Ln Compliance with the Canadian Privacy Legislation Sorne Supporting Forms May Have Been Removed from This .Dissertation

ln compliance with the Canadian Privacy Legislation sorne supporting forms may have been removed from this .dissertation.

While these forms may be included . in the document page count, their rernoval does not represent any loss of content from the dissertation.

Appealing to Higher Authority:

Explaining the Federal Government's Appeals to the Supreme Court of Canada in Charter Cases

Matthew A. Hennigar

Department of Political Science Mc Gill University, Montreal October 2002

A thesis submitted to the Faculty of Graduate Studies and Research in partial fulfillment of the requirements of the degree ofDoctorate ofPhilosophy

395 Wellington Street 395, rue Wellington Ottawa ON K1A ON4 Ottawa ON K1A ON4 Canada Canada

Your file Votre référence ISBN: 0-612-88489-9 Our file Notre référence ISBN: 0-612-88489-9

The author has granted a non L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sell reproduire, prêter, distribuer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/film, de reproduction sur papier ou sur format électronique.

The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts from it Ni la thèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou aturement reproduits sans son permission. autorisation.

Canada ABSTRACT

The adoption of the Charter ofRights and Freedoms has increased the policy

importance of litigation, and, accordingly, of those government lawyers who handle su ch

litigation. In particular, policies and laws may stand or faIl based on whether the

government appeals its losses in the lower courts to the Supreme Court of Canada. This

dissertation seeks to determine whether the federal government' s appeal decisions are

based on a rational weighing of multiple factors related to costs, case importance, and the

prospect of securing access to, and victory in, the Supreme Court. This first requires

ascertaining the locus of litigating authority and the decision making process within the

government. Surveys and personal interviews of government counsel reveal that the

Department of Justice Canada, which holds a near-monopoly over Ottawa's litigation,

combines decentralisation by region and policy area in the conduct of litigation with a

centralised appeal decision making pro cess. Quantitative and qualitative analyses of the

government's lower court losses and Supreme Court appeals confirm the dissertation's

central assumption that the federai government' s appeals are calculated choices, with case importance, avoiding losses on appeal, and protecting governmental authority emerging as the strongest motivations. Further, the dissertation situates these findings in the context of the contemporary "dialogue" debate, providing the first systematic analysis of government dialogue with lower appeal courts, and incorporating the concept of appeals as a form of dialogue. The study finds a significant degree of dialogue between the federal government and the lower courts, especially following highly activist judicial rulings. This is consistent with findings from the preceding sections, that Justice Department lawyers actively defend the government' s policy-making authority.

Il RÉSUMÉ

L'adoption de la Charte canadienne des droits et libertés a augmenté l'importance

politique du litige, et, en conséquence, des avocats gouvernementaux qui traitent ce litige.

En particulier, la probabilité que les politiques et les droits survivent ou périssaent peut

dépendre sur si le gouvernement fait appel à la Cour Suprême lors de ses pertes dans les

cours inférieures. Cette thèse vise à déterminer si les arrêts d'appel du gouvernement du

Canada sont baser sur plusiers facteurs concernant les coûts, l'importance de l'affaire, et la probabilité de gagner entrée à, et réussir dans, la Cour Suprême. Ce premier exige la détermination de la portée d'autorité gouvernemental de mettre en litige, et le processus de décisions gouvernementales. Les interviews avec des avocats gouvernementaux révèlent que le Ministère de Justice Canada, qui possède un quasi-contrôle sur le litige du gouvernement fédéral, joint la décentralisation par région et par catégorie politique dans le traitement du litige avec un processus centralisé pour les arrêts d'appel. Les analyses quantitatives et qualitatives des pertes et des appels par le gouvernement confirment l'hypothèse centrale de la thèse que les appels sont des choix refléchis. Les motivations les plus fortes pour faire appel sont l'importance du cas, le désir d'éviter les pertes en appel, et la protection de l'autorité gouvernemental. En plus, la thèse situe ces découvertes dans la contexte du débat contemporain sur le "dialogue," et documente la première analyse systématique du dialogue entre le gouvernement et les cours inférieures, incorporant l'idée de l'appel dans la définition de dialogue. Cette étude trouve un degré

élevé de dialogue, surtout à la suite d'arrêts judiciaires hautement activistes, ce qui est en accord avec les observations des sections précédentes, notamment que les avocats de

Ministère de Justice Canada défendent activement l'autorité politique du gouvernement.

III CONTENTS

Acknowledgments v

Introduction 1

1. From Independence to Litigation Strategy: Scholarly Treatment of 11 Government Lawyers

2. The Federal Government' s Appeals Pro cess 42

3. Modelling the Decision to Appeal: Factors and Hypotheses 59

4. Analysing the Decision to Appeal: Data and Statistical Results 84

5. Factors in the Decision to Appeal: A Second Look 116

6. Government Appeals as "Dialogue": Expanding a Contemporary Debate 151

Conclusion 187

Bibliography 196

IV ACKNOWLEDGMENTS

Many people have provided support and advice during the writing of tbis

dissertation who deserve my sincere thanks. F oremost among my many friends and family

in this regard is my partner Kathleen, who, as a feUow academic, was particularly

understanding of the travails of a doctorate, but never let me lose perspective. Deepest

thanks as weU to my parents and brothers Jay and Troy, who were probably starting to worry that 1 was going to be a student forever. Thank you aU for your love and support.

1 am grateful to have benefitted from comments on preliminary drafts from

several individuals over the course ofwriting this dissertation, including Elisabeth

Gidengil, Steven Wasby, Matthew Wetstein, Cythia Ostberg, Troy Riddell, and Gerald

Baier. 1 owe particular thanks to Elisabeth Gidengil, for giving me the opportunity to greatly improve my familiarity with methods of quantitative analysis through an extended

research assistantship. Thanks too to the Social Sciences and Humanities Research

Council of Canada, the Stanfield Dissertation Fellowship, and the Department ofPolitical

Science at McGill University for their financial support.

Several other people deserve acknowledgment for their help in a variety offorms over the past six years. 1 greatly appreciate the time and effort taken by those Justice

Department officiais who replied to my surveys and who consented to be interviewed; their co-operation was essential to understanding the bureaucratie processes of federal government litigation, and, in sorne cases, to distinguishing between theory and practice.

Sincere thanks to the Department ofPolitical Science's administrative staff, especially

v Helen Wilicka, for help with every conceivable facet of the program. Thanks also to my

dissertation committee, composed of Elisabeth Gidengil, Richard Janda, Christopher

Manfredi, Hudson Meadwell, and Richard Schultz for their dedication and insightful

comments.

Finally, special thanks to my supervisor Chris Manfredi. In addition to his consistently valuable advice on the dissertation, he encouraged my development as an academic in many ways during my time at McGill University. While promoting individu al initiative, he supported my conference work on seve rai occasions (often financially), provided welcome feedback on publication efforts, and facilitated the improvement of my teaching skills. 1 was fortunate to work as Chris's teaching assistant twice, during which time 1 learned a great deal about how to lecture effectively. Later, Chris and the

Department ofPolitical Science gave me the opportunity to teach several classes at

McGill, for which 1 am profoundly grateful. Lastly, he has served as an excellent example of academic professionalism, by combining deeply-held personal beliefs with a commitment to fairness and balanced scholarship. Thanks again for everything.

VI INTRODUCTION

On August 6, 1992, in Haig v. Canada,l the Ontario Court of Appeal declared that

the Canadian Human Rights Act' s failure to prohibit discrimination against gays and

lesbians violated the Charter ofRights and Freedoms. The court took the further,

controversial step of amending the Act by "reading in" such protection, even though the

Charter' s authors had explicitly excluded sexual orientation only a decade earlier. As the

Supreme Court of Canada had not yet addressed whether the Charter's equality rights

extended to sexual orientation, the decision ofOntario's top court represented the highest-

ranking judicial ruling on the issue. Furthermore, the decision could have far-reaching consequences for every federal institution and law regulated by the Canadian Human

Rights Act. Yet, despite the case's manifest importance, the federal government decided not to appeal the decision to the Supreme Court of Canada, effectively conceding the loss.

The Haig incident is difficuIt to reconcile with the view held by sorne interest group litigants, and even sorne Crown attorneys, that governments typicalIy defend their legislation from constitutional chalIenges. 2 For example, in 1993, interest group lawyer

Elizabeth ShiIton criticized the government' s response to Charter litigation, complaining,

[1992] 94 D.L.R. (4th) 1. 2 Elizabeth Shilton (1993) ''Charter Litigation and the Policy Proccsses of Goyermnent: A Public Interest Account," The Impact o/the Charter 0/1 the Public Policy Pro cess, eds. Patrick Monahan and Marie Finkelstein (North York: York University Centre for Public Law and Public Policy); Graeme G. Mitchell (1993) "The Impact of the Charter on the Public Policy Process: The Attorney General." The fmpact o/the Charter 011 the Public Policy Pro cess. 81. "There is Iittle evidence that anybody ever pauses to reflect on the merits of the case or the

politicaljudgment involved in the decision to defend.,,3 However, further evidence reveals

that the federal government is actually a highly selective appellant, appealing only a

fraction of its Charter-related losses to the Supreme Court of Canada.

This dissertation addresses the important question raised by Haig: what factors underlie the federal government' s decision to appeallower court los ses to the Supreme

Court? In so doing, the dissertation explores a neglected dimension of government litigation behaviour under the Charter. Ameriean seholar Donald L. Horowitz observed a quarter-century ago, "[t]he inereasing involvement of the judiciary in bureaucratie work affects the role of government lawyers. As litigation becomes more important, anticipating it and handling it also become more important tasks. The lawyer' s overall role may become more significant.,,4 This is undoubtedly true of government lawyers in

Canada sinee the advent of the Charter and subsequent explosion of rights-based judicial review. Although sorne recent work casts much-needed light on the way government

Iawyers anticipate Charter litigation in the policy drafting pro cess, 5 there has been virtually no systematic empirical analysis of how they "handle" constitutionallitigation.

