. 389  This position has some resemblance to that of Deschamps J in Dunsmuir, supra n7 at paragraphs 158-164. Deschamps J would not, however, dispense with the concept of deference altogether.  137  resources of any deference doctrine, even if the latter is elaborate and complex”390. Such an approach would seemingly relieve the court of any duty to respect the decision-making capabilities of administrators.  In my opinion, this argument demonstrates a similar kind of misunderstanding of the concept of “deference” that, as we will see, is frequently made by Australian academics and judges. My research has not discovered any writer who advocates for a complete abdication of judicial responsibility by courts in favour of administrative decision-makers – not even Harry Arthurs, who having written that “[t]here is no reason to believe that a judge whose entire professional life has been spent dealing with disputes one by one should possess an aptitude for issues which arise often because an administrative system dealing with cases in volume has been designed to strike an appropriate balance between efficiency and effective rights of participation”391, probably comes closest. It is also interesting to note that the Supreme Court of Canada in Dunsmuir expressly noted that the nature of the question at hand is one of the factors to be considered in determining the standard of review, although Binnie J gave this factor a greater level of importance than Bastarache and LeBel JJ 392. Allan’s focus is therefore taken into account by at least Canadian courts, even if it is not given the degree of prominence he seeks.  When viewed as a form of “respectful attention”393 and not “servility”394 or “abdication of a constitutionally conferred judicial function”395, I would argue that deference to administrative  390  Allan, supra n388 at 52. 391  Harry Arthurs, “Protection Against Judicial Review”, (1983) 43 Canadian Bar Review 277 at 289. This passage was quoted with approval by the Supreme Court of Canada in National Corn Growers, supra n138 at 1376. 392  Dunsmuir, supra n7, paragraphs 138 (Binnie J) and 64 (Bastarache and LeBel JJ). 393  Dyzenhaus, “Politics of Deference”, supra n221; quoted with approval in Dunsmuir, supra n7 at paragraph 48. 394  Prolife Alliance, supra n331 at paragraph 75.  138  decision-makers is not only desirable but necessary for the functioning of the modern administrative state. While it is certainly true that the mere fact that a democratically elected government confers a particular decision-making power on a person or body does not automatically validate that decision, nor can that fact be completely ignored by a court. As we will see, this is a lesson that Australian commentators could also take to heart.  4.7 Summary  We have seen in this chapter that there has been a significant convergence in Canadian and UK administrative law since the 1980s. The UK was probably slower to get going on a substantive review and deference basis to judicial review – Budgaycay396 was not decided until 1987, as compared with the 1979 decision in CUPE – but both countries now use an approach involving a standard of review of reasonableness for most administrative decisions, and a proportionality approach for decisions involving fundamental rights (those protected by the Charter in Canada, and the HRA in the UK 397 ). Neither jurisdiction now attempts to argue that there is more than one standard of reasonableness. The most significant remaining difference is that the UK uses a variable scale of reasonableness in non-HRA decisions, while Canada still refuses (in the main) to admit that reasonableness is a continuum.  Both jurisdictions have moved to a rights-based approach to judicial review. This is most clearly seen in the UK in Budgaycay, where the House of Lords, instead of simply examining the  395  Hayne J, supra n3 at 75. 396  Supra n282. 397  Doré, supra n21 (Canada); Daly, supra n12 (UK).  139  powers of the decision-maker, focused on the impact of the decision on the applicant. Similar reasoning can be seen in Dunsmuir and Khosa 398  in Canada, even though the individual applicants were unsuccessful in both of those cases. A corollary of this is that when a decision does impact on fundamental rights, particularly those protected by the Charter or the HRA, the decision-maker must provide justification for doing so, by way of written reasons, specifying the evidence before him or her. This requirement can be seen most clearly in Smith 399  and Denbigh 400  in the UK, and Khosa and Newfoundland Nurses 401  in Canada.  Both jurisdictions recognise that the courts may sometimes lack expertise possessed by administrators. This has been a staple of Canadian administrative law since CUPE, and deference to expertise is particularly marked in relation to economic 402 , labour 403  and security 404  decision-makers. The House of Lords has admitted that some decisions simply cannot be made by courts 405 , and although British judges have at time expressed a dislike of the word “deference”406, and have not simply adopted the European approach, a domestic form of deference has been put in place nonetheless.  The main deficiency in UK administrative law, in my opinion, is that judges still insist that they do not review the merits of a decision 407 . Review of a decision on a Wednesbury, variable  398  Supra n29. 399  Supra n293. 400  Supra n372. 401  Supra n263. 402  Southam, supra n172. 403  Toronto (City) Board of Education, supra n103. 404  Suresh, supra n186. 405  Cambridge District Health Authority, supra n298. 406  Prolife Alliance, supra n331 at paragraph 75. 407  Daly, supra n12.  140  unreasonableness, or proportionality ground all involve a consideration of the merits of the decision, with more or less deference given to the decision-maker. As we have seen, some commentators have attempted to argue that a high degree of deference somehow transforms review of the merits of a decision into review on the basis of an error of law, and this is a contention with which I cannot agree. I will expand on this argument in the remainder of this thesis.  I will now examine whether this convergence between Canadian and UK law has affected two other common law jurisdictions, these being New Zealand and Australia. I will demonstrate that New Zealand, perhaps somewhat reluctantly, is converging to the same kind of approach as Canada and the UK, while Australia is still dragging its feet.   141  CHAPTER FIVE DEFERENCE IN NEW ZEALAND  The development of judicial review in New Zealand has proceeded along lines very similar to that of the UK, at least in the lower courts. In particular, both countries have adopted a form of “variegated review” or sliding scale of reasonableness, at least at the level of the lower courts, and both are at least moving towards a proportionality approach to judicial review, at least where human rights are in issue. Unlike Canada, both countries are unitary, as opposed to federal, states. Like the UK, New Zealand has a statutory bill of rights, and courts have generally scrutinised decisions impacting on those protected rights more thoroughly than in other cases. Both countries also have a fairly new final appellate court. The Supreme Court of New Zealand came into existence on 1 January 2004 with the passage of the Supreme Court Act 2003, which abolished appeals to the British Privy Council. In the UK the Supreme Court of the United Kingdom was established on 1 October 2009 by means of Part III of the Constitutional Reform Act 2005, subsuming the functions of the Judicial Committee of the House of Lords.  Michael Taggart is just one commentator who has noted the similarity between the UK and New Zealand in the development of substantive review 408 :  Ten years ago, when writing about the scope of judicial review on questions of law, I noted that the doctrine of deference was well established in Canada and the United States, but not explicitly recognized at all in the administrative laws of New Zealand, Australia, and the United Kingdom. Now the word “deference” is very much in the air in the United Kingdom and New Zealand, although the Australians still consider it to be blasphemy. A broadly similar pattern can be seen with the doctrine, or rather the methodology, of “proportionality”. Ten years ago,  408  Michael Taggart, “Proportionality, Deference, Wednesbury”, [2008] New Zealand Law Review 423 at 424-5  142  it was seen as a foreign will-o’-the-wisp, emanating from continental European legal systems and the European Union, and not easily transplantable to common law soil. Nowadays, the “Eurospeak” of proportionality abounds in the United Kingdom and is increasingly heard in New Zealand public law.   One difference between the UK and New Zealand is the number of administrative law cases that have reached the highest courts. New Zealand has a population of around 4.5 million people compared to the UK’s 62.3 million and Canada’s 34.6 million409, meaning there are simply fewer cases to be heard. Glazebrook J of the New Zealand Court of Appeal, writing extradjudicially, has commented as follows 410 :  As a small democracy, New Zealand faces particular problems in developing a coherent theory of judicial review ... These difficulties were outlined by Hammond J in his judgment in Lab Tests Auckland v Auckland District Health Board 411 . Because there is a limited number of cases that come before the courts, there is less opportunity to develop the law. Indeed, as Hammond J remarked: “There is … an intermittent and somewhat mad-headed chase after the ‘latest case’ on the part of the bar and commentators, and seminars sprout up as if there has been a seismic shift when one case is decided”412. This has the effect that each individual case has a disproportionate impact on the law in this country. As a result, it seems that progress occurs in occasional leaps and bounds, rather than at the more sedate pace that is usually acceptable for those wearing judicial robes.   The smaller number of cases that come before the superior courts in New Zealand is one reason why the law of substantive review in New Zealand is somewhat unclear. As we will see, the lower courts, have generally adopted a variable scale of reasonableness review, and appear to be moving towards a “proportionality” approach (two approaches which I have already argued  409 410  Justice Susan Glazebrook, “To the Lighthouse: Judicial Review and Immigration in New Zealand”, Paper written for the Supreme Court and Federal Court Judges’ Conference, Hobart, 24 to 28 January 2009 at 3. 411  [2008] NZCA 385. In the same case, Hammond described judicial review in New Zealand as “stand[ing] somewhat in a fog of mushy dogma” (paragraph 373). 412  Ibid at 396.  143  exhibit little if any difference to each other). However, the Supreme Court is yet to issue an unambiguous pronouncement on the matter.  In this chapter, I will examine New Zealand jurisprudence on substantive review since the coming into effect of the Bill of Rights Act 1990 (BORA). In particular, I will examine the adoption of the concepts of deference and proportionality in that country.  5.1 Deference and variable unreasonableness in New Zealand law – the beginnings  5.1.1 Wellington City Council v Woolworths New Zealand Ltd (no 2)  Unlike the UK, which saw a shift to proportionality analysis in Human Rights Act cases only three years after that act came into effect, the concepts of deference and proportionality seem to have been slower to take off in New Zealand. Instead, New Zealand courts adopted the Wednesbury test, which was most recently restated in Wellington City Council v Woolworths New Zealand Ltd (No 2) 413 . Dean Knight states that “in rejecting a challenge to the merits of a rates-setting decision of a local authority, Richardson P endorsed the highly deferential form of Wednesbury reasonableness”414, an acknowledgement that even Wednesbury review is a form of deferential substantive review. In another article, Knight describes the Woolworths judgement as follows 415 :  413  [1996] 2 NZLR 537. 414  Dean Knight, “A Murky Methodology: Standards of Review in Administrative Law”, (2008) 6 New Zealand Journal of Public and International Law 180 at 182-3. 415  Dean Knight, “Dunne v Canwest TVWorks Ltd: Enhancing or Undermining the Democratic and Constitutional Balance?”, (2005) 21 New Zealand Universities Law Review 712 at 719-20.  144   The leading statement of the standard of unreasonableness in Wellington City Council v Woolworths New Zealand Ltd (No 2) is well known, and characterises the stringent threshold in a number of ways: “something overwhelming”, “outrageous in its defiance of logic or accepted moral standards”, “absurd”, “a pattern of perversity”, and “outside the limits of reason”. Likewise, “arbitrary” has a similarly high threshold. Lord Cooke characterised the ground of arbitrariness as having the “connotation of the despotic” and amounting to “caprice rather than a reasoned preference and balancing”.  However, Woolworths was a last gasp of the Wednesbury principle – a sort of attempted counter- reformation against a deference principle that was already gaining ground, and which ultimately failed. The modern situation in New Zealand substantive review is again summed up by Michael Taggart 416 :  As regards exercise of discretion, until very recently that was the province of Wednesbury unreasonableness. The courts now openly accept that (un)reasonableness is not a monolithic concept, and in truth never was. There is a sliding scale or rainbow of possibilities, with intensity of review varying depending on context and other factors. The dividing lines between law, discretion, and fact are notoriously grey and porous.  5.1.2 Origins of deference and variegated review in New Zealand  The first clear articulation of a doctrine of substantive review coupled with a concept of deference was seen in an earlier Court of Appeal decision, Thames Valley Electric Power Board v NZFP Pulp and Paper Ltd. Lord Cooke, the President of the Court of Appeal, stated as follows 417 :  The merit of the substantive unfairness ground is that it allows a measure of flexibility enabling redress for misuses of administrative authority which might  416  Michael Taggart, “Administrative Law”, [2006] New Zealand Law Review 75 at 83. 417  [1994] 2 NZLR 641 at 653.  145  otherwise go unchecked ... It seems safe to say that the limits or categories of substantive fairness will never be defined with exhaustive precision … At times it becomes necessary to give especial weight to human and civil rights. At times the emphasis will be more on non-interference with the legitimate exercise of governmental or other administrative discretion. It is equally safe to say, however, that never has it been suggested that the mere personal opinion of a Judge that a decision is unfair will justify holding it invalid. Nor is that ever likely to be suggested. The functions of exercising administrative discretion and judicially reviewing its exercise are fundamentally different. The line is not always easy to draw, but it has to be drawn.  This judgement is fairly similar to the British “variegated review” that began with Budgaycay418 and continues to the present day in non-Human Rights Act cases. It does, however, seem to focus on “misuse” of administrative authority leading to breaches of human rights, which reads something like a mala fides argument rather than simple unreasonableness.  Lord Cooke failed to carry the day in the Woolworths 419  decision, but the Court of Appeal’s rebuff to the variegated review principle did not last long. The concept of deference was “floated” again just one year later in Fulcher v Parole Board420, in which Thomas J declared that deference to an administrative decision-maker could be necessary in some circumstances. However, his Honour declined to defer to the decision of the Parole Board in the case at bar, stating that a reviewing court had to be careful “not to abdicate its ultimate responsibility for the decision”421. Leaving aside the fact that a reviewing court, more or less by definition, does not have “ultimate responsibility” for an administrative decision, this case is at least shows the beginnings of the acceptance of a deference concept.   418  Supra n282. 419  Woolworths (No 2), supra n413. 420  (1997) 15 CRNZ 222 (CA). 421  Ibid at 245.  146  Thomas J repeated his attack that same year in the better known decision of Waitakere City Council v Lovelock 422 . Dean Knight describes the judgement as follows 423 :  The particular target of his criticism was the subjective element inherent in Wednesbury reasonableness, along with its semantic complexity. Although he proposed a uniform standard divorced from the exaggerated Wednesbury language, he envisaged that degrees of tolerance would vary according to the circumstances. He was also critical of the lack of transparency in judicial supervision and suggested the courts should develop a series of principles which inform the reasonableness analysis  Again, this kind of reasoning appears very similar to the British “variegated review”, based on a single standard of “reasonableness” rather than “proportionality”. The Court of Appeal again hinted at the possibility of variable standards of reasonableness in Pharmaceutical Management Agency Ltd v Roussel Uclaf Australia Pty Ltd, where the court noted that in some cases, especially those involving human rights, “a less restricted approach, even perhaps, to use the expression commonly adopted in the United States, a ‘hard look’424, may be needed”425. However, this comment was strictly obiter.  5.2 A sudden rush – the Court of Appeal in 1999-2000  Significant support for the concept of a variable standard of reasonableness in judicial review came in four cases in the space of twelve months. In Ports of Auckland Ltd v Auckland City Council 426 , Baragwanath J of the High Court first attempted to move New Zealand substantive  422  [1997] 2 NZLR 385 at 400-1. 423  Knight, “A Murky Methodology”, supra n414 at 186. 424  This is a reference to the decision of the US Supreme Court in Motor Vehicle Manufacturers Association v State Farm Mutual Automotive Insurance Co (1983) 463 US 29. 425  [1998] NZAR 58 at 66. 426  [1999] 1 NZLR 601.  