The dissertation provides novel insights into this important form of government behaviour, not only by examining litigation strategy, but by identifying for the first time the institutional structure and pro cesses by which the Government ofCanada's appeal

3 Shilton (1993), 173. ~ Donald L. Horowitz (l977b) The Jurocracy: Government Lmvyers, Agency Programs, and Judicial Decisions (Lexington, Mass: Lexington Books), 2 [emphasis added]. 5 James B. Kelly (l999a) "Bureaucratie Activism and the Charter of Rights and Freedoms: the Department of Justice and its entry into the centre of goYcrnment," Canadian Public Administration ~2: 476-511; James B. Kelly (1999b) "The Charter of Rights and Freedoms and the Reba1ancing of Liberal Constitutiona1ism in Canada, 1982-1997," Osgoode Hall LmyJournal37: 625-695.

2 decisions are made. In the chapt ers that follow, 1 construct and test a model of

government appeal decision making that takes as its central assumption that such decisions

are made rationally, with government lawyers and their political superiors weighing the

costs and benefits of appealing. The factors 1 examine include the costs associated with a

loss in the lower court, case importance, the likelihood the Supreme Court will agree to hear the case, and the prospect of winning on appeal.

The model is adapted to the context of Canadian Charter litigation from studies of

American government litigation and judicial decision making. These studies, in particularly Christopher Zorn's recent work on the U.S. Solicitor-General's Supreme

Court appeal decisions, conclude that government lawyers are "'procedurally rational' when they decide to interact with the Court.,,6 Much like the U.S. Government, the

Canadian federal government's litigation before the nation's highest court is largely the product of centralized decision making by a professionalised legal bureaucracy. Justice

Department lawyers responsible for appeal decisions in both countries have extensive experience, expertise, resources, and a professed desire to maintain their office's credibility with the court. The dissertation therefore provides an excellent opportunity to test a model oflitigative behaviour developed in the U.S. in a comparative context. In so doing, 1 help address a problem recently identified by the president of the APSA's Law and Courts Section, that hypotheses and theories based on the U.S. context have largely

6 Christopher Zorn (2002) "U.S. Government Litigation Strategies in the Federal Appellate Courts." Political Research Quarter(v 55(1): 1~5-166; Eric N. Waltenburg and Bill Swinford (1999) "The Supreme Court as a Policy Arena: The Strategies and Tactics of State Attorneys GeneraL" Policy Studies Journal 27: 255.

3 been developed without passing through the crucible of comparative application.7

The dissertation also addresses, albeit less directly, a topic which exemplifies the current law and politics literature's neglect of government litigation: the growth of judicial power. A large body of recent work on judicial activism in the "rights revolution," led by Charles Epp's comparative analysis, and by F.L. Morton and Rainer Knopffin

Canada, emphasizes the courts' passive nature, or that judges require other actors to bring cases for adjudication. 8 Scholars typically identifY the crucial role of interest groups in this regard. Epp, for example, concludes that Canada's "rights revolution" is explained primarily by a "support structure for legal mobilization, consisting of rights-advocacy organizations," or what Morton and Knopffterm, more provocatively, the "Court Party." 9

This conclusion ignores that governments are by far the most frequent appellant in rights litigation. By definition, most Charter right claims are directed at government institutions, officiaIs, or legislation, and with unmatched financial and personnel resources, government is, to use Galanter's famous term, the uItimate "repeat player."lO By the same token, the

7 Lee Epstein (1999) "The Comparative Advantage," Law and Courts: Newsletter of the Law and Courts Section ofthe American Po/itical Science Association 9 (3): 1, 3-6; see also Stacia Haynie and C. Neal Tate (2000) "Comparative Judicial Politics in the Year 2000 and the Role of Research Collaboration between Indigenous and American Scholars," presented to the annual meeting of the American Politieal Science Association, Washington, De. 8 Charles R. Epp (1998) The Rights Revolution: Lmvyers, Activists, and Supreme Courts in Comparative Perspective (Chicago: University of Chicago Press); F.L. Morton and Rainer Knopff (2000) The Charter Revolution and the Court Party (Toronto: Broadview Press). 9 Epp (1998), 3. See also Ian Brodie (2001) "Interest Group Litigation and the Embedded State: Canada's Court Challenges Program," Canadian Journal ofPolitical Science 34: 357-376;); Gregory Hein (2001) "Interest Group Litigation and Canadian Democracy," Judicial Power and Canadian Democracy, eds. Paul Howe and Peter H. Russell (Montreal: McGill-Queen's Press); Kent Roach (1994) "The Role ofLitigation and the Charter in Interest Advocacy," Eqllity and COllllllllnity: The Charter, Interest Advocacy and Representation, ed. F. Leslie Seidle (Montreal: Institute for Research on Public Policy); and Miriam Smith (1999) Lesbian and Gay Rights in Canada: Social Movements and Equality-Seeking, 197/-1995 (Toronto: University of Toronto Press). 10 Marc Galanter (1974) "Why the 'Hayes' Come Out Ahead: Speculations on the Limits of Social Change." Law & Socie~v Review 9: 95-160.

4 government has a vested interest in broad, long-termjurisprudential development, which provides an incentive to bring cases to the highest court in the land. Thus, the rise of the judiciary has, ironically, been made possible to a large extent by the very institution that stands to lose authority because ofjudicial activism.

Finally, the dissertation speaks to the dominant contemporary debate in the

Canadian judicial politics field, regarding the appropriateness of the "dialogue" metaphor to describe the judicial-Iegislative relationship under the Charter. 1 expand the terms of this debate in two related directions, first, by providing the only systematic examination of dialogue with lower appeal courts, and second, by developing the contention (first suggested by Manfredi and Kellyll) that appeals to the SCC are a form of governmental expression within the dialogue.

The dissertation's themes are elaborated over the next six chapters. The first chapter reviews the scholarly treatment of government lawyers and litigation to date. To the limited extent the se topics have been addressed in Canada, the focus has been on the

Attorney General' s degree of independence from the political leadership, and more recently, the government's strategy ofinfluencingjudicial interpretation of the constitution through systematic litigation, or "micro-constitutional politics. ,,12 Of more direct relevance to the dissertation is the recent American literature on litigation behaviour, from which 1 develop the model of government appeal decision making. Notably, aIl three

Il Christopher P. Manfredi and James B. Kelly (1999) "Six Degrees of Dialogue: A Response to Hogg and Bushell," Osgoode Hall Law Journal 37: 513-527. 12 Christopher Manfredi (1997) "Institutional Design and the Politics of Constitutional Modification: Understanding Amendment Failure in the United States and Canada," Law & Society Rel'iew 31: 115.

5 bodies of work emphasize that government litigation choices are the product of reasoned

decision making based on a variety of factors.

Chapter Two details the Government of Canada' s appeal decision process. This entails, first, identifying where litigation authority resides within the government; 1 find that it is, primarily, in the Department of Justice Canada. The chapter examines the internaI organization of the Department, and finds that the handling of Supreme Court litigation is fairly decentralized-in stark contrast to the elite cadre oflawyers in the U.S.

Solicitor-General's Office-but the decision whether to appeal is highly centralized, involving top Justice Department bureaucrats and political officiaIs. This centralized institutional configuration facilitates calculated decision making based on broader political and jurisprudential considerations, in addition to case-specifie factors.

Chapters Three through Five constitute the theoretical and empirical core of the dissertation. Chapter Three identifies the institutional-rational choice conceptual framework that informs the dissertation' s main hypotheses. The chapter also elaborates and operationalises the factors that 1 hypothesize influence the government' s appeal decisions, many ofwhich are drawn from the anecdotal and social scientific literature reviewed in Chapter One. The factors reflect a variety of considerations, inc1uding immediate and long-term costs (particularly to government authority), jurisprudential development, and strategie factors su ch as the odds of gaining access to, and winning in, the Supreme Court of Canada.

The dissertation's quantitative analysis component is reported in Chapter Four, and consists of summary data, logistic regression, and statistical simulations. The dissertation employs case data from decisions of the "penuItimate" appeal courts, or those second in

6 rank only to the Supreme Court of Canada. These include the highest courts of appeal in

each province and territory, the Court Martial Appeal Court, and the Federal Court of

Appeal. The chapter establishes the dissertation's case selection criteria, which isolate ail

federal government losses involving a Charter claim in these courts between 1982 and

2000. I first report summary statistics from the data set, including the federal

government' slow appeal rate, despite its relatively few 10sses. 13 Bivariate and

multivariate analysis, consisting of correlations and binary logistic regression using SPSS,

provide statistical support for the calculated decision making model, but not ail individual

factors. Challenges to government authority and the odds of winning represent the

greatest influences, while more legalistic motivations, such as resolving jurisprudential

conflict among lower courts, are not supported by the evidence. The chapter ends by

calculating the effect of selected individual factors on the odds of appealing, by running

simulations where the factor is present or absent.

Chapter Five complements the previous chapter by providing qualitative analysis of two of the statistically strongest factors, namely, novel constitutional interpretations in conjunction with the presence of a third-party intervener, and intrusive judicial remedies such as "reading in." The chapter examines cases containing these elements in greater detail to determine what is really at stake, and with an eye to uncovering further patterns or relationships. Haig emerges as an exceptional case, as it is the on1y concession (non- appeal) wh en the lower court rendered a novel interpretation with an intervener present.

In addition, I examine cases with unexpected outcomes, such as concessions despite

13 As predicted by Galanter's (1974) "party capability theory," the Canadian govemment enjoys a remarkable succcss rate in the penultimate appeal courts.

7 compelling reasons to appeaI (as in Haig) and appeals where the reason for doing so is

initially unclear. The findings reinforce those in Chapter Four, that the federal

government' s appeal decisions appear to be motivated by rational considerations, such as

concerns about winning, minimizing losses, and, relatedly, protecting the discretion of

legislative and administrative actors.

Chapter Six places the dissertation's findings in the context of the contemporary

"dialogue" debate. The chapter first reviews the existing literature on the dialogue

metaphor, and concludes that the debate is primarily over the definition of "dialogue," or

specifically, what constitutes a vaIid governmental response to a judicial ruling.