147  review towards a “varying intensity” model of the UK variety. Ports of Auckland concerned a development application granted by the council, which the company claimed would cause interference with operation of the port, and was therefore “unreasonable”. Baragwanath J reviewed the Wednesbury unreasonableness approach and then stated as follows 427 :  While Judges of this Court do not in general claim the specialist qualifications and experience of the Environment Judges appointed under s.250 of the [Resource Management Act 1991], who have the benefit of sitting with Environment Commissioners contributing the qualifications described in s.253, the business of construing documents and of assessing the prospects of success in injunction proceedings is very much the business of the High Court. The present case is towards the opposite end of the spectrum considered by the President in Wellington City Council v Woolworths. I prefer therefore to employ the lower- level test applied in Re Erebus Royal Commission [1983] NZLR 662 (PC) at p 681, namely whether the decision is “based upon an evident logical fallacy”.  Re Erebus Royal Commission; Air New Zealand v Mahon 428  concerned the Royal Commission into the Mount Erebus disaster, in which an Air New Zealand sightseeing plane crashed into Mount Erebus in Antarctica, killing all on board. The Commission made a number of damning findings against Air New Zealand and related parties, and they sought judicial review of the Royal Commissioner’s findings, arguing amongst other things that the Commissioner’s fact finding process was erroneous. The matter proceeded to the Privy Council, where Lord Diplock commented on review of findings of fact as follows 429 :  As Courts whose functions in the instant case have been restricted to those of judicial review, both the Court of Appeal and this Board are disentitled to disturb findings of fact by the decision-maker whose decision is the subject of review, unless (1) the procedure by which such findings were reached was unlawful (in casu by failure to observe the rule of audi alteram partem) or (2) primary facts were found that were not supported by any probative evidence or (3) the  427  Ibid at 606. 428  [1983] NZLR 662 (PC). 429  Ibid at 681.  148  reasoning by which the decision-maker justified inferences of fact that he had drawn is self-contradictory or otherwise based upon an evident logical fallacy.  In other words, Baragwanath J in Ports of Auckland seems to have taken Lord Diplock out of context. The Erebus Royal Commission case was concerned with the ability of courts to disturb the findings of fact of decision-makers, but Baragwananth J has expanded the “evident logical fallacy” test to cover the substance of a decision of an administrator. It is quite a leap.  There were in fact three other very important decisions on judicial review handed down in New Zealand in the space of twelve months around this time. Besides Ports of Auckland, the Court of Appeal found in Peters v Davison 430  that there is no longer any distinction between jurisdictional and non-jurisdictional errors in New Zealand law, following the earlier Court of Appeal decision in Bulk Gas Users Group Ltd v Attorney-General 431 , and of course Anisminic 432 . Peters v Davison also identified the rule of law as a substantive issue in New Zealand law, which led Philip Joseph to argue that this decision altered the rationale for judicial review in New Zealand from a simple ultra vires analysis to a “rule of law” basis433.  The third case, Dunlea v Attorney-General 434, involved a claim that a police “raid” on a property was unreasonable and violated the applicant’s rights under s.21 of the BORA, which provides that “everyone has the right to be secure against unreasonable search or seizure, whether of the person, property, or correspondence or otherwise”. While none of Wednesbury, Woolworths or  430  [1999] 2 NZLR 164. 431  [1983] NZLR 129. 432  Supra n85. 433  Joseph, supra n270 at 359-60. 434  [2000] 3 NZLR 136.  149  Ports of Auckland is cited in the judgement, it seems clear that Thomas J, who wrote a separate concurring judgement, was of the view that “reasonableness” has different meanings depending on the rights in issue in the case. Bradley Selway QC, after stating that “the new English approach clearly permits merits review subject only to whatever forbearance the judge, as a matter of policy, is prepared to give”435, then goes on to say that “the New Zealand courts have adopted a similar approach to that of the English courts”436, citing Peters v Davison and Dunlea v Attorney-General in support of this proposition. It seems that this statement was somewhat premature.  Finally, the Court of Appeal also had cause to consider the concept in Pring v Wanganui District Council 437 . As we have already seen in Canadian and UK decisions 438 , Blanchard J, writing for the court, noted that the “weight” to be given to particular factors in the decision-making process was a matter for the decision-maker, and was not to be reviewed by the courts on a “reasonableness” analysis439. However, his Honour added as follows immediately afterwards:  Having said that … the court will scrutinise what has occurred more carefully and with a less tolerant eye when considering whether the decision was one open to the consent authority on the material before it than it will do in a case where the decision which is being questioned required the balancing of broad policy considerations and there was less direct impact upon the lives of individual citizens.  In the end, the issue was somewhat moot, as the Court of Appeal found that the decision in question was reasonable in any sense of the word. However, the four 1999-2000 cases, taken  435  Selway, supra n360 at 224. 436  Ibid. 437  [1999] NZRMA 519. 438  Suresh, supra n186; Khosa, supra n29; Daly supra n12. 439  Supra n437 at paragraph 7.  150  together, seemed to have set up the concept of a variegated standard of reasonableness review for the future.  5.3 A revival of variegated reasonableness? Wolf v Minister for Immigration  Wolf v Minister for Immigration 440  concerned a German citizen who was alleged to have entered New Zealand in 1986 with forged documentation, having escaped prison in his own country. He married a New Zealand citizen and had two children together with his wife. The couple separated in 1995, but he continued to have access to the children. In 1997 his (now ex) wife informed the New Zealand Immigration Service (NZIS) about Mr Wolf’s past, which three years later resulted in a decision to cancel his entry permit. This decision was upheld by the Deportation Review Tribunal (DRT). Mr Wolf’s case proceeded to the High Court, where he argued that the DRT’s decision was both unreasonable and disproportionate, relying on s.9 of the BORA and the impact on his relationship with his children in particular 441 .  Wild J first addressed the issue of whether there is in fact any difference between reasonableness and proportionality. His Honour noted that “views differ as to exactly what proportionality involves and as to the extent, if any, to which it differs – or should continue to be differentiated – from Wednesbury unreasonableness”442, a question which seems to have a fairly obvious answer. There may be little if any difference between a variegated unreasonableness analysis and  440  [2004] NZAR 414. 441  The facts of the case therefore bore a resemblance to two well-known New Zealand cases concerning the impact of removal of a parent on their children, Tavita v Minister of Immigration [1994] 2 NZLR 257, and Puli’uvea v Removal Review Authority (1996) 2 HRNZ 510. 442  Supra n440 at paragraph 26.  151  proportionality, but there does appear to be a significant difference between a decision that is so unreasonable that no reasonable decision-maker could make it, and one that is disproportionate. Despite this unpromising beginning to his analysis, Wild J then goes on to discuss some of the early British proportionality decisions, such as R (Alconbury Developments Ltd) v Secretary of State for the Environment, Transport and the Regions 443  and R v Secretary of State for the Home Department, ex parte Daly 444, describing the latter as an “excellent example of proportionality at work”445. At paragraph 35 Wild J then states as follows:  [T]he role for proportionality in current New Zealand public law is unclear. Is it simply one criterion in assessing whether a decision is unreasonable, or is it truly a separate principle? As the law in New Zealand currently stands, I think it best to take the cautious approach of acknowledging that traditional (Wednesbury) grounds of review and proportionality are different, and may therefore produce different outcomes.  Again, this passage really only makes sense if Wild J is talking about a variable reasonableness analysis (as he later more clearly does), and if that is what his Honour had in mind, then his decision squares with the judgement of Lord Steyn in Daly 446 . Wild J then decided that proportionality had no role to play in the case at bar, stating as follows at paragraph 36:  I do not consider proportionality has any real application to Mr Wolf’s case. His appeal did not involve the DRT assessing whether the means to achieve some legislative aim were proportionate to that aim. I accept that there may have been an element of that in the Minister’s decision on 10 April 2000 to revoke the appellant’s residence permit. But … that decision is not the subject of this appeal. The issue for the DRT was simply whether it was “satisfied that it would be unjust or unduly harsh for the appellant to lose the right to be in New Zealand indefinitely”.   443  Supra n328. 444  Supra n12. 445  Supra n440 at paragraph 32. 446  Supra n12 at paragraph 28.  152  This brings us back to the “possible, acceptable range of outcomes” problem. In this case, only two outcomes were possible – the decision to deport Mr Wolf could be either upheld or set aside. Is this passage a statement that proportionality is not really an issue when the range of possible outcomes is so limited? In any event, having made that finding, Wild J then went on to find that Article 9 of the BORA, which stated that everyone has the right not to be subjected to “disproportionately severe treatment or punishment” was not engaged, and there was no breach of the BORA in the DRT’s decision447.  Wild J cited Puli’uvea v Removal Review Authority448 in support of his analysis. At paragraph 38 he reproduced the following passage from the judgement of Keith J 449 :  The action of removing Mrs Puli’uvea cannot be said to begin to obtain the high threshold required by the prohibition in the New Zealand Bill of Rights Act on disproportionately severe treatment. The cases here and elsewhere expand on such constitutional guarantees by using expressions such as “treatment that is so excessive as to outrage standards of decency”.  Wild J did, however, clearly endorse a variable scale of reasonableness review. At paragraph 47 his Honour noted as follows:  I consider the time has come to state – or really to clarify – that the tests as laid down in GCHQ and Woolworths respectively are not, or should no longer be, the invariable or universal tests of “unreasonableness” applied in New Zealand public law. Whether a reviewing Court considers a decision reasonable and therefore lawful, or unreasonable and therefore unlawful and invalid, depends on the nature of the decision: upon who made it; by what process; what the decision involves (ie its subject matter and the level of policy content in it) and the importance of the decision to those affected by it, in terms of its potential impact upon, or consequences for, them.  447  Supra n440 at paragraphs 37-38. 448  Supra n441. 449  Ibid at 525.  153   This passage is very similar to the Supreme Court of Canada’s decision in Baker v Canada (Minister of Citizenship and Immigration) 450 , where the court found that the seriousness of the impact of the decision on the individual in question will affect both the level of procedural protection 451  and whether the decision could be regarded as “reasonable” on substantive review 452 . It is also redolent of Dunsmuir, in both the majority judgement 453 , and perhaps even more so in the judgement of Deschamps J 454 . We can now see the beginnings of a convergence in Canadian, UK and New Zealand law.  In a very lengthy paragraph 48 of the judgement, Wild J identifies four reasons for the shift to a variable standard. Firstly, the Wednesbury test would not survive the scrutiny of the BORA, an argument that is somewhat similar to Smith and Grady v United Kingdom 455 . Secondly, a variable scale had been recognised for at least 20 years, most prominently in Thames Valley Electric Power Board v NZFP Pulp and Paper Ltd 456 . Woolworths should be confined to its facts, ie that a high level of deference should be given in relation to development and ratings decisions 457 . Thirdly, New Zealand academic commentators supported the concept, and finally  450  Supra n60. 451  Ibid at paragraph 25. 452  Ibid at paragraphs 60 and 67. 453  Supra n7 at paragraph 64. 454  Ibid at paragraph 158. 455  Supra n349. 456  Supra n417. 457  This suggestion was followed in Discount Brands Ltd v Northcote Mainstreet Inc [2004] 3 NZLR 619.  154  the same concept had been endorsed in a number of foreign jurisdictions, including the UK 458  and Canada 459 .  Wild J found that the DRT’s decision passed the traditional Wednesbury test, but also found that “a lesser test is appropriate here, because the decision involves the deportation of the appellant, and the consequent breakup of a New Zealand family unit”460. The decision was therefore set aside on both substantive and procedural grounds.  5.4 Increasing intensity of anxiety?  As previously noted, there has been a lack of authoritative guidance from the Supreme Court of New Zealand on either the “variable standard of review” or proportionality issues since it commenced operation on 1 July 2004. There have been two cases since that date in which the Supreme Court could have considered the issue, but the majority at least did not find it necessary to do so.  5.4.1 Discount Brands  Discount Brands Ltd v Westfield (NZ) Ltd 461  was another development case, and again the issue was the reasonableness or otherwise of the development application decision. The key provisions  458  Daly, supra n12; Mahmood, supra n324. 459  Baker, supra n60, and Suresh, supra n186. Wild J regarded both of these cases has having some similarity to the case at bar. 460  Supra n440 at paragraph 65. 461  [2005] 2 NZLR 597.  155  of the Resource Management Act 1991 that were in issue in this case were ss.93 and 94(2), and which relevantly provided as follows.  93. Notification of applications  (1)  Once a consent authority is satisfied that it has received adequate information, it shall ensure that notice of every application for a resource consent made to it in accordance with this Act is – (a)  served on every person (other than the applicant) who is known by the authority to be an owner or occupier of any land to which the application relates; and …. (e)  served on such persons who are, in its opinion, likely to be directly affected by the application, including adjacent owners and occupiers of land, where appropriate  94(2) An application for a resource consent need not be notified in accordance with section 93, if the application relates to a discretionary activity or a non-complying activity and – (a)  the consent authority is satisfied that the adverse effect on the environment of the activity for which consent is sought will be minor; and (b)  written approval has been obtained from every person whom the consent authority is satisfied may be adversely affected by the granting of the resource consent unless the authority considers it is unreasonable in the circumstances to require the obtaining of every such approval.  The reference to “adequate information” in s.93(1) relates to information received by the “consent authority” (the council) in relation to the “resource consent” (basically an application for development).  Discount Brands had lodged an application with the North Shore council, and sought to avoid the disclosure requirements by means of s.94. The council had notified Discount Brands’ immediate neighbours of the proposal, but did not notify Northcote, which was a community body set up to  156  enhance and preserve the character of the North Shore shopping district. It also did not notify Westfield, which owned two large shopping malls in the general area. Both parties sought review of the council’s decision, arguing that it had not met the requirements of s.94.  The Supreme Court upheld the appeal, but the majority did so on the basis that the council had misinterpreted s.94(2) of the Resource Management Act 1991. Using language that might horrify some Canadian academics, Elias CJ stated at paragraph 5 that “I am of the view that the council failed to address the right question under s 94(2)(a)”, and her Honour also noted that “I do not consider that answers are helpfully advanced by consideration of the scope and intensity of the High Court’s supervisory jurisdiction to ensure reasonableness in substantive result in the exercise of statutory powers”. In other words, the council had simply misinterpreted its own powers, and any “standard of review” was irrelevant. Michael Taggart argues that this means that New Zealand courts will apply a correctness standard of review to questions of law 462 .  Blanchard J upheld the appeal on the basis that it was not sufficient that the decision-maker had “some probative evidence”463, as held by the Court of Appeal. His Honour stated as follows464:  The Court of Appeal in this case erred in thinking that the references in Pring 465  to making a decision “upon the basis of the material available” to the decision maker, to the need for “some material capable of supporting the decision” and to considering “whether the decision was one open to the consent authority on the material before it” provided guidance in the present case. They do not, because, as has just been pointed out, they are related to a subsequent stage in the process, after the requisite information has been obtained.   462  Taggart, “Administrative Law”, supra n416 at 77. 463  Supra n461 at paragraphs 63-4. 464  Ibid at paragraph 118. 465  Supra n437.  