Disagreement over the prevalence of dialogue is secondary, and flows directly from these

different definitions. 1 further observe that pro-dialogue arguments rest heavily on the potential use of the legislative override (or s.33 "notwithstanding" clause), but offer no

suggestion on how to rehabilitate that provision, which is practically unavailable in politically salient cases due to (misguided) public opposition to its use. The chapter then provides the first systematic analysis of government dialogue with lower appeal courts, incorporating the concept of appeals from these courts as a form of governmental expression, in that an appeal signaIs the government's ability to register disagreement with the ruling. While finding a significant degree of dialogue between the federal government and the lower courts overall, the chaptèr reveals that Ottawa is particularly likely to respond to more intrusive, or activist, judicial re-writing oflegislation. This is consistent with findings from the preceding chapt ers, that the government actively defends its own policy-making authority.

8 The methodology in this dissertation has three components. The first is an

extensive review of the Canadian and comparative literature on decision making in the context oflitigation, and of the relationship between the elected and judicial branches under the Charter. This review situates the dissertation within several scholarly debates, and helps generate the hypotheses tested herein. The second methodological component is the construction of a data set containing aIl Charter decisions by the penultimate courts of appeal involving the federal government as a party between 1982 and 2000. Of central importance to the dissertation is the subset of cases where the federal government lost, and the related data set of federal government appeals to the Supreme Court of Canada.

These data are the empirical basis for the dissertation's conclusions regarding government appeal decision making.

The third and final methodological component is a mail-back survey, conducted in the Spring of2001, ofa sample of70 Department of Justice Canada counsel. The surveys inquired into the conduct of Charter litigation as it proceeds from trial through appeals, the role of counsel in the decision to appeal, and the Department of Justice' s appeal process. FoIlow-up interviews with senior Department counsel and National Litigation

Committee members were conducted in July 2001 and April 2002, focusing on the factors considered in the decision appeal, and the relationship between central Department officiaIs and regional office counsel. The information gathered through surveys and interviews provide much of the basis for the information in Chapter Two.

While the dissertation ultimately seeks to improve understanding of a heretofore neglected dimension of Charter politics-litigation by government-the study is first and foremost an analysis of decision making behaviour. That analysis supports the conclusion

9 that federal government appeal decisions are largely calculated choices, based on institutional self-interest and strategie factors. The dissertation secondarily provides important insights into the pro cess by which cases are brought to the Supreme Court of

Canada, and the relationship between the judicial and executive branches. The government's role in Charter litigation reveals that this relationship is more complex than the existing literature suggests, and points to the need for further study of this dimension of law and politics.

10 CHAPTERONE

From Independence to Litigation Strategy: Scholarly Treatment of Government Lawyers

An impressive body of work has emerged in the last twenty years devoted to understanding the impact of the Charter of Rights and Freedoms on government behaviour. Such work typically deals with the government's capacity to act in light of institutional or behavioural changes precipitated by the adoption of the Charter. For example, reflecting the long-standing centrality of federalism in Canadian politics, several authors consider the Charter' s potential to erode the federal division of power by limiting provincial autonomy, through "centralizing" or "nationalizing" rights-based judicial review.! Others contend that the Charter has undermined executive federalism as a mechanism for constitutional reform, by engendering greater citizen and interest group

Peter H. Russell (1983) "Politicai Purposes of the Canadian Charter of Rights and Freedoms," Canadian Bar Review 61 (1983): 1-33; Rainer Knopffand F.L. Morton (1985) "Nation-Building and the Canadian Charter ofRights and Freedoms," Constitutionalism, Citizenship and Society in Canada, eds. Alan Cairns and Cynthia Williams (Toronto: University of Toronto Press); F.L. Morton (1987) "The Political Impact of the Canadian Charter of Rights and Freedoms," Canadian Journal of Political Science 20: 31-55; F.L. Morton (1995) "The Effect of the Charter of Rights on Canadian FederaIism," Publius: The Journal ofFederalism 25 (Summer): 173-88; Janet Hiebert (1994) "The Charter and Federalism: Revisiting the Nation-Building Thesis," Canada: The State ofthe Federation 1994, eds. Douglas M. Brown and Janet Hiebert (Kingston: McGiIl-Queen's University Press); Janet Hiebert (1996) Limiting Rights: The Dilemma ofJudicial Review (Montreal: McGill-Queen's University Press); Shannon Smithey (1996) "The Effects of the Canadian Supreme Court's Charter Interpretation on Regional and Intergovernmental Tensions in Canada," Publius: The Journal ofFederalism 26: 83-100; Yves de Montingy (1997) "The Impact (Real or Apprehended) of the Canadian Charter of Rights and Freedoms on the Legislative Authority of Quebec," Charting the Consequences: The Impact of Charter Rights on Canadian Law and PoUlies, eds. David Schneiderman and Kate Sutherland (Toronto: University of Toronto Press); James Kelly (2001) "Reconciling Rights and Federalism during Review of the Charter of Rights and Freedoms: The Supreme Court of Canada and the Centralization Thesis, 1982-1999," Canadian Journal of Po/itical Science 34: 321-356.

Il defensiveness of "their" rights.2 The greatest effort, however, has been directed at determining whether and to what extent the Charter has constrained the policy-making capacity of elected governments by expanding substantive judicial review of legislation and government actions. While there is an empirical grounding to this work/ greater emphasis has been placed on normative questions raised by critics of the court from the

Left4 and Right,5 and their detractors, who either support the court's activities6 or contend

2 Patrick Monahan (1991) Meech Lake: The Inside Story (Toronto: University of Toronto Press); Alan C. Cairns (1991) Disruptions: Constitutional Struggles, from the Charter to Meech Lake, ed. Douglas E. Williams (Toronto: McClelland & Stewart); Alan C. Cairns (1 992a) The Charter Versus Federalism: The Dilemmas ofConstitutional Reform (Montreal: McGill-Queen's University Press); Peter H. Russell (1993a) ConstitutionalOdyssey: Can Canadians Become a Sovereign People?, 2"d edition (Toronto: University of Toronto Press); Peter H. Russell (1993b) "The End of Mega Constitution al Politics in Canada?" The Charlottetown Accord, the Referendum, and the Future of Canada, eds. Kenneth McRoberts and Patrick 1. Monahan (Toronto: University of Toronto Press); Errol P. Mendes (1993) "Sinking Again into the Quagmire ofConflicting Visions, Groups, Underinclusion, and Death by Referendum," The Charlottetown Accord, the Referendum, and the Future afCanada; Christopher P. Manfredi and Michael Lusztig (1998) "Why Do Formai Amendments Fail? An Institutional Design Analysis," World PoUtics 50: 377-400. 3 F.L. Morton, Peter H. Russell and Troy Q. Riddell (1994) "The Canadian Charter ofRights and Freedoms: A Descriptive Analysis of the First Decade, 1982-1992," National Journal of Constitutional Law 5: 1-58; Hiebert 1996; James Kelly (1998) "Charter Activism and Canadian Federalism: Rebalancing Liberal Constitutionalism in Canada, 1982-1997" (Ph.D. dissertation, McGill University); James Kelly (1999b) "The Charter of Rights and Freedoms and the Rebalancing of Liberal Constitutionalism in Canada, 1982-1997," Osgoode Hall Law Journal 37: 625-695. 4 Andrew Petter (1986) "The Politics of the Charter," Supreme Court Law Review 8: 473-505; W.A. Bogart (1994) Courts and Country: The Limits ofLitigation and the Social and Political Life of Canada, (Toronto: Oxford University Press); Michael Mandel (1994) The Charter of Rights and the Legalization ofPoUtics in Canada (Toronto: Thompson Educational Publishing); Allan C. Hutchinson (1995) Waitingfor CORAF: A Critique ofLaw and PoUtics (Toronto: University of Toronto Press); Joel Bakan (1997) Just Words: Constitutional Rights and Social Wrongs (Toronto: University of Toronto Press). 5 Rainer Knopffand F.L. Morton (1992) Charter PoUtics (Scarborough: Nelson Canada); Christopher P. Manfredi (1993a) Judicial Power and the Charter: Canada and the Paradox ofLiberal ConstitutionaUsm, Ist edition (Toronto: McClelland & Stewart); Christopher P. Manfredi (2001) Judicial Power and the Charter: Canada and the Paradox of Liberal Constitutionalism, 2nd edition (Toronto: Oxford University Press); F.L. Morton and Rainer Knopff(2000) The Charter Revolution and the Court Party (Toronto: Broadview Press). 6 David Beatty (1987) Putting the Charter to Work: Designing a Constitutional Labour Code (Montreal: McGiIl-Queen's University Press); David Beatty (1990) Talking Heads and the Supremes: The Canadian Production of Constitutional Review (Toronto: Carswell); David Beatty (1995) Constitutional Law in Theory and Practice (Toronto: University of Toronto Press); Lorraine Eisenstat Weinrib (1999) "The Activist Constitution." Policy Options 20: 27-30; Kent Roach (2001) The Supreme Court on Trial: Judicial Activism or Democratic Dialogue (Toronto: Irwin Law).

12 that the limits on legislative/executive power have been exaggerated.7 Most recently, the

issue of legislative capacity has arisen in the debate over the appropriateness of the

"dialogue" metaphor to describe judicial-Iegislative (executive) power relations when

legislation is subject to Charter-based judicial review. 8

Despite this concern for government capacity, relatively little attention has

focused on the role of specifie institutions within government in Charter politics. An

exception is the body of work-much of it anecdotal9-on the role of government

lawyers in the policy-drafting process. Of particular note is the debate over whether the

federal Justice Department has exercised its monopoly over legal advice in the drafting

process to become a central agency. \0 One area that remains virtually unexamined,

7 Patrick Monahan (1987) Politics and the Constitution: The Charter, Federalism and the Supreme Court of Canada (Toronto: Carswell); Richard Sigurdson (1993) "Left- and Right-Wing Charterphobia in Canada: A Critique of the Critics," International Journal ofCanadian Studies 7-8: 96-115; Hiebert 1996; Peter H. Russell (1994) "The Political Purposes of the Charter: Have They Been Fulfilled?" Protecting Rights and Freedoms: Essays on the Charter's Place in Canada's Political, Legal and Intellectual Life, ed. Phillip Bryden et al. (Toronto: University of Toronto Press); Peter W. Hogg and Allison Bushell (1997) "The Charter Dialogue Between Courts and Legislatures (Or Perhaps the Canadian Charter of Rights Isn't Such a Bad Thing After AIl)," Osgoode Hall Law Journal 35: 75- 124; Kelly (1999b). 8 Hogg and Bushell (1997); Christopher P. Manfredi and James Kelly (1999) "Six Degrees of Dialogue: A Response to Hogg and Bushell," Osgoode Hall Law Journal 37: 513-527; Peter W. Hogg and Allison Thornton (1999) "Reply to 'Six Degrees of Dialogue, ", Osgoode Hall Law Journal 37: 529- 536; F.L. Morton (1999) "Dialogue or Monologue?" Policy Options 20: 23-26; Christopher P. Manfredi and James Kelly (2001) "Dialogue, Deference and Restraint: Judicial Independence and Trial Procedures," Saskatchewan Law Review 64: 323-346; Roach (2001); Hiebert (2002) Charter Conflicts: What is Parliament's Role? (Montreal-Kingston: McGill-Queen's University Press). 9 Mary Dawson (1992) "The Impact of the Charter on the Public Policy Process and the Department of Justice," Osgoode Hall Law Journal 30: 595-603; Sheila A. Gallagher (1993) "The Public Policy Process and Charter Litigation," The Impact ofthe Charter on the Public Policy Process, eds. Patrick Monahan and Marie Finkelstein (North York: York University Centre for Public Law and Public Policy); Graeme G. Mitchell (1993) "The Impact of the Charter on the Public Policy Process: The Attorney General," The Impact ofthe Charter on the Public Policy Process; John Tait (1995) "Policy DeveJ.opment and the Charter," Perspectives on Public Policy: 1-24; Julie Jai (1997-98) "Policy, Politics and Law: Changing Relationships in Light of the Charter," National Journal ofConstitutional Law 9: 1-25. 10 Dawson (1992); Monahan and Finkelstein (1993); lan Brodie (1995) "The Charter and the Policy Process," presented at the annual meeting of the Canadian Political Science Association, Montreal; Kelly (1998); James Kelly (1999a) "Bureaucratie Activism and the Charter of Rights and Freedoms:

13 however, is the conduct of Charter litigation by government lawyers.

This chapter reviews the existing literature on government litigation, which in

Canada has focused on one oftwo themes: the Attorney General's independence from politicalleadership in the conduct of litigation, or the government' s strategy of shaping constitutional meaning (and with it, public policy) by influencingjudicial interpretation through "micro-constitutional politics," or systematic litigation by "repeat player" litigants. Il The final section of the chapter reviews analytical work on government lawyers in other countries, which pro vides the theoretical and empirical work most directly relevant to the dissertation. The primary focus here is on the United States, and studies analysing the factors influencing the litigation decisions of state Attorneys

General 12 and the federal Solicitor General. 13 Another body of literature relevant to government litigation, that pertaining to the "dialogue" between the judicial and elected branches of the Canadian state, will be discussed in Chapter Six, which is devoted to that issue.

the Department of Justice and Its Entry Into the Centre of Government," Canadian Public Administration 42: 476-511; Kent Roach (2000) "The Attorney General and the Charter Revisited," University of Toronto Law Review 50: 1-40. Il Marc Galanter (1974) "Why the 'Haves' Come Out Ahead: Speculations on the Limits of Social Change," Law & Society Review 9: 95-160. 12 Eric N. Waltenburg and Bill Swinford (1999) "The Supreme Court as a Policy Arena: The Strategies and Tactics of State Attorneys General," Policy Studies Journal 27: 242-259. 13 William E. Brigman (1966) "The Office of the Solicitor General of the United States" (ph.D. dissertation, University of North Carolina (Chapel Hill)); Donald L. Horowitz (1977) The Jurocracy: Government Lawyers, Agency Programs, and Judicial Decisions (Lexington, Mass: Lexington Books); Peter Irons (1982) The New Deal Lawyers (Princeton: Princeton University Press); Peter Irons (1983) Justice at War (New York: Oxford University Press); Rebecca Mae Salokar (1992) The Solicitor General: The Politics of Law (Philadelphia: Temple University Press); Christopher Zorn (1997) "U.S. Government Litigation Strategies in the Federal Appellate Courts" (Ph.D. dissertation, Ohio State University); Christopher Zorn (2002) "U.S. Government Litigation Strategies in the Federal Appellate Courts," Political Research Quarterly 55(1): 145-166.

14 J. Attorney General Jndependence

There is general agreement thatjudicial review under the Charter of Rights and

Freedoms has enhanced the policy influence of the courts. Similarly, Charter litigation has enhanced the policy influence of the Attorneys General (AGs) who conduct such litigation on behalf oftheir governments. In addition, as diseussed in the next section, government lawyers play a key role in operationalising the Charter, by providing the court with possible interpretations of Charter provisions. By influencing the seope of

Charter rights, AGs help determine how the Charter shapes--and limits--government poliey. In light of the policy consequences of Charter litigation, how much independence should AGs enjoy from their politicalleaders in the conduct of such litigation?

According to the late John Edwards, the foremost scholar of Attorneys General in

Canada (and indeed, the Commonwealth), there is no consensus on the answer to that question: "Let us assume that a eonflict of opinion arises by virtue of the Attorney

General's assessment ofthe legal and policy issues .... There appears to be considerable confusion as to whose interpretation of the law, whose assessment of the policy implications of the legal eonflict should prevail.,,14 The two leading advocates of AG independenee in the conduet of Charter litigation are Edwards and former AG Ontario Ian

Scott, although they rest their arguments on somewhat different grounds.

Scott's emphasis is solely on Charter litigation in criminal cases, where he cites the AGIs historie independence in criminal prosecution and his or her role as "guardian of

14 J.U.J. Edwards (\ 987) "The Attorney General and the Charter of Rights," Charter Litigation, ed. Robert J. Sharpe (Toronto: Butterworths), 51.

15 constitutional values."15 That is, Scott contends that the presence of Charter c1aims does not fundamentally alter the AG's role in criminal cases. He reiterates the three basic justifications for AG independence in the criminal context: "the attorney general must be guided solely by considerations that are independent of his affiliation with a political party or the government" (i.e., non-partisanship)16; "what the law, inc1uding the Charter of Rights, requires in an individual case"; and "what the public interest requires."17 What

"the law requires" is left undefined, but Scott elaborates "guarding the public interest" extensively. He cites the c1assic exposition by British Parliamentarian Lord Shawcross on the AG's responsibility to prosecute in the public interest:

The true doctrine is that it is the dut y of the Attorney General, in deciding whether or not to authorize the prosecution, to acquaint himself with aIl the relevant facts, inc1uding, for instance, the effect which the prosecution, sueeessfui or unsueeessfui as the case may be, would have upon public morale and order, and with any other consideration affeeting public poliey .... The responsibility for the eventual deeision rests with the Attorney General, and he is not to be put ... under pressure by his [sic] colleagues in the matterY

15 lan Scott (1989) "Law, Policy and the Role of the Attorney General: Constancy and Change in the 1980s," University of Toronto Law Journal 39, 109. See also Philip Stenning (1986) Appearingfor the Crown (Cowansville, PQ: Brown Legal Publications); Law Reform Commission of Canada (1990), Controlling Criminal Prosecutions: The Attorney General and the Crown Prosecutor, Law Refonn Working Paper 62, Ottawa: The Commission; Don Stuart (1995) "Prosecutorial Accountability in Canada," Accountability for Criminal Justice: Selected Essays, ed. Philip Stenning (Toronto: University of Toronto Press); Susan Chapman and John Mclnnes (l996)"The Role of the Attorney General in Constitutional Litigation: Re-Defining the Contours of the Public Interest in a Charter Era," The Charter's Impact on the Criminal Justice System, ed. Jamie Cameron (Toronto: Carswell); and Roach (2000). 16 Scott (1989), 119; he cites Edwards, the leading scholar on the AG's office, in identifying that this principle was firmly established in Britain in 1924. Eisewhere, Edwards conclu des that "[h]istorically, there can be no room for doubt that above ail the duties associated with the Attorney General. .. was the exclusive responsibility for making prosecutorial decisions"; J.LI.J. Edwards (1980) Ministerial Responsibility for National Security as if relates ta the Offices ofthe Prime Minister, Attorney General, and Solicitor General ofCanada (Ottawa: Supply and Services Canada), 15. 17 Scott (1989), 115. 18 Scott (1989), 120.

16 Despite Scott' s contention to the contrary, each of the traditional justifications for

prosecutorial independence become problematic when Charter issues are raised. Each is

discussed in tum below.

The Principle ofIndependence !rom Partisan Interests

Scott defines partisan interests as the goveming party's determination to remain in

power, ultimately by maintaining and fostering public support. 19 The concem in cri minai

litigation is with those occasions where public opinion or influential interest groups may

encourage a govemment to pursue unwarranted prosecution. The prosecution (and

persecution) of Jehovah's Witnesses and Communists by the Duplessis govemment in

Quebec are good examples. Exc1udingsuch partisan considerations from prosecutorial

decisions certainly contributes to the rule of law, and, in tum, helps maintain public

confidence in the criminal process. However, Charter litigàtion is more complex, as it

raises fundamental questions about the very nature of the administration of justice, as

weIl as about the role of govemment and the distribution of power between the courts and

the elected government. Public opinion, while not determinative, is relevant to such

fundamental questions. By extension, as the people's elected representative, govemment

has a legitimate right to contribute to debates on these issues, through its legal representative in court.

19 Scott (1989), 121.

17 The Need ta Guard Canstitutianal Values

Scott's conception ofthe AG as the "guardian of constitutional values" (that is, to do "what the law requires") rests on the erroneous assumption that Charter rights are fixed and unequivocal. As former United States Supreme Court Justice William Brennan writes, "the Constitution does not take the form of a litany of specifies. There are, therefore, very few cases where the constitutional answers are clear.,,20 In the Canadian context, Knopff and Morton write, "The wording of the Charter is a tolerably clear expression of the regimes' core principles," but

the kinds of questions most likely to arise for judicial review under the Charter do not concem such basic challenges to the "central core" of liberal democratic principles. The bulk of Charter jurisprudence tums on questions encountered "as we move out from the central core ofthese values" toward the periphery, where the Charter's meaning is "ambiguous and indeterminate."21

Similady, legal scholar Jeremy Webber writes:

Legal texts are never self-executing. They never specify, with complete clarity, all the situations in which they apply. Instead, they are phrased in general terms, describing the general type of situation they govern, with what typical effect. This is especially true of constitutional norms, which are usually highly abstract, affirming broad principles applicable to a wide variety of circumstances.22

The indeterminacy of what Charter rights mean is clearly indicated by the fact that

Supreme Court Justices frequently disagree about the scope ofrights, whatjustifies their

20 William 1. Brennan Jr. (1996) "How the Supreme Court Arrives at Decisions," American Government: Readings and Cases (l2th edition), ed. Peter Woll (NY: HarperCollins), 455. 21 Knopffand Morton (1992),147. 22 Jeremy Webber (1994) Reimagining Canada: Language, Culture,Community and the Canadian Constitution (Montreal: McGill-Queen's University Press), 244.