157  In other words, Blanchard J found that the Court of Appeal had conflated two distinct steps in the decision-making process. The council had to consider both ss.93 and 94 of the Act, and the former required not just “some probative evidence”, but adequate evidence, before moving on to consideration of s.94. That is, the council had to be satisfied that it had “adequate information” about the development proposal, in accordance with s.93, before it could even consider s.94. In Canadian terms, this could read like a “preliminary question” line of reasoning, which also resembles the Metropolitan Life 466  kind of approach to substantive review that has been rejected by the Canadian courts since. Michael Taggart, though, sees this approach as more akin to a “jurisdictional fact” approach467:  Blanchard J saw the Court of Appeal’s error in muddling two distinct steps in the decision-making process. In accordance with s 93, the information received had to be adequate. The Court of Appeal had misunderstood and therefore misapplied a dictum from Pring 468, wrongly holding that “some probative evidence” would satisfy the adequacy requirement at the first step under s 93 (para 118). It was not a matter of simply having some evidence of probative value (as the Court of Appeal held), there actually had to be adequate information. Several of the Judges described this in terms of a condition precedent before s 94(2) could operate … This language is reminiscent of the old doctrine of jurisdictional fact, that some fact or state of affairs had to be correctly ascertained in order to validly proceed to the second step of satisfaction under s 94(2).  The concept of the jurisdictional fact will be discussed more fully in the next chapter of this thesis, as the Australian High Court has recently made it clear that the “jurisdictional fact” doctrine is well and truly alive in that country 469 .   466  Supra n16. 467  Taggart, “Administrative Law”, supra n416 at 77. 468  Supra n437. 469  M70/2011 and M106/2011 v Minister for Immigration and Citizenship [2011] HCA 32.  158  The only member of the court who clearly addressed the deference and standard of review issue was Keith J, who regards what he saw as procedural failings on the part of the council as a matter going to substantive review. While Canadian courts have at times blurred the distinction between process and substance themselves – Baker v Canada (Minister of Immigration and Citizenship) 470  being the classic example, particular in so far that deference can be provided to a tribunal’s choice of procedures – they have not generally applied any “standard of review” analysis to procedural fairness issues. If anything, recent cases, most particularly Alberta (Information and Privacy Commissioner) v Alberta Teachers’ Association471, have attempted to reassert the distinction between procedure and substance. The approach of Keith J also seems to conflict with the House of Lords’ approach in R (on the application of Begum) v Headteacher and Governors of Denbigh High School 472 , where the court found that matters of procedure were not to be examined on a reasonableness or proportionality basis. This may be an area where New Zealand law remains different to its near legal neighbours in Canada and the UK.  5.4.2 Unison Networks Ltd v Commerce Commission  The Supreme Court did appear to apply a deferential form of review, without using that terminology, in Unison Networks Ltd v Commerce Commission 473 . Broadly speaking, the Commission had the power to investigate prices charged by electricity companies that were in excess of published thresholds. When the Commission investigated Unison for charging prices in  470  Supra n60. 471  Supra n179. 472  Supra n372. 473  [2008] 1 NZLR 41.  159  excess of the threshold, Unison sought judicial review of the thresholds themselves, arguing that they were set in a manner inconsistent with the Commission’s enabling legislation.  In Canada, this would appear to be a textbook “deference to expertise” case, and so it proved here. McGrath J, writing for the court, set out the basic legal principles 474 :  Often, as in this case, a public body, with expertise in the subject-matter, is given a broadly expressed power that is designed to achieve economic objectives which are themselves expansively expressed. In such instances Parliament generally contemplates that wide policy considerations will be taken into account in the exercise of the expert body’s powers. The courts in those circumstances are unlikely to intervene unless the body exercising the power has acted in bad faith, has materially misapplied the law, or has exercised the power in a way which cannot rationally be regarded as coming within the statutory purpose.  At paragraph 77 McGrath J stated that the Commission could have exercised its power in one of two ways (two “possible, acceptable outcomes”?), and it had chosen one of those approaches. There was therefore no ground for setting the decision aside.  This case is interesting as an example of the ability of the Supreme Court to apply a principle of deference without using that terminology. If McGrath J had stated that “deference may also be warranted where an administrative tribunal has developed particular expertise in the application of a general common law or civil law rule in relation to a specific statutory context”475, the same result would have been reached.    474  Ibid at paragraph 55. 475  Toronto (City) v CUPE, supra n102 at paragraph 72.  160  5.4.3 Ye v Minister for Immigration  In Ye v Minister for Immigration 476  the Supreme Court was concerned with the removal from New Zealand of two Chinese “overstayers”, who now had three New Zealand citizen children. The couple resisted their removal on a number of grounds, most notably that their children would be regarded as “black children”, born in contravention of the Chinese one child policy, in China. In issue was s.47(3)  of the Immigration Act 1987, which provided as follows:  An appeal may be brought only on the grounds that there are exceptional circumstances of a humanitarian nature that would make it unjust or unduly harsh for the person to be removed from New Zealand, and that it would not in all the circumstances be contrary to the public interest to allow the person to remain in New Zealand.   Elias CJ pointed out that the term “appeal” in s.47(3) is somewhat misleading477, given that the section referred to an “appeal” to the Removal Review Authority (RRA), which is an administrative tribunal and not a court. However, nothing seems to turn on this.  The issue of the standard or intensity of review was squarely raised in argument before the Supreme Court, and was received by the judges with varying degrees of enthusiasm 478 .  Mr Bassett: If I now can move just to the question of intensity of review, my submission is that this case could have been, and still can be, decided on the application of orthodox judicial review principles, however it is my submission that a hard look approach or a heightened level of intensity of review is nevertheless appropriate in a case such as this ...   476  [2010] 1 NZLR 104. 477  Ibid at paragraph 2. 478  Ye v Minister of Immigration (NZSC, transcript, 21-23 April 2009, SC 53/2008) at 179-182.  161  Tipping J:  I would hope I would always have a hard look, the question is more, isn’t it, to the standard to which you hold the decision making? …  Elias CJ:  I don’t know that degrees of reasonableness help either.  Tipping J:  No.  Elias CJ:  It’s just, it’s got to be contextual. What is reasonable takes its colour from the context. Really, there’s so much dancing around on the heads of pins in this area.  Mr Bassett:  I think the lingo, if you like, Your Honour, is a question of deference, less deference where the rights are more fundamental and more deference where it’s --  Elias CJ:  That’s a dreadful word.  Tipping J:  It’s a controversial word. I understand the concept, you’re more exacting, if you like, the more fundamental -- it’s a more exacting test, or --  Anderson J:  It connotes the extent to which a Court’s prepared to interfere … however you describe it.  Tipping J:  And the Court must interfere where it must. You either feel driven to interfere or you don’t, and that will depend on what sort of a right it is and what the whole shebang is, I think this is a --  Anderson J:  It’s really intensity of anxiety.  It is particularly interesting that Tipping J starts to talk about “the standard to which you hold the decision-making”, only to be seemingly slapped down by Elias CJ. Dean Knight has described the view of Elias CJ on issues of varying standards of review 479 :  The Chief Justice is one of the most vocal critics of variable intensity. As can be seen from the exchange in Ye, she thinks that deference is a “dreadful” word and the concept of degrees of unreasonableness amounts to “dancing around on the heads of pins”. Elsewhere, she has complained that the idea of a spectrum of  479  Dean Knight, “Mapping the Rainbow of Review: Recognising Variable Intensity”, [2010] New Zealand Law Review 393 at 402.  162  unreasonableness was “a New Zealand perversion of recent years”480. She has similarly shunned attempts to develop notions of curial deference. In Discount Brands Ltd v Westfield (New Zealand) Ltd 481 , she rejected attempts to argue the propriety of the non-notification of resource consents in terms of intensity. Her Honour said that the questions turned on “the legislation and its application in context” and were not “helpfully advanced by consideration of the scope and intensity of the High Court’s supervisory jurisdiction to ensure reasonableness in substantive result in the exercise of statutory powers”482.  It is also interesting, however, to note that Elias CJ states that “what is reasonable takes its colour from the context”. This is a statement that, despite her rejection of the “spectrum of reasonableness”, could almost be taken straight from the judgements of Binnie J in Dunsmuir483 or Alberta Teachers 484 . In other words, she seems to recognise that a varying standard of reasonableness is necessary, and her real objection might be to over-complicating the issue by providing for specific standards of review or specific levels of deference. Elias CJ would no doubt have been horrified by the Canadian courts’ attempts to differentiate between “patent unreasonableness” and “reasonableness simpliciter”, and perhaps Dunsmuir might have allayed her concerns to some degree.  Given this background, it is hardly surprising that the Supreme Court in Ye also avoided any express discussion of deference or varying standards of review. Tipping J, writing for the majority, seemingly decided that Ye was not the best vehicle to argue for a formal system of variable review, noting that “the best approach in this Court, both for the parties and for the law generally, is to concentrate on the essential points which must be addressed in order to resolve  480  Astrazeneca Ltd v Commerce Commission (NZSC, transcript, 8 July 2009, SC 91/2008) at 52. 481  Supra n461. 482  Ibid at paragraph 5. 483  Supra n7 at paragraphs 150-152. 484  Supra n179 at paragraphs 67 and 68.  163  these particular cases” and that “the crucial issues turn on relatively straightforward principles of judicial review”485. Instead, Tipping J simply found that the RRA had misinterpreted s.47(3) of the Immigration Act 1987, and in particular that “we do not consider that … a general concern about the integrity of New Zealand’s borders and its immigration system will be enough in itself to demonstrate that it would be contrary to the public interest to allow a person fulfilling the first criterion to remain in New Zealand”486. The RRA had failed to examine the facts of the individuals’ case, and therefore it had failed to ask itself the right question487.  It would appear, then, that the New Zealand Supreme Court has yet to find an appropriate vehicle to squarely consider the issue of whether variable standards of reasonableness review exist in New Zealand law. Despite the reservations of Elias CJ about any concept of deference, the Supreme Court has never said that Wolf 488  was wrongly decided, but nor has it been endorsed. In short, it must be said that the state of New Zealand law on whether there is such a thing as a variable standard of reasonableness review is rather confused, and is still awaiting definitive Supreme Court authority on the topic. Given the current composition of the Supreme Court bench, it would appear that if the issue was to be definitively addressed at present, the court would come down against any “variegated standard” concept.   485  Supra n476 at paragraph 11. 486  Ibid at paragraph 37. 487  Ibid at paragraph 45. 488  Supra n440.  164  5.5 Conclusions  New Zealand also seems to be moving in the same direction as the UK and Canada, certainly at the lower court level. The Supreme Court has yet to show any real enthusiasm for the deference principle, and in fact Elias CJ of the Supreme Court has described it as a “dreadful word”489. Baragwanath J, of the High Court (which is the New Zealand equivalent of the Federal Court) and then the Court of Appeal, has been a particular champion of the variable reasonableness / proportionality approach to judicial review, most notably in Progressive Enterprises v North Shore City Council 490  (High Court) and X v Refugee Status Appeal Authority 491  (Court of Appeal). In the latter case, Baragwanath J cited Dunsmuir as authority for the ability of a court to decide some matters on a reasonableness basis and some on a correctness standard, before finding that decisions of the RSAA should be subjected to a correctness standard of review, because of the importance of the decision to the applicant 492 . Considering these lower court cases, Dean Knight has written as follows 493 :  In the High Court, judges have embraced the concept of variegated forms of unreasonableness, with little dissent. The high threshold of Wednesbury has been explicitly loosened, although the definition of the new standard of reasonableness has not been firmly settled. In the High Court, concepts like “hard look”494, “anxious scrutiny”495, or simple or intermediate unreasonableness496 are now routinely deployed, particularly in immigration cases and those dealing with technical environmental planning cases.   489  Supra n478 at 180. 490  [2006] NZRMA 72. 491  [2009] NZCA 488. 492  Ibid at paragraph 271. 493  Knight, “Rainbow of Review”, supra n479 at 408. 494  Roussel Uclaf, supra n425 at 66. 495  Bugdaycay, supra n282 at 531. 496  Wolf, supra n440 at paragraph 48.  165  It does seem unlikely, however, that the Supreme Court will formally adopt this kind of reasoning while Elias CJ remains on the bench 497. Even so, Elias CJ has admitted that “what is reasonable takes its colour from the context”498, which is an implicit recognition of some kind of variable scale of review, of much the same kind as the admission of Bastarache and LeBel JJ in Dunsmuir 499 . I would predict that the sheer weight of decisions pushing up from below will eventually force the Supreme Court of New Zealand to formally adopt at least a sliding scale of reasonableness, if not a form of proportionality review (at least for decisions impacting on rights protected by the BORA, as Canada and the UK have done), in the near future, or perhaps when Elias CJ retires! Adoption of such an approach would make Australia even more of an outlier in administrative law terms than it is now, and it is to the situation in Australia that we will now turn.     497  See also on this point Knight, “Rainbow of Review”, supra n479 at 402. 498  Supra n478 at 180. 499  Supra n7 at paragraphs 64 and 74.  166  CHAPTER SIX – DEFERENCE IN AUSTRALIA  As previously noted, Australia is somewhat of an “outlier” in matters of judicial review of administrative decisions in contemporary common law jurisdictions. Michael Taggart has described the Australian approach to judicial review as follows 500 :  There is a sharp distinction between questions of law – meaning the correct interpretation of statutory text and common law rules – and exercise of discretionary power. As regards the former … the Australian courts insist on having the last word on “correctness” (there is no deference: Marbury v Madison and all that).  As regards discretion, the courts could not defer more, in theory at least. Within the four corners of the power the decision-maker is free to decide as he or she likes. Once the decision-maker has applied the right legal test, the application of that test and the weight given to the relevant factors are a matter solely for the decision-maker … The court would not second-guess (or judge) under the guise of judicial review questions of fact, policy, weight or otherwise intrude into the merits.  I will address the Australian position as follows. First, I will consider the seminal case of Corporation of the City of Enfield v Development Assessment Commission 501  (“Enfield”), in which the High Court of Australia roundly rejected any application of the deference approach in Australia. I will then examine in more detail why this decision was reached, considering such issues as the concept of the “judicial power of the Commonwealth” in Australia, the historical approach to judicial review in Australia, and the seemingly vehement reaction by Australian commentators against any concept of deference. I will then address the question of whether the attempts by the Australian courts to sharply distinguish “law” and “merits” is really sustainable, and argue that a doctrine of substantive review similar to Canada’s is the best way forward.   500  Taggart, “Australian Exceptionalism”, supra n379 at 13. 501  Supra n2.  167  6.1 The Enfield decision  6.1.1 Prelude – Chevron  Before examining the Enfield decision, it is necessary to briefly examine the decision of the US Supreme Court in Chevron USA Inc v Natural Resources Defense Council Inc 502 . The reason for this is that when an attempt was made to create an express deference approach for Australian judicial review in 2000, this was the case referred to, rather than extant Canadian authority such as CUPE 503  or Pushpananthan 504 . Oddly, Canadian authority seems to have been hardly considered at all (and even then rejected) until the High Court’s decision in Minister for Immigration and Citizenship v SZMDS 505 . However, the rejection of Chevron can also be regarded as a rejection of the Canadian deference approach.  The facts in Chevron were fairly complex, but can be summarised as follows. Amendments made to the federal Clean Air Act in 1977 506  required certain states to establish a program whereby polluting industries were required to purchase permits for “new or modified major stationary sources” of air pollution. However, regulations issued by the Environmental Protection Authority (EPA) allowed a polluter to make certain modifications to its plant without applying for a permit, if the overall pollution level was not increased as a result. The issue before the court basically boiled down to whether a corporate group could be regarded a “stationary source” of air  502  Supra n13. 503  Supra n5. 