18 violation under section 1, and what sorts of remedies should follow from unreasonable

violations. 23 Moreover, as noted at the outset, the AG plays a pivotaI role in determining

what the Charter means in practice. Scott's assertion in this regard over-simplifies

Charter litigation.

There is also an institutionally-based objection to this part of Scott's argument.

Nothing within the Charter alters the fundamentally adversarial nature oflitigation. If the

AG fails to provide possible interpretations of Charter rights that would save the

government policy at stake, the court cannot make fully-informed decisions; Justices

Lamer and La Forest said as much in their criticism of the federal government's

concession in Schachter. 24 By declaring policies unconstitutional prior to adjudication-

what Scott calls the "quasi-judicial function"25-AGs in fact usurp the interpretive

function of the judicial branch.

The Need ta Guard the Public Interest

According to Scott, "as a guardian of the public interest, the attorney general must act in accordance with the interests of those whom the government represents, and not simply in the interest of the government to which he belongs."26 The problem lies in that

23 See F.L. Morton, Matthew Hennigar and Shawn Ho (1996) "The Role of Govemments in Charter Litigation: Charter Politics as Centre-Periphery Politics," presented at the annual meeting of the Canadian Political Science Association, St. Catharines, for a discussion of the many interpretive rules the court must address in the course of Charter adjudication. 24 [1992] 2 S.C.R. 679; Lamer wrote, "[the concession] precludes this court from examining the s.15 issue on its merits, whatever doubts might or might not exist about the fin ding below. Further, the appellants' choice not to attempt a justification under s.1 at trial de prives the court of access to the kind of evidence that a s.1 analysis would have brought to light." (689) 25 Scott (1989), 119. 26 Scott (1989), 120.

19 the public interest is inherently a political issue rather than a legal one. As Scott

acknowledges, " [t]he decision to prosecute also entails what must be called a 'political'

dimension with respect to the application of the law and may, in sorne cases, have

important social repercussions that must be reviewed in order to fulfil the attorney

general's role as the protector of the public interest." [emphasis added]27 This is equally

true-if not more so-for the AG's conduct of Charter litigation.

Given this political element, from what source does the AG's authority to

determine the public interest independently derive? It is not from his or her role as the

government's lawyer, but from his or her explicitly political role as a Cabinet minister: at

both the federal and provinciallevels, the AG simultaneously holds the portfolio of

Minister of Justice. Insofar as the AG-Justice Minister is responsible for the criminal

law, independence in criminal prosecutions may be justifiable, but by extension, independence when litigating in non-criminal policy areas (such as Employment

Insurance, govemment pension programs, or language education policy) is not.

However, even as Scott observes, criminal cases may involve complex policy questions outside the boundaries oftraditional criminallaw. He cites as an example the

Morgentaler case on abortion: "Is it about criminallaw? Is it about what a health service is? Is it about access to health services? Is it about what priorities will be established for this health service as opposed to another health service? In truth, the case is about an of these things at one level or another.,,28 In such cases, it is unclear why the AG's

27 Scott (1989), 121. 28 Scott (1989), 110-111.

20 assessment of the public interest should outweigh that of the executive as a whole, or of the ministers responsible for other affected policy areas.

In contrast to Scott, Edwards does not limit his support for AG independence in

Charter litigation to criminal cases. Edwards contends that the British constitutional tradition confers on Canadian AGs both independence and a "larger constitutional duty" to act as guardian of the public interest:

If he views his functions [in Charter litigation] as restricted to that of ensuring that the government is represented by counsel. .. it is my opinion that the Attorney General would be in serious dereliction of his larger constitutional dut y to ensure that the wider public interest is adequately represented. The do or must be left open for the extraordinary demonstration of the Attorney General's independent status ... to argue the case on behalf of the public interest. 29

There are several flaws in Edwards's position. First, the British tradition he cites is in fact a tradition of prosecutorial independence in criminallaw, and so subject to the same criticisms raised against Scott' s argument. 30 Even if one ignores this problem, his contention that the AG should "guard" the public interest is untenable. As noted above,

Charter litigation frequently involves policy issues outside the purview of the Justice

Minister; in non-criminal cases, this is true by definition. More fundamental, however, is the criticism that Edwards's argument is based on the "hotly contested" notion of a single public interest. 31 Roach, echoing Stenning32 contends, "the ide a of the public interest is

29 Edwards (1987), 53. 30 See J.LU. Edwards (1964) The law officers of the Crown: a study ofthe offices ofAttorney-General and Solicitor-General ofEngland with an account ofthe office ofthe Director of Public Prosecutions of England (London: Maxwell & Sweet), and J.LU. Edwards (1984) The Attorney General, Politics and the Public Interest (London: Maxwell & Sweet). 31 Roach (2000), 26-27. 32 Stenning (1986), 304.

21 based on a consensus view of politics. Always suspect, this view has become

increasingly unrealistic in the last two decades as various groups have made their c1aims

and their allegiance to certain parts of the Charter c1ear.,,33 Furthermore, as the public

interest is pluralistic, the AG has no monopoly over either its interpretation or

representation; instead, the public interest can be represented by the public, by relaxing

rules of standing and encouraging third-party interventions in litigation.34 Indeed, as

Huscroft writes, "Developments in the law of standing ... suggest that, although the

Attorney General exercises important powers in the public interest, these powers can no

longer be described as exc1usive."35

A final criticism of Edwards, as weIl as of Scott, relates to their use of the British

tradition to justify AG independence in Charter litigation.36 While the British legal

system provided the blueprint for Canadian legal institutions, these two systems have

diverged dramatically since 1867. In particular, British tradition lacks both entrenched rights and substantive judicial review of such rights-that is, judicially-enforced constitutional supremacy, in contrast to Parliamentary supremacy-and so provides little guidance for the AGIs role in constitutionallitigation. 1 argue elsewhere that a better analogy to Charter litigation cornes from our own tradition of constitutionallitigation in

33 Roach (2000), 27. 34 Roach (2000), 25. 35 Grant Huscroft (1996) "The Attorney General and Charter Challenges to Legislation: Advocate or Adjudicator?" National Journal ofConstitutional Law 5: 129-130. 36 Notably, Peter Hogg maintains that the convention of independence in the prosecution of offenses has only been established in Canada since 1978. See Law Reform Commission of Canada (1990), 8-14.

22 division of powers cases. 37 As Brian Dickson, former Chief Justice of the Supreme Court

of Canada, admonishes,

It has long seemed to me that many commentators are unaware that since 1867 Canadian courts have, in division of powers cases, decided whether Parliament in a particular piece of legislation has infringed provincial authority ... or whether a provinciallegislature has improperly entered the federal domain .... Many cornmentators rush to assert that the Charter marks the advent of a new era without fully exploring the nature of the era that preceded.38

Like Charter rights, the federal division ofpowers-spelled out primarily in sections 91

and 92 of the Constitution Act, 1867-is constitutionally entrenched and part of the

"supreme law" identified by section 52(1) of the Canada Act, 1982. As such, the

Constitution authorizes judicial review on the basis of federalism just as it does for the

Charter, although federalism litigation has been conducted for much longer, beginning

shortly after Confederation. The difference is that, where Charter litigation examines

whether government policies violate the rights of citizens, federalism litigation is

concemed with whether one level of govemment is legislating in the jurisdiction of the

other. However, both types of litigation affect policy areas that are beyond the scope of

the criminallaw and the ministerial jurisdiction of AGs. Federalism litigation has

involved such varied policy areas as natural resources,39 unemployment insurance,40

monetary policy,41 fiscal policy,42 minority language rights,43 constitutional amending

37 Matthew A. Hennigar (1998) "Attorney General- Executive Relations in Charter Litigation," presented at the annual meeting of the Canadian Political Science Association, Ottawa. 38 Brian Dickson (1994) "The Canadian Charter of Rights and Freedoms: dawn of a new era?," Review of Constitutional Sludies 2: 4-5. 39 Natural Products Marketing Act Reference [1937] A.C. 377, C/GOL v. Saskatchewan [1978] 2 S.C.R. 545, et al. 40 Employment and Social /nsurance Act Reference [1937] A.C. 355. 41 Reference re Alberta Statutes [1938] 2 S.C.R. 100. power,44 marketing of agricultural products,45 and regulation of aeronautics46 and

telecommunications.47 As weIl, both Charter and federalism litigation often entail

attempting to influence the interpretation of constitutional text.

My brief examination elsewhere of government litigation in six major federalism

cases, ranging in time from 1938 to 1985,48 reveals a clear precedent of AGs tailoring

their legal arguments in constitutionallitigation to serve the government's political

interests. A good example of AG-Executive co-ordination is provided by a series of

cases dealing with minority language rights, in which the AG Canada made questionable

legal arguments to advance the Trudeau government's constitutional unity strategy of reciprocity between Quebec and "English" Canada in the treatment oftheir respective official language minorities. Constitutionallitigation was crucial to this strategy even before the adoption of the Charter's language rights (sections 16-23). In Blaikie,49 the provisions of Quebec' s Bill 101 requiring the exclusive use of French in official legislative and judicial record-keeping were found ultra vires because, as the AG Canada argued, section 133 of the B.NA. Act, 1867 stipulates that the federal and Quebec

42 Anti-Inflation Act Reference [1976] 2 S.C.R. 373. 43 AG Quebec v. Blaikie [1979] 2 S.C.R. 1016, AG Manitoba v. Forest [1979] 2 S.C.R. 1032, Reference re Manitoba Language Rights [1985] 2 S.C.R. 374. 44 Patriation Reference [1981] 1 S.C.R. 753, Senate Reference [1980] 1 S.C.R. 54, Quebec Veto Reference [1982] 2 S.C.R. 793. 45 Reference re F arm Products Marketing Act (Ontario) [1957] S.C.R. 198, Chicken and Egg Reference [1971] S.C.R. 689. 46 Reference re Regulation and Control ofAeronautics in Canada [1932] A.C. 54. 47 Reference re Regulation and Control ofRadio Communication in Canada [1932] A.C. 304. 48 Reference re Alberta Statutes [1938] 2 S.C.R. 100; Anti-Inflation Act Reference [1976] 2 S.C.R. 373; Patriation Reference [1981] 1 S.C.R. 753; AG Quebec v. Blaikie [1979] 2 S.C.R. 1016; AG Manitoba v. Forest [1979] 2 S.C.R. 1032; and the Reference re Manitoba Language Rights [1985] 2 S.C.R. 374. The focus on reference cases is not arbitrary. The fact that references are initiated by political executives provides a strong guarantee that politicalleaders are aware of and comprehend the issues at stake, and how the law relates to the policy issues. 49 [1979] 2 S.C.R. 10 16.

24 legislatures must keep records in both French and English.50 At roughly the same time, the AG Canada intervened in Forest to support a challenge to Manitoba's English analog of Bill 101, the Official Language Act, 1890. 51 The key issue in Blaikie had been whether section 133 was part of Quebec's "provincial constitution" and therefore unilaterally amendable by the Quebec government. The Court concurred with the AG Canada that it was not, as section 133 also applied to Parliament and was not located in the "provincial powers" section of the BNA Act. In Forest, the AG Canada employed what Mandel accurately characterizes as "sorne half-hearted and completely unconvincing technical arguments" to draw an analogy between section 133 and section 23 of the Manitoba Act,

1870, which legally recognized the province and established bilingualism, but was part of the provincial constitution. 52

The Manitoba Language Rights Reference is something of a footnote to Forest, as the case addressed the legal implications of invalidating the OLA, 1890 for aIl subsequent unilinguallegislation. Ottawa initiated the reference and argued for invalidation of aIl unilinguallegislation after negotiations with the Manitoba government and francophone groups over a translation schedule were scuttled by the province.53 As in Forest, the AG

Canada's legal arguments reflected the government's strategy of linguistic reciprocity by emphasizing the "striking similarity" between section 23 of the Manitoba Act, 1870 and

50 Raynold Langlois, Attorney General of Canada Factum in Blaikie, 23. Supreme Court of Canada File No. 15495. 51 Manitoba's Act, inspired by the declining francophone proportion of the population in the late 19th century, actually banned the use of French altogether in legislative record-keeping, court proceedings and the publication of laws. 52 Mandel (1994), The Charter of Rights and the Legalization of Politics in Canada, Toronto: Thompson Educational Publishing, 139. 53 Ottawa proposed a two-year period of "temporary validity" to allow for translation. The Court accepted the princip le, but extended the period to five years.

25 section 133 of the ENA (now Constitution) Act, 1867. However, the case is notable by itself, as it demonstrates AG-executive co-ordination in constitutionallitigation is not restricted to the pre-Charter era.

Finally, Huscroft provides an alternative argument which implicitly responds to both ScottiEdwards and Hennigar. 54 In contrast to Hennigar, he argues that the AG should be independent from the political executive, but in contrast to ScottiEdwards, only to the extent necessary for the AG to adhere to the will of the legislature. As a member ofParliament, a government Minister, and the Chief Law Officer of the Crown, the AG has a dut y to represent the legislature, which requires defending impugned legislation and prosecuting existing laws. Huscroft criticizes Scott' s argument for its "presumption that the public interest lies only in vindicating the Charter rights of individuals," which

"overlooks the constitutional interest inherent in defense of the legislative process," in particular "the effect of denying the legislative branch a voice in the judicial process."55

Put simply, legislatures should repeal or amend "obviously unconstitutional" laws, not leave it to the AG to orchestrate their judicial nullification.

Huscroft's concern about undermining the legislature is compelling in those cases, like M v. H, where the legislature has c1early rejected the government's attempt to amend the impugned legislation, or where the law was adopted through a free vote. 56

54 Grant Huscroft (1996) "The Attorney General and Charter Challenges to Legislation: Advocate or Adjudicator?", National Journal ofConstitutional Law 5: 125-162. 55 Huscroft (1996),154-155. 56 [1996] 27 O.R. (3d) 593 (Ontario Court (General Division)). Following a free vote in which the Ontario legislature rejected sweeping amendments proposed by the NDP government that would have extended rights and obligations to same-sex couples equivalent to those of opposite-sex couples, then AG Marion Boyd declared that she might refuse to detènd unamended laws. She didjust that when the government tiled its trial factum in M. v. H. Following the NDP's electoral defeat, the new PC government withdrew Boyd's factum and refiled with an argument that defended the impugned law.

26 However, such cases are rare, given the tendency toward majority government, which,

bolstered by party discipline, typically prevents overt dis agreements between the

executive and the legislative majority. More commonly, the question is whether the AG

should defend legislation promulgated by previous governments and legislatures, where

Huscroft's justification for AG fidelity to the legislature is weak. Specifically, he bases his argument that "the Attorney General is also the Legislature's lawyer" on the AG's responsibility for drafting private members' Bills, providing advice to the Legislature and legislative committees about proposed legislation, and her accountability as a member of

Cabinet (and ofParliament) to the House. 57

However, none ofthese roles suggest that the AG should be boundto the will of the Legislature, let alone past Legislatures. The AG's advisory functions c1early do not generate such a dut y, and Stenning provides ample evidence that Ministerial accountability does not either. 58 Stenning, citing Edwards, observes that Ministerial accountability in a parliamentary system means answerability for one's actions after-the- fact, rather than strict adherence to the will of the legislature in the conduct of litigation:

"To be explicit, it is conceived that, after the termination ofthe particular criminal proceedings, the Attorney General...is subject to questioning by members of the House in the same way as any other Minister of the Crown. Like any other Minister they are answerable for their ministerial actions" [emphasis in original]. 59 Even if one ignores this argument, it is unclear from the perspective of democratic the ory why the AG should

57 Huscroft (1996), 128. 58 Stenning (1986), 301-306. 59 Cited in Stenning (1986), 302.

27 place the interests ofpast legislatures over those of the current government (which, by definition in a parliamentary system, is the current legislative majority, barring minority government or a caucus revoIt).

Notwithstanding the foregoing technical points, the core of Huscroft's thesis remains compelling: the government's use of the AG and the courts to amend or repeal legislation undermines the legitimacy of representative institutions. This is particularly troubling given the current view-largely deserved-that legislatures are ineffective because oftheir domination by the executive, and increasingly First Ministers.60

Nonetheless, this is insufficientjustification for binding AGs to the legislature rather than the executive, outside of the rare circumstances noted above. Elections clearly empower governments to pursue their policy goals. The crucial issue, then, is whether elections only authorize the use of parliamentary mechanisms, or whether litigation is a legitimate policy tool for executives: a question whose answer is beyond the scope ofthis piece.

As this dissertation deals with the appeal decisions of government lawyers, it is important to bear in mind the degree of independence from politicalleaders exercised in such decisions. The preceding section reminds us that there are both policy and institutional consequences of the litigation choices taken by government lawyers, and that significant normative concerns about democracy and representing the public interest are raised when lawyers act independently.

60 Donald J. Savoie (1999) Governingji'om the Centre: The Concentration of Power in Canadian Po/itics (Toronto: University of Toronto Press).

28 II. Constitutional Litigation Strategies of Canadian Governments

A. Anecdotal Evidence

There is an unfortunately small amount of anecdotal evidence from federal and provincial Crown counsel about their handling of Charter litigation.61 However, these

accounts support the view hypothesized in this dissertation, that litigation decisions are just that: calculated decisions, rather than fixed procedures. For example, contrary to the allegations of sorne interest group litigants,62 government lawyers contend that their

strategy, when faced with constitutional challenges to legislation, is not automatically defensive, but flexible, involving multiple considerations.63 In particular, government lawyers are prepared to concede that "the challenged law is bad and .. .it is appropriate for the government not to defend but rather to amend or repeal the offending law as quickly as possible.,,64 That said, several authors concur with Mitchell's comment that "if at all possible, the impugned statute will be defended."65 He continues,

First, as counsel for the government, it is not our province to waive the application of a valid provincial or federallaw simply because we might have reservations about it surviving judicial scmtiny. Rather, if the law can reasonably be defended, it is our dut y to argue that it is consistent with the Charter and obtain a legal mIe clarifying its constitutionality. Second, the Attorney General, through his counsel, brings a unique perspective to Charter litigation. Quite naturally, the Applicant represents the interests and concerns of the individual. While sensitive to these concerns, counsel for the Attorney General must represent society's interests before the court.66

61 Jim MacPherson (1986) "The Case for the Defendant Government in Charter Litigation," Litigating the Values ofa Nation: The Canadian Charter of Rights and Freedoms, eds. Joseph Weiler and Robin Elliot (Toronto: Carswell); Scott (1989); Dawson (1992); Gallagher (1993); Mitchell (1993). 62 Shilton (1993). 63 Dawson (1992); MacPherson (1986). 64 MacPherson (1986), 379. 65 Mitchell (1993), 8l. 66 Mitchell (1993),81-82.

29 Similarly, Gallagher argues the Attorney General will be expected to defend a law "if the

Iegislation under atlack is relatively new," and has therefore been subjected to a thorough internaI review on constitutional grounds, before promulgation.67 While governments quite understandably prefer to defend their laws, the anecdotal evidence suggests that such defenses are neither automatic nor universal.

B. Micro-Constitutional Politics

The academic work most explicitly associated with government litigation strategy is that concerning "micro-constitutional politics," or constitutional rule-making through litigation. This term, coined by Manfredi, describes constitutionallitigation where "the general objective ... is to institutionalize specifie policy preferences by manipulating and transforming existing constitutional rules without the constraints imposed by the formaI amendment process."68 Central to the concept of micro-constitutional politics is a distinction between first- and second-order rules. In the case of the Charter, first-order rules are the rights and freedoms themselves, which are neither unambiguous nor self- executing. Second-order rules, created through judicial rulings, govern interpretation and enforcement offirst-order rules. As Manfredi phrases it, "[a]s an institutional design strategy, litigation"-that is, micro-constitutional politics-"manipulates second-order rules in order to modify first-order rules.,,69 In contrast, formaI constitutional amendment, or macro-constitutional politics, seeks to alter first-order rules directly.