504  Supra n170. 505  Supra n4. 506  Pub. L 95-95, 91 Stat 685.  168  pollution – Chevron had argued that although it had increased pollution at one site, it had reduced it by a comparable amount at another site in the same state, and therefore did not require a permit.  Stevens J, writing for the court, commented that “[w]e have long recognized that considerable weight should be accorded to an executive department’s construction of a statutory scheme it is entrusted to administer, and the principle of deference to administrative interpretations”507, clearly demonstrating the use of the word “deference”. It is also notable that later US cases have even used the phrase “Chevron deference”508.  Stevens J then found that the EPA had been given a broad discretion in deciding how the 1977 amendments should be implemented 509 . The key passage of the judgement is as follows 510 :  First, always, is the question whether Congress has directly spoken to the precise question at issue. If the intent of Congress is clear, that is the end of the matter; for the court, as well as the agency, must give effect to the unambiguously expressed intent of Congress. If, however, the court determines Congress has not directly addressed the precise question at issue, the court does not simply impose its own construction on the statute, as would be necessary in the absence of an administrative interpretation. Rather, if the statute is silent or ambiguous with respect to the specific issue, the question for the court is whether the agency’s answer is based on a permissible construction of the statute.  Stevens J remarked further that “a court may not substitute its own construction of a statutory provision for a reasonable interpretation made by the administrator of an agency”511. That is, an  507  Supra n13 at 844. 508  US v Mead Corporation (2001) 533 US 218 at 226. 509  Supra n13 at 862. 510  Ibid at 842-3. 511  Ibid at 844.  169  administrative decision should not be disturbed if the enabling statute is “silent or ambiguous” with respect to a particular issue, and the administrative decision-maker’s finding is based on a “permissible” or “reasonable” construction of the statute. The resemblance to CUPE, decided just five years earlier, is marked. Stevens J ultimately found that the EPA’s interpretation of the Clean Air Act, permitting Chevron to take the action it did, was reasonable and upheld its decision 512 .  This case has attracted significant academic attention, and I do not intend to make a detailed analysis of the commentary. However, Robert Dolehide has commented as follows 513 :  Chevron thus established an extraordinarily deferential standard of review for agency statutory interpretations. As Professor Elizabeth Foote has observed, “[c]ourts rarely invoke unreasonableness as a ground for setting aside agency action”514. Indeed, Foote concluded that “it is hard to find a single case in which the Court deemed the agency action unreasonable at the second step of the [Chevron] test”515. Accordingly, Chevron created a judicial review doctrine that holds that the interpretation of ambiguous regulatory and administrative statutory provisions is primarily the province of the agencies and not the courts.  Chevron can be seen to have overturned the “hard look” approach endorsed by a case decided just one year earlier, Motor Vehicle Manufacturers Association v State Farm Mutual Automotive Insurance Co 516. Interestingly, though, the “hard look” terminology lives on in Australian (and at times New Zealand) jurisprudence.   512  Ibid at 866. 513  Robert Dolehide, “A Comparative Hard Look at Chevron: What the United Kingdom and Australia Reveal About American Administrative Law”, (2010) 88 Texas Law Review 1381 at 1386. 514  Rosemary Foote, “Statutory Interpretation or Public Administration: How Chevron Misconceives the Function of Agencies and Why It Matters”, (2007) 59 Administrative Law Review 673 at 708. 515  Ibid at 709. 516  Supra n424.  170  6.1.2 Enfield – the facts517  In Enfield, a waste management company applied to the Development Assessment Commission (DAC) in South Australia for approval to alter a waste treatment plant located within the local government area of Enfield Council. In the absence of approval by the DAC the development was prohibited. Before assessing the development against the relevant development plan, the DAC was required by subregulation 16(1) of the Development Regulations (SA) to “determine the nature of the development”. A development for “special industry”, defined in part as one causing fumes or producing conditions which may become offensive, would be a “non- complying” development for the purposes of the Development Act 1993 (SA). If the proposed waste plant was a non-complying development, s.35(3)(a) of that Act prohibited the DAC from granting provisional development plan consent unless the Minister and relevant council consented.  Enfield Council claimed that the application was properly classified as for a “non-complying” development, attracting the requirement specified under s.35(3)(a) of the Act that the consent of the Minister and the Council be obtained. However, the DAC decided the application was for general industry rather than special industry and therefore was not a non-complying development. On this basis the consent of the Enfield Council would not be required. The DAC granted the waste management company provisional consent.   517  Much of this section is paraphrased from Margaret Allars, “Chevron in Australia: A Duplicitous Rejection?”, (2002) 54 Administrative Law Review 569 at 581-2.  171  Enfield Council sought a declaration that the provisional consent was ultra vires and an injunction to restrain action being taken upon it. At first instance, the Supreme Court of South Australia granted the relief sought 518 . The primary judge (Debelle J) held that the development fell within the definition of “special industry” because the industry involved would produce conditions which would on occasion be offensive.  The Full Court of the Supreme Court of South Australia reversed the decision 519 , holding that it was inappropriate for the Court to admit evidence as to whether the development was properly classified as “special industry” and that the Court should defer “in grey areas of uncertainty to the practical judgment of the planning authority”520. The Court should only interfere if it appeared from the evidence before the DAC that that the DAC had made an obvious and clear departure from the requirements of the planning legislation 521 . The primary judge should not have “descend[ed] into the planning merits”522 since “without such an obvious and clear departure, the court on judicial review will defer to the judgment of the planning authority on planning issues”523.  Exactly why the Full Court used the word “defer” is not immediately clear. Chevron is not cited in the judgement, and nor is any other overseas authority. The High Court had not, at that time, made any pronouncement on the desirability or otherwise of the use of that word. It may be that term “deference” was seen to be a particular requirement of the Development Act 1993 (SA), as  518  Enfield City v Development Assessment Commission (1996) 91 LGERA 277. 519  Enfield City Corporation v Development Assessment Commission (1997) 69 SASR 99. 520  Ibid at 119. 521  Ibid at 115. 522  Ibid at 121. 523  Ibid.  172  another case dealing with s.35 of that Act, Corporation of the City of Kensington and Norwood v Development Assessment Commission, saw Williams J, with whom Cox and Prior JJ agreed, state that “[t]here will be cases where upon the application of s35(2) the court may now be entitled to defer to the judgment of the Planning Authority in a way which was not possible under the former legislation”524. In another planning case considering the same legislation, Mount Gambier Shopping Centres Pty Ltd v Village Fair Shopping Centres Pty Ltd and City of Mount Gambier, Matheson J stated that “[i]n grey areas of uncertainty, the court would be more likely to defer to the practical judgment of the responsible planning authority”525. The term “defer” does not appear to have arisen in other contexts in South Australia prior to Enfield.  It is also interesting here to note that the Full Court focused on one of the favourite approaches of Australian judicial review – to argue that there is a clear distinction between the law and the merits of a particular administrative decision. “Deference” here is equated with an avoidance of examining the “merits” of the case. However, despite the Full Court’s willingness to use a form of deference terminology in its decision, “deference”, as the term is used in Canada and the UK, is part of the substantive review of a decision, which does involve examination of the merits (even though the UK courts still insist otherwise). In Australian law as it stands, the merits of a decision are simply not to be considered, and “deference” is neither here nor there as a result.   524  (1998) 70 SASR 471 at 483. 525  [1996] SASC 5441 at paragraph 28.  173  6.1.3 High Court decision Majority judgement  The majority of the High Court was constituted by Gleeson CJ and Gummow, Kirby and Hayne JJ. After reviewing the legislation and the decisions of the lower courts, their Honours turned their attention to the concept of “deference”. Their Honours introduced the concept as follows526:  In the written submissions, reference was made to the applicability to a case such as the present of the doctrine of “deference” which has developed in the United States.  However, this Chevron doctrine, even on its own terms, is not addressed to the situation such as that which was before Debelle J. Chevron is concerned with competing interpretations of a statutory provision not, as here, jurisdictional fact-finding at the administrative and judicial levels.  The majority here drew a distinction between cases such as Chevron (and CUPE), where the issue was the interpretation of legislation by an administrative decision-maker with the delegated responsibility to make decisions under that legislation, and findings of “jurisdictional facts”. In the next paragraph, the majority elaborates on this, saying that “Chevron applies in the United States where the statute administered by a federal agency or regulatory authority is susceptible of several constructions, each of which may be seen to be (as it is put) a reasonable representation of Congressional intent”527. The High Court here seems to be saying that it will have the final say on any matter going to a tribunal’s jurisdiction, an approach rather redolent of Metropolitan Life 528 . What if the legislation conferring the decision-maker’s jurisdiction is itself capable of a  526  Supra n2 at paragraph 40. 527  Ibid at paragraph 41. 528  Supra n16.  174  number of reasonable interpretations? Will any credence be given to the decision-maker’s finding on its own jurisdiction, in interpreting legislation with which it is no doubt familiar?  The majority then turned to a theme that is, as we will see, common in Australian discussion of deference – the idea that giving any deference to an administrative decision-maker is somehow an abdication of judicial responsibility. At paragraph 41, the majority cited an extrajudicial article by Breyer J of the US Court of Appeal for the First Circuit, in which his Honour stated that the deference doctrine amounts to “a greater abdication of judicial responsibility to interpret the law than seems wise, from either a jurisprudential or an administrative perspective”529. At paragraph 42, their Honours cited Professor Keith Werhan as stating that 530 :  Before Chevron, the traditional approach viewed the interpretation of ambiguous laws to be a “question of law”; after Chevron, this task became simply a “policy choice”.531  Having transformed the legal into the political, the Justices ceded interpretative authority to the agencies.  Again, I would reply that if an agency makes a completely unreasonable interpretation of even its own enabling statute, no-one would argue that the courts should not intervene, to ensure that the agency acts within the power given to it by Parliament. Is it really “ceding” authority, however, to give a reasonable degree of “weight” to an agency’s interpretation of its “home” statute?  The High Court made the following observation at paragraph 42:   529  Justice Stephen Breyer, “Judicial Review of Questions of Law and Policy”, (1986) 38 Administrative Law Review 363 at 381. 530  Keith Werhan, “Delegalizing Administrative Law”, (1996) University of Illinois Law Review 423 at 457. 531  Chevron, supra n13 at 844-5.  175  An undesirable consequence of the Chevron doctrine may be its encouragement to decision-makers to adopt one of several competing reasonable interpretations of the statute in question, so as to fit the facts to the desired result.  In a situation such as the present, the undesirable consequence would be that the decision- maker might be tempted to mould the facts and to express findings about them so as to establish jurisdiction and thus to insulate that finding of jurisdiction from judicial examination.  One would have thought that administrative decision-makers have an incentive of this kind with or without any concept of deference. The situation could actually be made worse in a situation where an administrative decision-maker believes that a court will only accept one interpretation of a decision-making power. To give one example, in the Full Federal Court decision in Guo v Minister for Immigration and Ethnic Affairs, amongst many other things, found that a decision- maker should not come to an adverse view of an applicant’s credibility unless he or she was in a “positive state of disbelief”532 about his or her claims, as opposed to being merely doubtful about them. Between 26 February 1996, when this decision was handed down, and 13 June 1997, when the High Court overturned the Federal Court in a rare 7-0 judgement 533, the phrase “positive state of disbelief” occurred in no less than 315 reported decisions of the Refugee Review Tribunal (RRT), of a total of 5705 reported in that period. Since the High Court’s decision, the phrase has only occurred 192 times in published decisions, over a period of nearly 15 years 534 . This certainly looks like an attempt by members of the RRT to “mould the facts and to express findings about them” so “avoid judicial examination”.   532  (1996) 135ALR 421 at paragraph 20 of the judgement of Foster J. 533  Minister for Immigration and Ethnic Affairs v Guo (1997) 191 CLR 559. 534  These figures are taken from the RRT decision database at  176  The High Court then turned to the scope of judicial review in Australia, and in particular the prohibition on “merits review”535. The majority referred to ss.75(iii), 75(v) and 76(i)536 of the Australian Constitution as the sources of power for the High Court and other Federal Courts to review administrative decisions and the constitutionality of legislation and administrative action, and distinguished the prohibition on review of the merits from any kind of “deference” principle. Refusal to engage in merits review is not a form of deference, it simply represents the limit, as the High Court sees it, of judicial power in Australia 537 .  Oddly, the High Court went on to state that “in a proceeding in the original jurisdiction of a court on “appeal” from that tribunal, the “court should attach great weight to the opinion of the [tribunal]”538, and that “the weight to be given to the opinion of the tribunal in a particular case will depend upon the circumstances”539. The majority stated that these circumstances “will include such matters as the field in which the tribunal operates, the criteria for appointment of its members, the materials upon which it acts in exercising its functions and the extent to which its decisions are supported by disclosed processes of reasoning”540. These considerations appear to inform, at paragraph 46, a “deference to expertise” argument. The majority cited the following passage from Eclipse Sleep Products Inc v Registrar of Trade Marks 541  in support:   535  Enfield, supra n2 at paragraph 43. 536  Section 75 of the Australian Constitution provides for the original jurisdiction of the High Court, and is extracted in Chapter 1 of this thesis. Section 76 of the Constitution permits the Parliament to make laws granting further original jurisdiction to the High Court, in relation to “matters including arising under this Constitution, or involving its interpretation” (s.76(i)) or “arising under any laws made by the Parliament” (s.76(ii)). 537  The majority cites in support of this proposition Peko-Wallsend, supra n191, and Attorney-General (New South Wales) v Quin (1990) 170 CLR 1. 538  Enfield, supra n2 at paragraph 45, citing Registrar of Trade Marks v Muller (1980) 144 CLR 37 at 41. 539  Ibid at paragraph 47. 540  Citing Gummow J in Eshetu, supra n63 at 655. 541  (1957) 99 CLR 300 at 321-322.  177  By reason of his familiarity with trade usages in this country, a familiarity which stems not only from an examination of marks applied for and of the many trade journals which he sees, but from the perusal and consideration of trade declarations and the hearing of applications or oppositions, the Registrar is peculiarly well fitted to assess the standards by which the trade and public must be expected to estimate the uniqueness of particular indications of trade origin.  This line of reasoning would not look out of place in CUPE or Dunsmuir! Margaret Allars also sees the development of a principle of deference to expertise in the High Court in this and preceding cases, even though the judges themselves reject the use of such a term 542 . Professor Allars tracks the development of this approach from a number of trademark cases such as Eclipse Sleep Products Inc v Registrar of Trade Marks 543 , to the 1986 decision of Minister for Aboriginal Affairs v Peko-Wallsend Ltd 544 , to Eshetu 545 , which has resulted in an approach that is, as she sees it, similar to that adopted in Canada.  Ultimately, the majority set aside the decision of the Court of Appeal, thereby restoring the decision of Debelle J. Their Honours stated at paragraph 50 as follows:  However, it was the task of Debelle J to determine the question of the jurisdiction of the Commission upon the evidence as to “special industry” before him, as opposed to the probative material which had been before the Commission, and upon his construction of the relevant provision.  His Honour did so … If, at the end of the day, Debelle J had been in doubt upon a particular factual matter, it would have been open to his Honour to resolve that doubt by giving weight to any determination upon it by the Commission.     542  Allars, supra n517 at 585-7. 543  Supra n541. 544  Supra n191. 545  Supra n63.  178 Gaudron J  Gaudron J wrote a concurring judgement, focusing on the “jurisdictional fact” concept. In her Honour’s view, “once it is appreciated that it is the rule of law that requires the courts to grant whatever remedies are available and appropriate to ensure that those possessed of executive and administrative powers exercise them only in accordance with the laws which govern their exercise, it follows that there is very limited scope for the notion of ‘judicial deference’ with respect to findings by an administrative body of jurisdictional facts”546. That is, in the opinion of Gaudron J, if a statute requires certain facts to exist before an administrative body has jurisdiction, it is up to the courts alone to determine whether those facts exist and whether the decision-maker may lawfully embark on the inquiry – an approach that closely resembles the pre-CUPE “preliminary question” doctrine. However, Gaudron J also discusses the “weight” to be given to the opinion of the decision-maker 547 :  Where, as here, the legality of an executive or administrative decision or of action taken pursuant to a decision of that kind depends on the existence of a particular fact or factual situation, it is the function of a court, when its jurisdiction is invoked, to determine, for itself, whether the fact or the factual situation does or does not exist.  To do less is to abdicate judicial responsibility.  However, there may be situations where the evidence before the court is the same or substantially the same as that before the primary decision-maker and minds might reasonably differ as to the finding properly to be made on that evidence.  In that situation a court may, but need not, decline to make a different finding from that made by the primary decision-maker, particularly if the latter possesses expertise in the area concerned 548 . Even so, in that situation, the question is not so much one of “judicial deference” as whether different weight should be given to the evidence from that given by the primary decision-maker.  546  Supra n2 at paragraph 59. 547  Ibid at paragraph 60. 548  See R v Alley; Ex parte NSW Plumbers & Gasfitters Employees’ Union (1981) 153 CLR 376 at 390 per Mason J; R v Williams; Ex parte Australian Building Construction Employees’ and Builders Labourers’ Federation (1982) 153 CLR 402 at 411 per Gibbs CJ, Mason, Wilson, Brennan, Deane and Dawson JJ; R v Ludeke; Ex parte Queensland Electricity Commission (1985) 159 CLR 178 at 183-184.  179   Is there really any difference between “weight” and “deference” in this context? Surely according more or less weigh to the opinion of a decision-maker, particularly one with a particular expertise, is precisely the same thing as allowing a greater or lesser degree of deference? It looks more and more as if the Australian courts are simultaneously tying themselves up in semantic knots trying to avoid the using the word “deference”, and providing it in many cases anyway. Academic comment  Margaret Allars specifies three reasons why the High Court rejected the Chevron approach for Australia. Firstly, Chevron, even on its own terms, was not applicable to the situation. Chevron was concerned with competing interpretations of a statute, while Enfield was concerned with “the existence of a jurisdictional fact which was a precondition to the jurisdiction of the agency”549. Secondly, an application of Chevron may have the result that “an agency decision maker may be encouraged to adopt the competing interpretation which the facts will satisfy so as to produce the desired result, rather than determine the interpretation on the basis of proper principles of statutory interpretation”550. Finally, the Chevron approach would be antithetical to fundamental Australian principles of judicial review. Australia, following Marbury v Madison 551 , has taken the view that it is the role of court to declare and enforce the law: “This determines the  549  Allars, supra n517 at 584. 550  Ibid. 551  Supra n39.  180  limits of the function of courts in judicial review, requiring that they do not intrude upon the merits of administrative decisions.”552  Professor Allars’ first point will be addressed in a discussion of the jurisdictional fact doctrine later in this thesis. The second I have already addressed above, and my view is that the High Court is simply wrong on this point, in that the phenomenon they have identified exists regardless of any concept of deference. The final point is probably the most interesting one, and will require an examination of the judicial power of the Commonwealth, as set out in the Australian Constitution.  6.2 Merits review in Australian courts  6.2.1 The judicial power of the Commonwealth  A striking feature of Chapter III of the Australian Constitution is that the term “judicial power” is nowhere defined. It must therefore have been intended to be left to the High Court itself to determine what “judicial power” actually is. There is, however, a marked lack of authority on this point. Stephen Gageler SC, now the Solicitor-General of Australia and soon to be High Court judge 553, has commented that “[t]he largest and most emphatic words in the Constitution – take ‘judicial power’ and ‘absolutely free’ as well-worn examples – have no fixed or intrinsic  552  Allars, supra n517 at 585. 553  Stephen Gageler has been appointed to the position on the High Court to be vacated by Gummow J in October 2012. See “Gageler Appointed to High Court”, Sydney Morning Herald 21 August 2012.  181  meaning and it would be in vain to attempt to search for one”554. Tony Blackshield and George Williams QC have stated as follows 555 :  The characteristics and content of “judicial power” have not proved susceptible to precise definition. In Tasmanian Breweries, Windeyer J observed that “the concept seems … to defy, perhaps it were better to say, transcend, purely abstract conceptual analysis”556. In Chu Kheng Lim v Minister for Immigration, Local Government and Ethnic Affairs (1992) 176 CLR 1, McHugh J noted that “the line between judicial power and executive power in particular is very blurred”557, and that the classification of a power “frequently depends upon a value judgement … having regard to the circumstances which call for its exercise … The application of analytical tests and descriptions does not always determine the correct classification. Historical practice plays an important, and sometimes decisive, part”558. And in R v Quinn, ex parte Consolidated Foods Corporation (1977) 138 CLR 1, Aicken J concluded that “in substance, all that the courts have been able to say towards a definition has been the formulation of negative propositions by which it has been said that no one list of factors is itself conclusive and perhaps the presence of all is not conclusive”559.  Despite this unpromising background, Blackshield and Williams state 560  that the “classic” definition of judicial power is still that given by Griffith CJ in Huddart, Parker and Co Ltd v Moorehead, in which his Honour stated as follows 561 :  I am of opinion that the words “judicial power” as used in sec. 71 of the Constitution mean the power which every sovereign authority must of necessity have to decide controversies between its subjects, or between itself and its subjects, whether the rights relate to life, liberty or property. The exercise of this power does not begin until some tribunal which has power to give a binding and authoritative decision (whether subject to appeal or not) is called upon to take action.  554  Stephen Gageler SC, “Beyond the Text: A Vision for the Structure and Function of the Constitution”, (2009) 32 Australian Bar Review 138 at 141. 555  Tony Blackshield and George Williams QC, Australian Constitutional Law and Theory: Cases and Materials, Federation Press, 2006 at 664. 556  R v Trade Practices Commission, ex parte Tasmanian Breweries Ltd (1970) 123 CLR 361 at 394. 557  (1992) 176 CLR 1 at 67. 558  Ibid. 559  (1977) 138 CLR 1 at 15. 560  Blackshield and Williams QC, supra n555 at 662. 561  (1909) 8 CLR 330 at 357.  182   That is, unless there is a final determination of existing rights to be made, there is no exercise of “judicial power”. The High Court has also made clear a number of particular propositions in relation to what is or is not judicial review. For example, unlike in Canada, the High Court has found that giving an advisory opinion is not an exercise of judicial power 562 , but the making of control orders applied to terrorism suspects 563  and persons convicted of sexual offences after their release from prison 564  is.  Australian courts have also made it clear that judicial power may not be exercised by a body other than a Chapter III court, and a judicial body may not exercise executive power. This principle has been enunciated many times by the High Court, most notably in R v Kirby; ex parte Boilermakers’ Society of Australia565, which found that the Court of Conciliation and Arbitration could not exercise both the power to impose an award on the parties to an industrial dispute, and provide a final and binding legal interpretation of that award. The Boilermakers decision also makes clear that Chapter III is an exhaustive statement of the judicial power of the Commonwealth. The majority judges, Dixon CJ and McTiernan, Fullagar and Kitto JJ stated that Chapter III is “an exhaustive statement of the manner in which the judicial power of the Commonwealth is or may be vested ... No part of the judicial power can be conferred in virtue of any other authority or otherwise than in accordance with the provisions of Chap III”566.   562  Re Judiciary Act 1903-1920 and In re Navigation Act 1912-1920 (1921) 29 CLR 257 (popularly known as the “Advisory Opinions Case”). 563  Thomas v Mowbray [2007] HCA 33. 564  Fardon v Attorney-General (Queensland) (2004) 223 CLR 575. 565  (1956) 94 CLR 254. 566  Ibid at 270, cited with approval by Lim, supra n557 at paragraph 21.  183  An interesting illustration of this principle can be seen in Lim v Minister for Immigration, Local Government and Ethnic Affairs 567 . In that case, the applicant argued that a number of provisions of the Migration Act 1958 which provided for the mandatory detention of “designated persons” (roughly, persons who arrived in Australia without a visa or entry permit, most of whom would be boat arrivals, as opposed to persons who entered Australia legally and remained after their expiration of their entry permit) were unconstitutional on a number of grounds, including that orders for detention were inherently punitive in nature, and therefore amounted to an exercise of the judicial power of the Commonwealth. The High Court found that s.54L, which provided that a designated person must not be released from detention unless granted a visa or removed from Australia, and s.54N, which required an “officer” to detain a person reasonably suspected of being a designated person, without a warrant, were valid, as they were powers exercised incidentally to s.51(xix) of the Constitution 568 , and were not an exercise of judicial power. They could therefore be exercised by administrative decision-makers.  6.2.2 “Judicial power” and “merits review”  For the purposes of this thesis, the crucial issue is the distinction drawn by Australian courts between “merits review” and “judicial review”. There have been more cases than can possibly referred to in which courts have stated that they are not to interfere in the merits of a decision, but the reasons why this is the case, even leaving aside the pejorative use of the word “interfere”, are rather obscure.  567  Supra n557. 568  Paragraph 51(xix) permits the Commonwealth Parliament to make laws with respect to “naturalisation and aliens”.  184   Australian courts have generally taken the view that a court must stay out of consideration of the “merits” of a decision altogether. A frequently cited statement of the rule against merits review can be found in Attorney-General (NSW) v Quin, in which Brennan J (as he then was ) stated as follows 569 :  The duty and jurisdiction of the court to review administrative action do not go beyond the declaration and enforcing of the law which determines the limits and governs the exercise of the repository’s power. If, in so doing, the court avoids administrative injustice or error, so be it; but the court has no jurisdiction simply to cure administrative injustice or error. The merits of administrative action, to the extent that they can be distinguished from legality, are for the repository of the relevant power and, subject to political control, for the repository alone.  The key phrase is, of course, “to the extent that they [the merits] can be distinguished from legality”. Margaret Allars makes the following points on that issue570:  Three principles of judicial review qualify the operation of the legality/merits distinction. First, review for abuse of power where a decision is Wednesbury unreasonable is in practical terms review of the factual basis of the decision. The Wednesbury test of abuse of power permits the court to strike down a decision which is so unreasonable that no reasonable decision-maker could have reached it. This ground effectively sanctions as review for legality what is review of the merits in extreme cases of disproportionate decisions. Second, according to the “no evidence” principle, an agency makes an error of law in the course of making a finding of fact if there is a complete absence of evidence to support the factual inference. The third qualification to the legality/merits distinction is the jurisdictional fact doctrine.  More will be said about jurisdictional facts later in this chapter. Allars cites in support of her proposition that the Wednesbury test allows for review of “extreme cases of disproportionate  569  Supra n537 at 36. 570  Allars, supra n517 at 583-4.  185  decisions” the following fascinating passage from the judgement of Mason J (as he then was) in Minister for Aboriginal Affairs v Peko-Wallsend Ltd 571 :  The limited role of a court reviewing the exercise of an administrative discretion must constantly be borne in mind. It is not the function of the court to substitute its own decision for that of the administrator by exercising a discretion which the legislature has vested in the administrator. Its role is to set limits on the exercise of that discretion, and a decision made within those boundaries cannot be impugned … It follows that, in the absence of any statutory indication of the weight to be given to various considerations, it is generally for the decision-maker and not the court to determine the appropriate weight to be given to the matters which are required to be taken into account in exercising the statutory power … I say “generally” because both principle and authority indicate that in some circumstances a court may set aside an administrative decision which has failed to give adequate weight to a relevant factor of great importance, or has given excessive weight to a relevant factor of no great importance. The preferred ground on which this is done, however, is not the failure to take into account relevant considerations or the taking into account of irrelevant considerations, but that the decision is “manifestly unreasonable”.  Mason J starts with an orthodox statement that a court must not simply substitute its own opinion for that of an administrative decision-maker, but his Honour then admits that a court may set aside a decision on the basis that a decision-maker has given too little “weight” to a “relevant factor of great importance”. This is a clear admission that Australian courts do engage in review of the merits of a decision, even if only in limited circumstances. In fact, it could be argued that, given the prohibition on “reweighing” factors in the decision-making process in Suresh v Canada (Minister of Citizenship and Immigration) 572 , that Australian courts actually permit a greater intrusion into the merits of a case than Canada, in this area at least.   571  Supra n191 at 42. 572  Supra n186. Note the apparent reconsideration of this position, at least in relation to Charter rights, in Doré, supra n21.  186  In my opinion, the only difference distinguishing Wednesbury unreasonableness, variegated unreasonableness, proportionality and full review of the merits is the degree of deference provided to the decision-maker. That is, in Australia Wednesbury unreasonableness equates to the “patent unreasonableness” standard of review, while a correctness standard is applied to questions of law, for example. The judicial analysis is identical in each case, and the only difference is the degree of unreasonableness that must be demonstrated before the decision will be quashed. Australian courts simply provide a high degree of deference on findings of fact and matters of policy.  David Bennett QC, the then Commonwealth Solicitor-General, having defended the orthodox line early in his article, then makes a similar admission 573 :  The main problem arising in the application of the ground of unreasonableness is the subjectivity involved in drawing the line at which the merits of a decision end and the legality of the decision begins. The courts have made it clear that the ground of unreasonableness is extremely confined and requires something overwhelming, so that it should only be in exceptional circumstances that a court should interfere with the exercise of discretion by the decision-maker.  Stating that administrative discretion should only be interfered with in “exceptional circumstances” is the same thing as saying that a high degree of deference should be provided when reviewing the exercise of discretion. It is a review of the merits of the decision.   573  Bennett, supra n361 at 11.  187  6.2.3 “Merits review” and “review of the merits” distinguished  In my opinion, much of the difficulty in this area can be resolved by carefully distinguishing the terms “merits review” and “review of the merits”. David Bennett has defined the terms “merits review” and “judicial review” as follows574:  A merits review body will “stand in the shoes” of the primary decision-maker, and will make a fresh decision based upon all the evidence available to it. The object of merits review is to ensure that the “correct or preferable”575 decision is made on the material before the review body. The object of judicial review, on the other hand, is to ensure that the decision made by the primary decision-maker was properly made within the legal limits of the relevant power.  