67 Gallagher (1993), 92. 68 Christopher Manfredi (1997) "Institutional Design and the Politics of Constitutional Modification: Understanding Amendment Failure in the United States and Canada," Law & Society Review 31: 115. 69 Manfredi (1993b), 4.

30 Micro-constitutional politics often occurs incrementally over several cases, and

therefore usually involves a highly-organized and well-resourced "repeat player" litigant;

for example, rights advocacy organizations.70 LEAF (the Women's Legal Education and

Action Fund), the leading feminist litigation organization in Canada, was founded with

the explicit objective of influencing the second-order mIes sUITounding the Charter's

section 15 equality rights.71 Several studies have documented the micro-constitutional

strategies ofrights advocacy organizations, in areas including equality rights,72 minority

language education,73 and the Charter's section 1 "reasonable limits" clause.74

However, given its frequency before the Court, funding, personnel resources, and

expertise, the ultimate repeat player litigant is the government. Furthermore, "as the

targets of constitutional litigation, govemments are the most obvious systematic litigant

in micro-constitutional politics.,,75 In light ofthis, there has been surprisingly little

sustained analysis of governmental micro-constitutional strategy. Three exceptions are a

70 Galanter (1974). 71 M. Elizabeth Atcheson (1984) Women and Legal Action: precedents, resources and strategies for the jiaure (Ottawa: Canadian Advisory Council on the Status of Women); Mary Eberts (1986) "A Strategy for Equality Litigation Under the Canadian Charter of Rights and Freedoms," Litigating the Values of a Nation, eds. Joseph Weiler and Robin Elliot (Toronto: Carswell). 72 Lori Hausegger and Rainer Knopff (1994) "The Effectiveness ofInterest Group Litigation: An Assessment ofLEAF's Participation in Supreme Court Cases," presented at the annual meeting of the Canadian Political Science Association, Calgary; Lori Hausegger (1999) "The Impact of Interest Groups on Judicial Decision Making: A Comparison ofWomen's Groups in the U.S. and Canada" (Ph.D. Dissertation, Ohio State University). 73 Christopher P. Manfredi (1994b) "Constitutional Rights and Interest Advocacy: Litigating Educational Reform in Canada and the United States," Equity and Community: the Charter, Inferest Advocacy and Representation, ed. F. Leslie Seidle (Montreal: IRPP). 74 Troy Q. Riddell and F.L. Morton (1998) "Reasonable Limitations, Distinct Society and the Canada Clause: Interpretive Clauses and the Competition for Constitutional Advantage," Canadian Journal of Po/itical Science 31: 467-493. . 75 Christopher P. Manfredi (1995) "Constitutional Modification, Institutional Design and the Politics of Gender Equality," presented at the annual meeting of the Canadian Political Science Association, Montreal,7.

31 series of studies by HO,76 Hennigar,77 and Morton, Hennigar and H078 to deterrnine whether the micro-constitutional strategies of provincial and federal governments regarding the Charter are consistent over time, as well as consistent with their respective macro-constitutional strategies in the Patriation (1980-82), Meech Lake Accord (1987-

1990) and Charlottetown Accord (1992) periods.

Ho examines the strategies of Ontario-which macro-constitutionally supported greater judicial power under the Charter-and Alberta and Saskatchewan-which were strongly opposed to greater judicial power-from 1982 to 1992, and finds the strategies of all three to be consistent over time, even across party lines. The micro-constitutional strategy of each province mirrored their macro-constitutional strategy, and was similarly consistent over time. In contrast, Hennigar' s study of the federal government over the same period found far less consistency, at both the macro- and micro-constitutional levels. As is widely known, the Mulroney government's position in the Meech Lake and

Charlottetown Accords departed significantly from Trudeau's constitutional vision of entrenched rights, official bilingualism, and greater judicial power. However, Charter litigation in these periods tells a very different story. Under Trudeau, Ottawa's legal arguments in early non-language rights cases were indistinguishable from the anti-activist positions taken by Alberta and Saskatchewan, an orientation that continues through the

76 Shawn Ho (1994) "The Macro- and Micro-Constitutional Strategies of Provincial Governments in Charter Politics: A Study of Alberta, Saskatchewan and Ontario, 1982-1992" (M.A. Thesis, University of Calgary). 77 Matthew A. Hennigar (1996) "Litigating Pan-Canadianism: The Constitutional Litigation Strategy of the Canadian Federal Government in Charter Cases, 1982-1993" (MA Thesis, University of Calgary). 78 F.L. Morton, Matthew Hennigar and Shawn Ho (1996) "The Role of Governments in Charter Litigation: Charter Politics as Centre-Periphery Politics," presented at the annual meeting of the Canadian Political Science Association, St. Catharines. Mulroneyera. In contrast, and as expected, the Trudeau government's factum in

Protestant Schools (the only language rights case ofthat period) strongly supported judicial activism for language rights. Surprisingly, though, this support continues under

Mulroney, in such cases as Ford (Quebec sign law) and Mahé (Alberta francophone education). There are thus two major disjunctures between Ottawa's micro- and macro- constitutional strategies in this period: "The first is the unanticipated strong support for language rights in Quebec under Mulroney. The second in the strength of the opposition to non-language rights under Trudeau.'.?9

Morton, Hennigar and Ho synthesize the latter two' s earlier works and add

Quebec, and conclude that the micro-constitutional politics of the Charter are largely a continuation of centre-periphery politics, with the Western provinces and Quebec allied in favour ofjudicial restraint, and opposed to pro-activist Ontario. Notably, however, the

"centre" does not usually inc1ude the federal government, except on language rights. The consistency of provincial strategies despite changes in governments further suggests that

"provincial interests have been stronger than partisan interests in shaping governments' approach to rights issues.',gO The authors conc1ude these findings "challenge the orthodox view that Charter poli tics is qualitatively distinct from the traditional Canadian poli tics of regionalism and provincial rights.',SI

79 Hennigar (1996), 131. 80 MOiton, Hennigar and Ho (1996), 23. 81 Morton, Hennigar and Ho (1996), 24.

33 III. Comparative Literature on Government Lawyers

The body of work most directly relevant to the dissertation cornes not from

Canada, but from abroad, specifically the U.S. With the exception of Edwards's highly detailed descriptions of the Attorney General and Director of Public Prosecutions in the u.K.,82 English-Ianguage literature on government lawyers cornes exclusively from the

American context. In addition to a wealth of anecdotal accounts from former office- holders83-primarily Solicitors General, the federal government's top litigator before the u.S. Supreme Court-and descriptions of governmentallegal bureaucracy,84 a relatively small body of work analyses litigation decisions by government lawyers. As Zorn contends, although U.S. scholars have long noted the importance of the Office of the

82 J.LI.J. Edwards (1964) The Law Officers ofthe Crown: a study ofthe offices ofAttorney-General and Solicitor-General ofEngland with an account ofthe office ofthe Director ofPublic Prosecutions of England (London: Maxwell & Sweet); J.LI.J. Edwards (1984) The Attorney General, Politics and the Public Interest (London: Maxwell & Sweet). 83 Archibald Cox (1963) "The Government in the Supreme Court." Chicago Bar Record 44: 221-29; Erwin N. Griswold (1969) "The Office of the Solicitor General-Representing the Interests of the United States Before the Supreme Court." Missouri Law Review 34: 527-537; Robert H. Bork (1973) "The Problems and Pleasures ofBeing Solicitor General," A.B.A. Antitrust Law Journal 42: 701-706; Rex E. Lee (1986) "Lawyering for the Government: Politics, Polemics & Princip le," Ohio State Law Journal 47: 595-601; Charles Fried (1991) arder and Law: Arguing the Reagan Revolution-A Firsthand Account (New York: Simon & Schuster). 84 Donald L. Horowitz (1977b) The Jurocracy: Government Lawyers, Agency Programs, and Judicial Decisions (Lexington, Mass: Lexington Books); James Eisenstein (1978) Counselfor the United States: us. Attorneys in the Political and Legal Systems (Baltimore: Johns Hopkins University Press); John A. Jenkins (1983) "The Solicitor General's Winning Ways," American Bar Association Journal 69: 734-738; Lincoln Capian (1987a) "The Annals of the Law-The Tenth Justice (1)," The New Yorker (10 August): 29-58; Lincoln Capian (1987b) "The Annals of the Law-The Tenth Justice (II)," The New Yorker (17 August): 30-62; Ronald S. Chamberlain (1987) "Mixing Politics and Justice: The Office of the Solicitor General," The Journal of Law and Politics 4: 379-428; Cornell Clayton (1992) The Politics ofJustice: The Attorney General and the Making of Legal Policy (New York: M.E. Sharpe Inc); Rebecca Mae Salokar (1992) The Solicitor General: The Politics of Law (Philadelphia: Temple University Press); Todd Lochner (1993) "The Relationship Between the Office of the Solicitor General and the Independent Agencies: A Reevaluation," Virginia Law Review 79: 549-582; Kirsten A. Norman-Major (1994) "The Solicitor General: Executive Policy Agendas and the Court," Albany Law Review 57: 1081-1109; Comell Clay ton, ed. (1995) Government Lawyers: The Federal Legal Bureaucracy and Presidential Polilics (Lawrence: University Press of Kansas); H.W. Perry, Jr. (1998) "United States Attorneys-Whom Shall They Serve?" Law and Contemporary Problems 61: 129-148.

34 Solicitor General (OSG) as a "cue" to the V.S. Supreme Court's decisions regarding certiorari (leave to appeal), amicus curiae (intervener) applications, and case disposition,85

"few studies have examined the federal government's activities as a litigant; in particular, very little systematic work exists on the decisions of government attorneys regarding appeal strategies."86 As the next chapter will examine the structure of the V.S. federal legal bureaucracy as a comparison to the Canadian, the focus in this section is on the more analytical literature.