That is, it is the role of a primary decision-maker, or review tribunal, to make a new decision on the evidence before it. This is the same principle that the House of Lords enunciated in Huang v Secretary of State for the Home Department 576 , where it found that the administrative adjudicator reviewing a primary decision had not fulfilled their function when they focused on whether there was an error in that primary decision. Instead, the adjudicator’s role was to make a new decision on the basis of all the evidence, including evidence that may not have been available to the Home Department, before them. It does not, however, follow that there is therefore no role in examining the merits of a case for a court. The court’s role is judicial review – it is not the role of a court to simply reopen a case and make any order it sees fit. Its role is to review the administrative decision before it. If the court stays out of substantive decision-making and limits itself to a review of the decision and, if the decision is to be set aside, remits it to the appropriate decision-maker for reconsideration, this is an exercise of judicial and not executive power, even  574  Ibid at 7. 575  Drake, supra n362 at 68. 576  Supra n363.  188  if the “substance” or the “merits” of the decision are in question. It does not offend the Boilermakers principle that judicial power cannot be exercised by any other body than a Chapter III court, and nor may an administrative body exercise anything other than executive power. Sun v MIEA  A consideration of two Australian cases illustrates this point. In Sun v Minister for Immigration and Ethnic Affairs 577  the full court of the Federal Court, in my view, correctly exercised judicial power and not merits review. The applicant in Sun had been before the Refugee Review Tribunal (RRT) three times. The first decision, made by Member Fordham, accepted the truth of most (although not all) of the applicant’s claims, but found that he was not a refugee regardless. As the Department prepared to remove Mr Sun from Australia, the Chinese consulate refused to issue him with a passport, claiming they could not identify him. Mr Sun took this as further evidence of persecution, and applied again for refugee status 578 . This second application was also refused by the Department, and then by a different member of the RRT, Ms Ransome. Ms Ransome’s decision was ultimately set aside by consent, on the fairly technical basis that she had referred to an incorrect provision of the Migration Act 1958 in her decision.  The matter then went back for a third time to the RRT, this time before Member Smidt. Ms Smidt undertook a de novo review, in the face of Mr Sun’s objection that she should accept Mr Fordham’s finding that he was telling the truth about most of his claims, and that the only issue  577  (1997) 81 FCR 71. 578  This course of action would now be prohibited by s.48A of the Migration Act 1958, which, broadly speaking, prohibits repeat claims for refugee status onshore. Section 48A did not exist at the time of Mr Sun’s second application.  189  was whether he was a refugee on the basis of those facts. The full Federal Court found no error in Ms Smidt’s approach to the matter in that sense579. Ms Smidt, unlike Mr Fordham, found that Mr Sun had fabricated most of his claims and again refused his application for review.  The full Federal Court, however, set Ms Smidt’s decision aside on procedural fairness grounds. The question that remained was what to do with Mr Sun. There was uncontradicted evidence before the court (and Ms Smidt) that Mr Sun was suffering from post-traumatic stress disorder 580 , and the court was clearly concerned about putting him through another RRT hearing. The leading judgement was given by Wilcox and Burchett JJ, but North J, who concurred in the result, added as follows on the disposal of the case in the final paragraph of the judgement 581 :  Finally, I wish to refer to the observation by Wilcox J that the Minister should consider exercising his power under s 417 582  in favour of the appellant. As the comprehensive analysis made by Wilcox J in his judgment reveals, the Court has had the opportunity to examine the entire history of the appellant’s involvement in the review system. The circumstances of this case are exceptional and call for a quick and humane conclusion in favour of the appellant. No doubt, in many approaches to the Minister, cases are urged as “special cases” which are special only in the eyes of their proponents. The history of this case does make it special. It is special because the appellant has special problems of depression and post- traumatic stress disorder arising out of the circumstances of the case. It is special because there have been a number of errors in the review system. A number of these errors make it oppressive to require the appellant to have to face another hearing.   579  Supra n577 at 83. 580  Ibid at 81-3. 581  Ibid at 137. 582  Section 417 of the Migration Act 1958 gives the Minister a non-compellable power to grant a visa to a person who is refused refugee status by the RRT, on humanitarian grounds.  190  North J seemed sorely tempted to make some kind of declaration that Mr Sun was a refugee, but declined to do so. Making an order to this effect would go beyond judicial review of an administrative decision, and would be an exercise of executive power. The Guo litigation  Sun should be compared to the Guo cases in the full Federal Court and then the High Court. In the Full Federal Court, Einfeld J, having first ruled that an asylum-seeker should be found to be a refugee unless the contrary could be proved beyond reasonable doubt 583 , then made orders to the effect that Mr Guo and his wife Ms Pan were refugees and “entitled to the appropriate entry visas”584. Foster J agreed with the orders proposed by Einfeld J585.  In a rare 7-0 judgement, the High Court 586  overturned both the “beyond reasonable doubt” approach to refugee decision-making proposed by Einfeld J, and the orders his Honour proposed. The majority judges (Brennan CJ and Dawson, Toohey, Gaudron, McHugh and Gummow JJ) found on the first point that “[i]ngenious as his Honour’s approach may be, it is not supported by the terms of the Convention or the proper approach to administrative decision making in this context”587. On the power to make orders, the majority stated as follows588:   583  Guo (Full Federal Court), supra n532, paragraph 25 of the judgement of Einfeld J. 584  Ibid at paragraphs 63 and 68. Incidentally, there is not and never has been such a thing as an “entry visa” – Einfeld J apparently conflated the terms “visa” and “entry permit”. The latter kind of document was abolished with the passage of the Migration Reform Act 1992, which came into effect on 1 September 1994, and the visa has been the sole authority for entry to Australia by a non-citizen since that date. 585  Ibid at paragraph 54 of the judgement of Foster J. 586  Guo (High Court), supra n533. 587  Ibid at 574. 588  Ibid at 579.  191  The orders of the Full Court included a declaration “that both appellants are refugees and are entitled to the appropriate entry visas”. A declaration in these terms lacked utility because it did not specify with reference to the legislation the “appropriate entry visas” nor did it indicate any ready means of identification thereof. A declaration so loosely framed is objectionable in form.  Moreover, a declaration, even if drawn in specific terms, should not have been made. The Tribunal was empowered by s 166BC(1) of the Act to exercise all the powers and discretions conferred upon the primary decision-maker. The Act provided (s 22AA) for determination by the Minister that a person was a refugee, but this power was exercisable upon the Minister being satisfied that a person had that status or character. The rights of the appellants to the issue of visas, which the Full Court purported to declare with present effect, would only arise upon satisfaction of statutory conditions including the determination by the Minister under s 22AA or by the Tribunal under s 166BC. In those circumstances, the appropriate course would have been for the Full Court to set aside the orders of Sackville J and to return the matter to the Tribunal for determination in accordance with law.  Kirby J concurred as follows 589 :  [I]t is sufficient in my view to say that it was not appropriate for the Federal Court to adopt the course which the majority did. The proper course, legal error having been found, was to return the matter to the Tribunal. In that way, each of the relevant organs of government performs the functions proper to it. The Judicial Branch authoritatively clarifies and declares the law as it applies to the facts found. The Executive Branch, by power vested in it by the Legislature, performs its functions according to the law as so clarified and declared. Neither branch usurps or intrudes upon the functions proper to the other 590 .   589  Ibid at 600. 590  Kirby J made similar comments in Coal and Allied Operations Pty Ltd v Australian Industrial Relations Commission (2000) 203 CLR 194 at paragraph 87:  A conclusion to the foregoing effect does not result in a substitution by a court performing judicial review of its opinion on the merits. A court has no such power. Nor does it involve an unauthorised conversion of a limited process of judicial review into, in effect, an appeal for which no legislative warrant is provided. All that it permits is a quashing of the decision and order to the decision-maker to ensure that the latter performs the functions and exercises the powers and discretions in a way that the law envisages. This is nothing more than insistence on the rule of law. The lawful discharge of those functions, powers and discretions remains for the decision-maker so empowered and it alone.  192  While disagreeing with Kirby J in the past, I cannot fault him here. It is no part of the judicial function to make a decision of an administrative nature such as the grant of a visa. This is indeed a breach of the principle of separation of powers. This does not mean, however, that a court has no place in reviewing the merits of a decision, and leaving the substantive decision to the duly designated administrative decision-maker. This kind of reasoning complies with the admonition of Mason J in Minister for Aboriginal Affairs v Peko-Wallsend Ltd that “[i]t is not the function of the court to substitute its own decision for that of the administrator by exercising a discretion which the legislature has vested in the administrator”591 while still ensuring that the courts can truly review the merits of the decision.  Kirby J also noted in Guo as follows 592 :  [C]are must be exercised in applying decisions about the available and appropriate remedy apt to an appeal when the process before the Court is that of judicial review. Whereas on appeal a court will often enjoy the power and responsibility of substituting its decision for that under appeal, judicial review is designed, fundamentally, to uphold the lawfulness, fairness and reasonableness (rationality) of the process under review. It is thus ordinarily an adjunct to, and not a substitution for, the decision of the relevant administrator.  The appeal to “fairness” and “rationality” in his Honour’s judgement is the language of “review of the merits”, as opposed to “merits review”. The point that the court is an “adjunct” to administrative decision-making is an important one – the court is not to simply substitute its view for that of the decision-maker, a sentiment similar to that expressed in the UK in A v Secretary of State for the Home Department 593 and in Canada in CUPE 594 , amongst other cases.  591  Supra n191 at 42. 592  Supra n533 at 600. 593  Supra n106.  193  Stating that a court should not simply substitute its opinion for that of the decision-maker is simply another way of stating that deference should be afforded.  6.3 Deference and standards of review in Australia  Having established that there is nothing in the Australian Constitution to prevent courts reviewing the merits of an administrative decision, and that courts do in fact engage in review of the merits despite their protestations to the contrary, it is now necessary to establish the situations in which Australian courts will show deference to a decision-maker, and when they will impose a standard of correctness. Robert Dolehide has summed up the Australian position by stating that “Australian courts retain the exclusive authority to decide questions of law but generally give near-complete deference to administrators’ policy decisions”595. That is, in Dolehide’s view, Australian courts impose a correctness standard on all interpretations of law (including those of a “home statute”), and a reasonableness standard on exercises of discretion. In my opinion, Dolehide is basically correct, although some further distinctions need to be drawn. I propose to examine the degree of judicial deference shown to determination of “jurisdictional facts”, matters of jurisdiction generally, other questions of law, expertise, discretion and fact-finding, and then the impact of privative clauses.   594  Supra n5. 595  Dolehide, supra n513 at 1389.  194  6.3.1 Jurisdictional facts Definition  One particular kind of interpretation of law on which the High Court has firmly imposed a standard of correctness is the interpretation by an administrative body of “jurisdictional facts”. The term “jurisdictional fact” was defined in Enfield as a “criterion, satisfaction of which enlivens the power of the decision-maker to exercise a discretion ”596 . More recently, in M70/2011 and M106/2011 v Minister for Immigration and Citizenship, the High Court described the term as follows 597 :  The term “jurisdictional fact” applied to the exercise of a statutory power is often used to designate a factual criterion, satisfaction of which is necessary to enliven the power of a decision-maker to exercise a discretion. The criterion may be “a complex of elements”598.  When a criterion conditioning the exercise of statutory power involves assessment and value judgments on the part of the decision- maker, it is difficult to characterise the criterion as a jurisdictional fact, the existence or non-existence of which may be reviewed by a court 599 .  The decision- maker’s assessment or evaluation may be an element of the criterion or it may be the criterion itself.  Where a power is expressly conditioned upon the formation of a state of mind by the decision-maker, be it an opinion, belief, state of satisfaction or suspicion, the existence of the state of mind itself will constitute a jurisdictional fact 600 . If by necessary implication the power is conditioned upon the formation of an opinion or belief on the part of the decision-maker then the existence of that opinion or belief can also be viewed as a jurisdictional fact.     596  Supra n2 at paragraph 28. 597  Supra n469 at paragraph 57. 598  Enfield, supra n2 at paragraph 28. 599  Australian Heritage Commission v Mount Isa Mines Ltd (1997) 187 CLR 297 at 303-304. 600  Eshetu, supra n63 at 651-654 per Gummow J; Graham Barclay Oysters Pty Ltd v Ryan (2002) 211 CLR 540 at 609 per Gummow and Hayne JJ.  195 M70 v Minister for Immigration  M70 is a particularly fascinating illustration of the jurisdictional fact concept. The case was concerned with s.198A of the Migration Act 1958, and in particular with the government’s so- called “Malaysia solution”, which involved processing of asylum-seekers who arrived illegally in Australia in Malaysia, in return for Australia accepting persons from Malaysia who had been determined by the United Nations High Commission for Refugees (UNHCR) as having refugee status. Subsection 198A(1) provided that “an officer may take an offshore entry person601 from Australia to a country in respect of which a declaration is in force under subsection (3)”. Subsection 198A(3) then provided as follows:  The Minister may: (a) declare in writing that a specified country:   (i) provides access, for persons seeking asylum, to effective procedures for assessing their need for protection; and   (ii) provides protection for persons seeking asylum, pending determination of their refugee status; and   (iii) provides protection to persons who are given refugee status, pending their voluntary repatriation to their country of origin or resettlement in another country; and   (iv) meets relevant human rights standards in providing that protection; and (b) in writing, revoke a declaration made under paragraph (a).  The Minister made a declaration on 25 July 2011 providing that Malaysia was a “declared country”. The applicants sought a declaration that the declaration was invalid on the basis that ss.198A(3)(a)(i) – (iv) were each jurisdictional facts that did not exist, or alternatively that the Minister had misconstrued the meaning of the provisions. The Minister argued that as long as he  601  An “offshore entry person” is a non-citizen who arrived without a visa at an “excised offshore place” – these terms are defined in s.5 of the Migration Act 1958. Both plaintiffs had arrived without a visa at Christmas Island, which is an excised offshore place under paragraph (a) of the definition of “excised offshore place” in s.5(1).  196  made a declaration in good faith, that was sufficient – in other words, ss.198A(3)(a)(i) – (iv) were simply relevant considerations for the Minister, not jurisdictional facts.  The majority, consisting of Gummow, Hayne, Crennan and Bell JJ, found that ss.198A(3)(a)(i) – (iv) were jurisdictional facts. At paragraph 109 their Honours noted as follows:  It may readily be accepted that requirements to exercise the power in good faith and within the scope and for the purposes of the Act constrain the exercise of the Minister’s power.  But the submissions on behalf of the Minister and the Commonwealth that sub-pars (i) to (iv) of s 198A(3)(a) are not jurisdictional facts should not be accepted.  To read s 198A(3)(a) in that way would read it as validly engaged whenever the Minister bona fide thought or believed that the relevant criteria were met.  So to read the provision would pay insufficient regard to its text, context and evident purpose.  Text, context and purpose point to the need to identify the relevant criteria with particularity.  At paragraph 116 their Honours added as follows:  Contrary to the submissions of the Minister and the Commonwealth, the matters stated in s 198A(3)(a)(i) to (iii) are not established by examination only of what has happened, is happening or may be expected to happen in the relevant country. The access and protections to which those sub-paragraphs refer must be provided as a matter of legal obligation.  At paragraph 118 their Honours stated that a country could only meet the requirements of s.198A(3) if that country was a signatory to the Convention Relating to the Status of Refugees, which Malaysia was (and is) not. Since the jurisdictional facts that allowed the Minister to make a declaration under s.198A(3) did not exist, the declaration could not be lawfully made and was invalid.   197  French CJ and Kiefel J gave separate concurring judgements. Kiefel J found that ss.198A(3)(a)(i) – (iv) were jurisdictional facts602, but that even if they were not, the Minister had misconceived his power under s.198A(3) by relying on an undertaking by the Malaysian government to comply with certain human rights requirements, stating that “the enquiry under s 198A(3)(a) is as to the state of the laws of the country proposed to be the subject of a declaration and it is to be undertaken at the date of such declaration”603.  French CJ found that ss.198A(3)(a)(i) – (iv) were not jurisdictional facts, but took a similar view to the alternative approach of Kiefel J, finding that “the declaration must be a declaration about continuing circumstances in the specified country … [i]t cannot therefore be a declaration based upon, and therefore a declaration of, a hope or belief or expectation that the specified country will meet the criteria at some time in the future even if that time be imminent”604. As Malaysia was not a signatory to the Convention, it could not meet the requirements of s.198A(3)(a) at the time the declaration was made 605 .  Heydon J dissented.  This is a clear example of a correctness standard of review being applied to an executive decision. Whether or not the Minister’s interpretation of s.198A was reasonable or not was not even discussed. In the view of five of the seven judges, the “jurisdictional facts” simply did not exist and that was the end of the matter. No deference was given. The obvious result is that Australian courts, despite stressing the difference between judicial and merits review, have now  602  Supra n469 at paragraph 155. 603  Ibid at paragraph 156. 604  Ibid at paragraph 62. 605  Ibid at paragraphs 66-7.  198  adopted a quite intrusive standard of judicial review. Michael Tolley explains the situation as follows 606 :  In Australia, the High Court explicitly rejected the Chevron doctrine and has adopted an approach that favours wider judicial control of administrative action. The approach, based on the doctrine of “jurisdictional fact,” allows courts to review administrative action authorised by statute. Parliament can, and often does, stipulate that any action that it authorizes depends on the existence of various preconditions. Where the power depends on the existence of objective facts, the court on review is given the final say as to whether the required facts exist. This doctrine of jurisdictional fact has been used (manipulated some critics would say) by courts to justify a wide range of review of administrative interpretation of statutes.  While Tolley’s article was published well before the decision in M70, I think that the current Australian government would certainly count as one of his “critics” after this judgement. “The Minister is satisfied that …”  Another possibility is that legislation will provide that a decision-maker may not undertake a certain action unless he or she is “satisfied” that certain circumstances exist. In that case, the “satisfaction” can be construed as a jurisdictional fact. The obvious question that follows is whether that “satisfaction” has to be reasonable in some sense.  The most recent pronouncement on this subject came in Minister for Immigration and Citizenship v SZMDS 607 . The case involved a Pakistani applicant for a Protection Visa 608 , who  606  Michael Tolley, “Judicial Review of Agency Interpretation of Statutes: Deference Doctrines in Comparative Perspective”, (2003) 31 Policy Studies Journal 421 at 427. 607  Supra n4.  199  claimed a well-founded fear of persecution on the basis of his membership of a particular social group, namely homosexuals. The RRT rejected his claim, refusing to accept that he was even homosexual. Section 65 of the Migration Act 1958 provided (and still provides) that if the Minister is “satisfied” that the applicant meets all criteria for the grant of a visa then he or she must grant it, and if not, the application must be refused.  The RRT decision was set aside by the Federal Court, which found that the “Tribunal’s conclusion that the applicant was not a homosexual was based squarely on an illogical process of reasoning”609. On appeal to the High Court, the Minister argued that the RRT’s findings were not illogical, and that even if they were, this did not amount to a “jurisdictional error” (a term that will be explained more fully in the next section).  The leading judgement was given by Crennan and Bell JJ, with whom Heydon J agreed. Gummow ACJ and Kiefel J gave separate reasons, concurring on this point. Crennan and Bell JJ started by referring to Minister for Immigration and Multicultural Affairs v SGLB 610 , which had found that the Minister’s satisfaction, referred to in s.65, was a jurisdictional fact. The key passage in the judgement relating to jurisdictional facts is at paragraphs 119 and 120, as follows:  119.  Whilst the first respondent accepted that not every instance of illogicality or irrationality in reasoning could give rise to jurisdictional error, it was contended that if illogicality or irrationality occurs at the point of satisfaction (for the purposes of s 65 of the Act) then this is a jurisdictional fact and a jurisdictional error is established. This submission should be accepted …   608  Under s.36 of the Migration Act 1958, the key criterion for the grant of a Protection Visa is that the applicant has been found to be a refugee as defined by the Convention Relating to the Status of Refugees. 609  SZMDS v Minister for Immigration and Citizenship [2009] FCA 210 at paragraph 29. 610  (2004) 207 ALR 12.  200  120. An erroneously determined jurisdictional fact may give rise to jurisdictional error. The decision maker might, for example, have asked the wrong question or may have mistaken or exceeded the statutory specification or prescription in relation to the relevant jurisdictional fact. Equally, entertaining a matter in the absence of a jurisdictional fact will constitute jurisdictional error.  That is, while Australian courts (as we will see) will generally give deference to findings of fact by administrative decision-makers, this is not the case with findings of jurisdictional facts. Therefore, illogicality or irrationality in finding of jurisdictional facts is a jurisdictional error and will result in the decision under review being set aside. However, Crennan and Bell JJ found that the RRT’s findings were open to it on the evidence before it, and that “a decision will not be illogical or irrational if there is room for a logical or rational person to reach the same decision on the material before the decision maker”611. The Federal Court decision was therefore set aside and the RRT decision restored.  The reasons of Gummow ACJ and Kiefel J are not relevantly different to those of Crennan and Bell JJ, but their Honours did note some developments overseas. At paragraph 26 they state that “[i]n England the distinction between jurisdictional and non-jurisdictional facts has fallen into deep disfavour and broader notions of the scope of judicial review have been developed and applied by the English courts”, and that the Minister’s submissions in the case at bar “in significant measure sought to discourage any such development by this Court”. The Minister’s submissions, to this extent at least, seem to have been accepted by the High Court.     611  Supra n4 at paragraph 135.  201  6.3.2 Deference on questions of law generally  While Canadian courts may give deference to an administrative decision-maker in at least some matters of law, most particularly when interpreting a “home statute”, Australian courts do not. In general, no deference on questions of law is given at all. Sackville J, writing extrajudicially, has stated as follows 612 :  But … two principles have been accepted, generally without challenge, as fundamental in determining the proper scope of judicial review. The first is that courts exercising powers of judicial review must not intrude into the “merits” of administrative decision-making or of executive policy making. The second is that it is for the courts and not the executive to interpret and apply the law, including the statutes governing the power of the executive. These can be regarded as the twin pillars of judicial review of administrative action in Australia.  Even more bluntly, Hayne J, also writing extrajudicially, has stated that “[t]he whole system of Government in Australia is constructed upon the recognition that the ultimate responsibility for the final definition, maintenance and enforcement of the boundaries within which governmental power may be exercised rests upon the judicature”613. In Enfield Gaudron J stated that “that there is very limited scope for the notion of ‘judicial deference’ with respect to findings by an administrative body of jurisdictional facts”614, which, as we have seen, is regarded as a question of law, not fact-finding.   612  Justice Ronald Sackville, “The Limits of Judicial Review of Executive Action – Some Comparisons between Australia and the United States”, (2000) 28 Federal Law Review 315 at 315-6. 613  Hayne J, supra n3 at 76. 614  Supra n2 at paragraph 59.  202  There is, of course, a real difficulty in determining the difference between an error of law and an error of fact in the first place. Sir Anthony Mason, the former Chief Justice of the High Court, has written extrajudicially as follows 615 :  The difficulty of distinguishing between questions of law, on the one hand, and questions of fact, not to mention questions of policy, is notorious. This difficulty unquestionably creates complications for a system of administrative law such as ours which requires questions of law and questions of fact to be treated differently. In the United States and Canada, the assumption that there is a distinction has been challenged. So far that is not the position in Australia, where the High Court has noted that the distinction “is a vital distinction in many fields of law”, while acknowledging that “no satisfactory test of universal application has yet been formulated”616.  6.3.3 Expertise  Despite the considerations mentioned above, there is a judicial trend in Australia to defer, at least on matters of fact and discretion, to expert decision-makers. This reasoning seems to have been clearly expressed for the first time in Collector of Customs v Agfa-Gevaert Ltd 617 , in which the High Court was concerned with the assessment by the Collector that certain goods imported by Agfa were subject to duty. The case turned on the interpretation of a Commercial Tariff Concession Order (CTCO), which was an instrument made under s.269C of the Custom Tariff Act 1987. The effect of a CTCO was that goods that would otherwise be subject to import duty were exempted.   615  Mason CJ, supra n382  at 55. 616  Citing Collector of Customs v Agfa-Gevaert Ltd (1996) 186 CLR 389 at 394. 617  Supra n616.  203  Agfa sought exemption from duty of products it called “types 8 and 9 photographic paper”. A CTCO exempted such products if they used a “silver dye bleach reversal process” and operated by “having the image dyes incorporated in the emulsion layers”. The High Court, in a unanimous judgement, found that the individual words in these terms should be defined in terms of their “trade meaning”, and stated as follows618:  [W]hen construing a composite phrase which does not have a trade meaning, it will ordinarily make sense for a court or tribunal to take notice of the trade meaning of a word or words within that expression, provided such an interpretation does not lead to a result which is absurd in the sense that the result may be unworkable or impracticable, inconvenient, anomalous or illogical, futile or pointless, or artificial. Consideration of the trade meaning of individual words in such cases is more likely than not to lead to the interpretation that the makers of the instrument had in mind.  Further, contrary to Agfa’s submission, using the trade meaning of individual words in a composite phrase having no special meaning as a whole does not involve a failure to construe the phrase “as a whole”. It simply does not follow, as a matter of logic or common-sense, that the division of a composite expression into parts which are interpreted by reference to their trade meaning, ordinary meaning or a combination thereof necessarily means that a court or tribunal has failed to construe an expression by reference to its meaning as a whole … It remains to determine whether the finding of the Tribunal was permissible as a matter of law. We think that it was.  Even though the words “deference” and “expertise” do not appear in the judgement, this decision reads very much as if the High Court reasoned that it should accept the interpretation given to the CTCO by the Controller of Customs, as that officer had expertise in the interpretation of technical terms such as “silver dye bleach reversal process” that the court did not.  Australian courts tend to refer to the “weight” to be given to certain findings of an administrative decision-maker, rather than “deference to expertise”. However, the two formulations lead to  618  Ibid at 409-10.  204  much the same results. In Minister for Immigration and Ethnic Affairs v Wu Shan Liang, Kirby J, who wrote a separate judgement from the majority but concurred in the result, commented as follows 619 :  [T]here are additional reasons for restraint and resistance to any temptation to turn a case of judicial review into, effectively, a reconsideration of the merits. Often, the decision-maker will have more experience in the consistent application of applicable administrative rules to achieve fairness to a wider range of people than typically come before the courts … In reviewing reasons and decisions of the delegates of the Minister, such as are in contest in this appeal, it is appropriate to take into account the fact that they were not untrained laymen. They had obvious expertise for the performance of their functions. By the evidence, they also had legal advice available to them.  Note how Kirby J links the requirement not to engage in “merits review” with respect for the expertise of decision-makers. Gummow J picked up this kind of reasoning and stated as follows in Minister for Immigration and Ethnic Affairs v Eshetu 620 :  [W]hilst it is for this Court to determine independently for itself whether in a particular case a specialist tribunal has or lacks jurisdiction, weight is to be given, on questions of fact and usage, to the tribunal’s decision, the weight to vary with the circumstances. The circumstances will include such matters as the field in which the tribunal operates, the criteria for appointment of its members, the materials upon which it acts in the exercise of its functions and the extent to which its decisions are supported by disclosed processes of reasoning. A similar doctrine has been developed by the Supreme Court of Canada, at least with respect to findings of non-jurisdictional fact 621 .  There is one somewhat confusing element of the judgement of Gummow J, in that “expertise” is not a free-standing “doctrine” in Canadian law. Expertise is instead one important facet of the  619  (1996) 185 CLR 259 at paragraph 25 of the judgement of Kirby J. 620  Supra n63 at 655. 621  Referring to Bradco, supra n167 at 335; Pezim, supra n168 at 591-592; Ross v New Brunswick School District No 15 [1996] 1 SCR 825 at 846-847; Westcoast Energy Inc v Canada (National Energy Board) [1998] 1 SCR 322 at 353-355, 414-415.  205  doctrine or principle of deference, in the sense that the degree of expertise on the part of the decision-maker directly affects the degree of deference provided. As we have seen, this is an unwritten constitutional principle that corresponds with the separation of powers and the principles of the rule of law and democracy.  It is also interesting that Gummow J would not automatically assume an administrative decision- maker or tribunal to be expert in its field, and would instead look for “corroborative” evidence. The focus on the means by which a tribunal’s members are appointed is particularly interesting, and may go some way towards addressing David Mullan’s concern about “political hacks” being appointed to tribunals 622 . In Enfield itself, the majority stated as follows 623 :  Questions may arise, within the jurisdiction of an administrative tribunal and upon a settled construction of the applicable legislation, as to the side of the line on which a case falls.  The question may be one to be decided on the particular primary facts which are largely undisputed and where little can be gained from a detailed examination of previous decisions.  In such instances, this Court has said that, in a proceeding in the original jurisdiction of a court on “appeal” from that tribunal, the “court should attach great weight to the opinion of the [tribunal]”624.  At paragraph 47 the majority made comments very similar 625  to those of Gummow J in Eshetu:  The weight to be given to the opinion of the tribunal in a particular case will depend upon the circumstances.  These will include such matters as the field in which the tribunal operates, the criteria for appointment of its members, the materials upon which it acts in exercising its functions and the extent to which its decisions are supported by disclosed processes of reasoning.  A similar view appears to be taken by the Supreme Court of Canada 626 .   622  Mullan, “Deference: Is It Useful Outside Canada?”, supra n125 at 52. 623  Enfield, supra n2 at paragraph 45. 624  Referring to Muller, supra n538 at 41. 625  Margaret Allars views the two tests as identical – see Allars, supra n517 at 587. 626 Referring to Bradco, supra n167 at 335; Pezim, supra n168 at 591-592; Ross, supra n621; Westcoast Energy, supra n621 at 353-355, 414-415.  206  Gaudron J noted that “there may be situations where the evidence before the court is the same or substantially the same as that before the primary decision-maker and minds might reasonably differ as to the finding properly to be made on that evidence”627. Her Honour added that “[i]n that situation a court may, but need not, decline to make a different finding from that made by the primary decision-maker, particularly if the latter possesses expertise in the area concerned”628.  Finally, Osland v Secretary to the Department of Justice 629  involved an application under the Freedom of Information Act 1982 (Victoria) for access to documents relating to a decision to refuse the applicant’s request for an executive pardon. Heather Osland had been convicted of the murder of her violent and abusive husband, in a case that resulted in an (unsuccessful) appeal to the High Court 630 . The Victorian Department of Justice had refused her FOI application on the basis that the documents she sought were protected by Legal Professional Privilege, and this decision was upheld by the Victorian Civil and Administrative Tribunal (VCAT). At paragraph 12 of the judgement, the majority (Gleeson CJ and Gummow, Heydon and Kiefel JJ) noted that the response to Mrs Osland’s petition was informed by its legal professionals, “their legal expertise being relevant to the weight to be attached to their opinions”. However, in this case the High Court found that the Victorian Court of Appeal should have examined the relevant documents itself to determine if privilege applied, and remitted the matter to the court for reconsideration 631 .  627  Enfield, supra n2 at paragraph 60. 628  Ibid. 629  [2008] HCA 37. 630  Osland v The Queen (1998) 197 CLR 316. 631  Supra n629 at paragraph 58.  207   In summary, Enfield saw the majority of the High Court adopt a form of deference to expertise, at least in relation to findings of fact and exercise of discretion. The High Court has specifically referred to (pre-Dunsmuir) Canadian jurisprudence to support this line of reasoning, and the refusal to formally move to a Canadian approach of substantive review in SZMDS 632  does not appear to have altered this reasoning.  6.3.4 Discretion and fact-finding generally  On the other hand, lower courts have been regularly warned by the High Court to defer to the written reasons of administrative decision-makers, on matters of fact and discretion, as far as possible. The best-known instance of the High Court making such a pronouncement was in Minister for Immigration and Ethnic Affairs v Wu Shan Liang 633 , yet another case involving a failed asylum-seeker in Australia. The full Federal Court 634  had found, despite the fact that the decision-maker had clearly stated that he found that the applicants did not have a “well-founded fear of persecution”, that the decision-maker had in fact decided the matter on a balance of probabilities standard, and not on the “real chance” test propounded by Chan v Minister for Immigration and Ethnic Affairs 635 .  On appeal, the majority of the High Court (Brennan CJ and Toohey, McHugh and Gummow JJ) stated that “the reasons of an administrative decision-maker are meant to inform and not to be  632  Supra n4 at paragraph 28. 633  Supra n619. 634  Wu Shan Liang v Minister for Immigration and Ethnic Affairs (1995) 130 ALR 367. 635  (1989) 169 CLR 379.  208  scrutinised upon over-zealous judicial review by seeking to discern whether some inadequacy may be gleaned from the way in which the reasons are expressed”636, and found that the Full Federal Court had erred in reading the reasons of the decision-maker in the way they had.  Kirby J concurred, noting that “[i]t is erroneous to adopt a narrow approach, combing through the words of the decision-maker with a fine appellate tooth-comb, against the prospect that a verbal slip will be found warranting the inference of an error of law”637, and that “[t]his admonition has particular application to the review of decisions which, by law, are committed to lay decision- makers, ie tribunals, administrators and others”638.  This form of reasoning is quite apparent in the Supreme Court of Canada decision in Newfoundland and Labrador Nurses’ Union v Newfoundland and Labrador (Treasury Board), where Abella J, writing for the court, emphasised that the purpose of reasons is to enable a court to follow the reasoning of the decision-maker, and to determine whether the decision reached falls within the Dunsmuir “possible, acceptable outcomes”639. Reasons are not required to be perfect640.  Wu is simply the best example of a long line of judicial reasoning on this point. For example, in Collector of Customs v Pozzolanic Enterprises Pty Ltd the Full Federal Court stated that “[t]he Court will not be concerned with looseness in the language of the Tribunal nor with unhappy phrasing of the Tribunal’s thoughts ... [t]he reasons for the decision under review are not to be construed minutely and finely with an eye keenly attuned to the perception of error”641. Finally, in a passage that appears frequently in Department of Immigration and Citizenship  636  Supra n619 at paragraph 31 of the majority judgement. 637  Ibid at paragraph 24 of the judgement of Kirby J. 638  Ibid. 639  Supra n263 at paragraph 16. 640  Ibid at paragraph 18. 641  (1993) 43 FCR 280 at 287.  209  training materials, the Federal Court stated in Obejas v Minister for Immigration and Ethnic Affairs that “[t]he reasons of the Tribunal are not a Statute … [t]hey are not to be parsed and analysed as if they were”642.  Kirby J has also noted that there is a distinct similarity between the North American approach to deference and the Wu Shan Liang principle. In Minister for Immigration and Multicultural Affairs v Singh 643  the Minister had argued that a reviewing court should show deference to the expertise of the Administrative Appeals Tribunal (AAT) in making a finding that an applicant for a Protection Visa was excluded from refugee status under Article 1F(b) of the Convention. Kirby J observed as follows 644 :  Where a repository of statutory power has been designated by the Parliament as the decision-maker, required to determine whether critical facts do or do not exist, courts, without clear authority to go further, should restrict their intervention to cases that fall within the categories that have been identified as evidencing legal error. In the United States, such restraint upon appellate intervention is often described in terms of the “deference” owed by courts of law to administrators entrusted with primary decision-making in that country. This principle is especially applicable in the context of immigration decisions. In this Court there are suggestions of a similar approach in the repeated expressions of caution against over-zealous scrutiny of administrative reasons, nominally for error of law, that finds such error in infelicitously expressed or otherwise imperfect reasons 645 .  Kirby J, however, found that the AAT member had misconstrued the meaning of Article 1F(b) and had therefore made an error of law. His Honour noted that “where the decision-maker has given reasons that indicate that the finding was arrived at by a misunderstanding of the  642  [1995] FCA 1458 at paragraph 21. 643  (2002) 209 CLR 533. 644  Ibid at paragraphs 131 and 132. 645  Referring to Wu Shan Liang, supra n619 at 271-272, 291.  210  applicable legal test, or where the finding resulted from a failure to apply correctly the language of that phrase to the facts as found, a court reviewing for error of law is entitled to intervene”646. The Minister’s appeal was therefore dismissed.  6.4 Privative clauses in the High Court  As we have seen, the High Court has generally taken the view that deference is not to be given on questions of law. The situation changes, however, when a privative clause is inserted into the relevant legislation. The clash between the insertion of a privative clause into legislation and the constitutional guarantee of judicial review in s.75 of the Constitution has been considered on several occasions by the High Court. I will now turn to examine the High Court’s approach to privative clauses, and suggest some reasons for the court’s recent dilution of their effect.  6.4.1 Cases from federation to 1945  Until 1945, the High Court’s approach was to find that privative clauses were unconstitutional because they offended s.75 of the Constitution. In R v Commonwealth Court of Conciliation and Arbitration; Ex parte Whybrow 647  the court simply noted that s.31 of the Conciliation and Arbitration Act 1904, which provided that “[n]o award of the Court shall be challenged, appealed against, reviewed, quashed or called into question in any other Court on any account whatsoever”, did not mention prohibition or mandamus, and made an order of prohibition against the Court of Conciliation and Arbitration. However, in R v Commonwealth Court of Conciliation  646  Supra n643 at paragraph 134. 647  (1910) 11 CLR 1.  211  and Arbitration (“the Tramways Case”)648, in the face of a better drafted privative clause, the court made itself unambiguously clear, with a unanimous finding that such clauses conflicted with s.75 of the Constitution and were invalid. Powers J stated that “[t]he power directly conferred on the High Court by the Court as original jurisdiction cannot be taken away by the Commonwealth Parliament”649. The Tramways Case was upheld as late as 1942, in Australian Coal and Shale Employees Federation v Aberfield Coal Mining Co Ltd 650 .  6.4.2 Reconciliation between the rule of law and Parliamentary supremacy: the Hickman case  However, from 1945 the court attempted to reconcile privative clauses and s.75 by finding that a privative clause could not oust judicial review, but it expanded the situations in which an administrative decision would be found to be valid by a court. This line of authority, known as the Hickman approach for the leading case, R v Hickman ex parte Fox and Clinton 651 , was basically the approach taken by the High Court from 1945 to 2003. The key part of the Hickman judgement can be found in the judgement of Dixon CJ as follows 652 :  Both under Commonwealth law, and in jurisdictions where there is a unitary constitution, the interpretation of provisions of the general nature of reg 17 is well established. They are not interpreted as meaning to set at large the courts or other judicial bodies to whose decision they relate. Such a clause is interpreted as meaning that no decision which is in fact given by the body concerned shall be invalidated on the ground that it has not conformed to the requirements governing its proceedings or the exercise of its authority or has not confined its acts within  648  (1914) 18 CLR 54. 649  Ibid at 86. 650  (1942) 66 CLR 161. 651  Supra n80. 652  Ibid at 614-5.  212  the limits laid down by the instrument giving it authority, provided always that its decision is a bona fide attempt to exercise its power, that it relates to the subject matter of the legislation, and that it is reasonably capable of reference to the power given to the body.  Later cases added that the administrative decision in question must also conform to mandatory (or “inviolable”) requirements within the Act itself653. For example, if the Act itself required that certain procedural rights be given to an applicant, failure to follow those procedures would have the result that the privative clause would not protect the decision. However, the principles basically remained unchanged until 2003.  6.4.3 The jurisdictional error qualification: the evisceration of Hickman in S157  Plaintiff S157/2002 v Commonwealth of Australia 654  involved a constitutional challenge to the validity of s.474 of the Migration Act 1958, and administrative challenges to a number of decisions that were defended on the basis of this section. Subsections 474(1) and (2) relevantly provided as follows at the time of the judgement:  (1)  A privative clause decision: (a)  is final and conclusive; and (b)  must not be challenged, appealed against, reviewed, quashed or called in question in any court; and (c) is not subject to prohibition, mandamus, injunction, declaration or certiorari in any court on any account.  (2)  In this section, privative clause decision means a decision of an administrative character made, proposed to be made, or required to be made, as the case may be, under this Act or under a regulation or other  653  R v Metal Trades Employers’ Association, ex parte Amalgamated Engineering Union, Australian Section (1951) 82 CLR 208 at 248. 654  Supra n81.  213  instrument made under this Act (whether in the exercise of a discretion or not) …  Subsection 474(3) made it clear that a decision to grant or refuse a visa was a “privative clause decision”. Subsections 474(4) and (5) listed a number of decisions that were taken not to be privative clause decisions, none of which is relevant for the purpose of this discussion.  The applicants argued that s.474 conflicted with s.75 of the Constitution and was therefore invalid, or alternatively that s.474 did not protect “jurisdictional errors”, a term that will be explained shortly. The High Court rejected the first argument but accepted the second, which left s.474 “on the books”, but rendered it of almost no effect.  The leading judgement was given by Gaudron, McHugh, Gummow, Kirby and Hayne JJ. At paragraph 73, their Honours stated that:  A privative clause cannot operate so as to oust the jurisdiction which other paragraphs of s 75 confer on this Court, including that conferred by s 75(iii) in matters “in which the Commonwealth, or a person suing or being sued on behalf of the Commonwealth, is a party”. Further, a privative clause cannot operate so as to allow a non-judicial tribunal or other non-judicial decision-making authority to exercise the judicial power of the Commonwealth 655 . Thus, it cannot confer on a non-judicial body the power to determine conclusively the limits of its own jurisdiction.  Their Honours stated at paragraph 76 that an administrative decision affected by jurisdictional error is a legal nullity, referring to Minister for Immigration and Multicultural Affairs v Bhardwaj 656. Therefore, a “decision” affected by a privative clause is only a putative decision and cannot be a “privative clause decision” for the purposes of s.474. When read in this way,  655  Referring to Boilermakers, supra n565. 656  (2002) 76 ALJR 598 at paragraph 51.  214  there was no conflict between s.474 and s.75 of the Constitution, and the provision was therefore constitutionally valid. Indeed, s.75(v) was reaffirmed to amount to “an entrenched minimum provision of judicial review”657, “assuring to all people affected that officers of the Commonwealth obey the law and neither exceed nor neglect any jurisdiction which the law confers on them”658, in a passage remarkably similar to the Canadian case of Crevier659.  The remaining issue was the definition of “jurisdictional error”. Curiously, none of the judgements referred to the High Court’s decision of just two years previously, Minister for Immigration and Multicultural Affairs v Yusuf 660 . In that case, McHugh, Gummow and Hayne JJ defined the term as follows at paragraph 82:  It is necessary, however, to understand what is meant by “jurisdictional error” under the general law and the consequences that follow from a decision-maker making such an error. As was said in Craig v South Australia 661 , if an administrative tribunal (like the Tribunal):  “falls into an error of law which causes it to identify a wrong issue, to ask itself a wrong question, to ignore relevant material, to rely on irrelevant material or, at least in some circumstances, to make an erroneous finding or to reach a mistaken conclusion, and the tribunal’s exercise or purported exercise of power is thereby affected, it exceeds its authority or powers. Such an error of law is jurisdictional error which will invalidate any order or decision of the tribunal which reflects it.”  “Jurisdictional error” can thus be seen to embrace a number of different kinds of error, the list of which, in the passage cited from Craig, is not exhaustive 662  … if an error of those types is made, the decision-maker did not have authority to make the decision that was made; he or she did not have jurisdiction to make it.   657  Supra n81 at paragraph 104. 658  Ibid at paragraph 105. 659  Supra n48. 660  (2001) 206 CLR 323. 661  Supra n121 at 179. 662  Aala, supra n49 at paragraph 163.  215  The kinds of “jurisdictional error” identified by Craig and Yusuf are very wide, and really endorse the Anisminic 663  approach as far as one can without expressly abolishing the distinction between jurisdictional and non-jurisdictional errors. The Full Federal Court has even recently found that a failure to afford an adjournment in proceedings, at least in some circumstances, can amount to a jurisdictional error and grounds to quash a decision 664 . Craig itself declined to apply Anisminic, as this case was concerned with the decision of an inferior court and not a tribunal, but it nowhere said that Anisminic was wrongly decided or had no application in Australia.  The judgement of Gleeson CJ is of interest primarily for his Honour’s attempt to distinguish between jurisdictional and non-jurisdictional errors. Gleeson CJ suggested that a jurisdictional error is one that is clear or “manifest” in some way665:  The concept of “manifest” defect in jurisdiction, or “manifest” fraud, has entered into the taxonomy of error in this field of discourse. The idea that there are degrees of error, or that obviousness should make a difference between one kind of fraud and another, is not always easy to grasp. But it plays a significant part in other forms of judicial review. For example, the principles according to which a court of appeal may interfere with a primary judge’s findings of fact, or exercise of discretion, are expressed in terms such as “palpably misused [an] advantage”, “glaringly improbable”, “inconsistent with facts incontrovertibly established”, and “plainly unjust”. Unless adjectives such as “palpable”, “incontrovertible”, “plain”, or “manifest” are used only for rhetorical effect, then in the context of review of decision-making, whether judicial or administrative, they convey an idea that there are degrees of strictness of scrutiny to which decisions may be subjected. Such an idea is influential in ordinary appellate judicial review, and it is hardly surprising to see it engaged in the related area of judicial review of administrative action.   663  Anisminic, supra n85. 664  Li, supra n64. 665  Supra n81 at paragraph 13.  216  The majority judges in S157 came to the conclusion that a failure of procedural fairness was a “jurisdictional error” and therefore s.474 did not protect the Tribunal decision from such a claim 66