Most studies of government litigation in the V.S. rely on surveys of, or interviews with, government lawyers. Examples include lawyers at severallevels, and in both the federal and state governments. Eisenstein is exceptional for his focus on V.S. Attorneys, or federal trial-Ievellitigators, as are Waltenburg and Swinford for their analysis of state

Attorneys Genera1. 87 The overwhelming majority of attention, beginning with Brigman88 and through Horowitz,89 Chamberlain90 and Salokar,91 is given to the OSG and the federal

85 Joseph Tanenhaus, Joseph Schick, Matthew Muraskin, and Daniel Rosen (1963) "The Supreme Court's Certiorari Jurisdiction: Cue Theory," Judicial Decision Making, ed. Glendon Schubert (New York: The Free Press); S. Sidney Ulmer (1984) "The Supreme Court's Certiorari Decisions: Conflict as a Predictive Variable," American Political Science Review 78 (December): 901-911; H.W. Perry, Jr. (1991) Deciding to Decide: Agenda Setting in the United States Supreme Court (Cambridge: Harvard University Press); Gregory A. Caldiera and John R. Wright (1988) "Organized Interest and Agenda Setting in the U.S. Supreme Court," American Political Science Review 82 (December): 1109-1127; Karen O'Connor (1983) "The Amicus Curiae Role of the U.S. Solicitor General in Supreme Court Litigation," Judicature 66: 256-264; Jeffrey A. Segal (1988) "Amicus Curiae Briefs by the Solicitor General During the Warren and Burger Courts," Western Political Quarterly 41: 135-144; Lee Epstein and Jack Knight (1999) "Mapping Out the Strategic Terrain: The Informational Role of Amici Curiae," Supreme Court Decision-Making: New Institutiona/ Approaches, eds. Corne Il W. Clayton and Howard Gillman (Chicago: University of Chicago Press). 86 Zorn (2002), 146 [emphasis in original]. 87 Eisenstein (1978); Eric N. Waltenburg and Bill Swinford (1999) "The Supreme Court as a Policy Arena: The Strategies and Tactics of State Attorneys General" Policy Studies Journa/27: 242-259. 88 William E. Brigman (1966) "The Office of the Solicitor General of the United States" (Ph.D. dissertation, University of North Carolina (Chapel Hill)). 89 Horowitz (1977b). 90 Chamberlain (1987), 392. 91 Salokar (1992).

35 Department of Justice (DOJ). The primary findings of survey research to date can be

summarized as follows.

First, government litigators are, to use Waltenburg and Swinford's phrase,

"'procedurally rational' when they decide to interact with the Court,,;92 that is, the

decision to appeal is a rational one, made by government lawyers weighing the costs and benefits of appealing, and the competing interests ofthose parties directly involved.

Eisenstein's surveys reveal V.S. Attorneys concerned with workload and their relationships with Department of Justice supervisors, the local Federal Circuitjudge, federal administrative agencies, and opposing counse1. 93 When interviewed by

Chamberlain, former Solicitor General Erwin Griswold recalled "such factors as the likelihood that an issue will recur, circuit conflicts, pervasiveness and breadth of the policy issue, the ability of the case to shape the fringes of the law, and the importance of the issue to the administration are an relevant to the Solicitor General when making his decision" whether to appea1. 94 Griswold further contended, however, that the Solicitor

General "will almost always petition for certiorari when an act of Congress is dec1ared unconstitutional.,,95 Based on her interviews of Solicitors General, Salokar cites several factors influencing the decision to appeal: the complexity of the case ("hard cases make bad law"); the degree to which the decision affects government affairs and power; the

92 Waltenburg and Swinford (1999), 255. 93 Eisenstein (1978),132-135. 94 Chamberlain (1987), 392. 95 Chamberlain (1987), 393. Griswold's comment is somewhat surprising, since it is easy to imagine situations where the executive branch has an incentive to use litigation (namely, no! petitioning for certiorari) instead of the Presidential veto, particularly during periods of divided govemment. Perhaps these factors explain Zom's tïnding, discussed below, that the SG does not, in fact, automatically appeal cases when legislation is invalidated.

36 prospect ofwinning on appeal (although a probable loss does not preclude an appeal); an

error by the lower courtjudge; the long-term development of the law; ensuring the law

applies to government in a uniform manner; maintaining the reputation and credibility of

the Solicitor General' office before the Court-which former SG Rex Lee admits is more

96 important than the merits of the recommendation to file certiorari ; and the Court's

workload.97

Second, of the myriad influencing factors, a strong probability ofwinning on

appeal seems to be the most important consideration for American government litigators, particularly the Solicitor General. Brigman concludes, "In the final analysis the decision to seek certiorari is based on subjective estimates of the probabilities ofvictory."98 The logic is that it is "better not to seek review than to lose on the merits.,,99 Similarly,

Waltenburg and Swinford's survey ofU.S. state Attorneys General found "the states are weIl aware that litigation entails risk. Namely, the state interest might lose, and even worse this disadvantaged position is then bolstered by the permanency accompanying a

Supreme Court ruling. As one state respondent explained, 'We will not put good precedent at risk."']OO This echoes Horowitz's earlier conclusions about the federal

Department of Justice and the Solicitor General:

96 Lee (1986), 598. 97 Salokar (1992), 112-113. Notably, these findings largely mirror Rathjen 's survey of private litigants regarding their decision to appeal to the U.S. Court of Appeals. However, as expected, private attorneys are more motivated by the client's ability to pay than government counsel; see Gregory Rathjen (1978) "Lawyers and the Appellate Choice: An Analysis of Factors Affecting the Decision to Appeal," American Politlà Quarter/y 6(4): 387-405. 98 Brigman (1966), 150. 99 Brigman (1966), 68. 100 Waltenburg and Swinford (1999), 154.

37 Not only are they concerned about the possibility of losing a given case, but they must also worry about the potential impact of a loss on aU their other cases. If they lose a given case, they may also lose in the process a rule of law favorable to the governrnent across the board, and, by arguing vigorously for a position the courts regard as untenable, they may simultaneously diminish their credibility with the courts in future cases. 101

Third and finaUy, governrnent lawyers in the U.S. are politicized, in the sense that

they are motivated by the policy agendas oftheir political superiors. Peter lrons's richly

detailed accounts reveal intense co-ordination between the federal administration and its

lawyers with respect to New Deallitigation, and, to a scandalously unethical degree, in

cases concerning the internrnent of Japanese-Americans in World War II. 102 Politicization

occurs even in the OSG, despite its reputation for "independence" and the fact that "the

decision to appeal [to the Supreme Court] is an entirely discretionary one confided in the

final instance to the Solicitor General.,,103 CapIan, arnong others, strongly criticized

Solicitors General Rex Lee and Charles Fried for "politicizing" the office by promoting the Reagan administration' s social policy agenda. 104 However, Salokar rebuts this charge convincingly:

The observation that the office was politicized during the Reagan Administration implies that it was not political in the past. This is simply untrue. The selection process that has been in place since the legislative drafting of the office in 1870 [Presidential nomination with Congressional confirmation] guarantees that solicitors general will act politicaUy when necessary and that their decisions will reflect the policy goals of the administration. The solicitor general will be at least sympathetic to, if not totally supportive of, the president's policy agenda. I05

lOI Horowitz (1977b), 45. 102 Irons (1982); (1983). 103 Horowitz (1977b), 47. 104 Capian (1987a); (1987b). See Norman-Major (1994) for a fuller discussion of the SG's independence and the politicization issue. 105 Salokar(1992),68.

38 Furthermore, as noted earlier, the Solicitor General tends to prote ct legislative capacity

more generally, by routinely challenging judicial nullifications in the lower courts. In

short, government lawyers in the V.S. are neither insulated from nor unconcerned with

the political interests of the administration.

While the conclusions drawn from survey research yield valuable insights into

government litigation, the work from the U.S. that bears most directly on the dissertation

is Zorn's examination ofOSG and DOl appeal decisions, which is unique for its use of previously unavailable internai data and statistical case analysis, rather than interviews with personnel. 106 Vsing a two-year sample (1993-1994) of appeal decisions and recommendations, he investigates three questions: (1) what factors encourage federal government lawyers to appeal?; (2) do the factors influence OSG counsel (who litigate the case before the Supreme Court) differently from DOl counsel (who litigate the case at the lower appeal court and make appeal recommendations to the OSG)-in other words, does the strength of these factors vary according to the decision-maker' s position in the bureaucratie hierarchy?; and (3) what factors influence venue choice by the OSG?J07 1 ignore the third question here as only the first two are relevant in the Canadian context.

Zorn hypothesizes four categories of factors that influence appeal decisions:

(1) Costs: the financial costs oflosing, in terms ofliability in civil cases or fiscal costs

associated with policy changes, and the policy cost of invalidated policies;

106 Christopher Zorn (1997) "U.S. Government Litigation Strategies in the Federal Appellate Courts," (Ph.D. dissertation, Ohio State University); Zorn (2002). 107 Losses in the U.S. court ofappeals may be appealed one oftwo ways: a request for a rehearing en banc (that is, by the entire U.S. court of appeal for that Circuit; U.S. court of appeal cases are heard initially by a three judge panel), or by filing a petition for certiorari in the Supreme Court.

39 (2) Salience: the importance of the case, measured by the presence of an amicus curiae or

constitutional rights daim, or whether the decision was officially reported in print;

(3) Reviewability: the likelihood the court will grant certiorari (that is, will hear the case

at aIl), measured by the presence of intercircuit conflict (disagreements between lower

courts); and

(4) Prospect ofWinning on Appeal: tied to the government's credibility with the Court

and the risk of bad precedent, the chance of winning is improved by a dissent on the

lower court, a reversaI in the lower courts as the case proceeded from trial to appeal

(as this implies the government won at sorne point), and a lower court decision

ideologically at odds with the Supreme Court's ideology.

Because all D.S. Government litigation in the Supreme Court is handled by the

Solicitor General, those individuals litigating the case are usually not those whose policies or direct interests are at stake. As such, Zorn observes, there are multiple participants in the process, each ofwhom are influenced by these four factors. However, the relative weight attached to each factor varies depending on the institutional position of the participant. \08 Counsel for the Solicitor General should be most concerned with maintaining their credibility with the Court (which underlie reviewability and the prospect of winning), case salience and financial costs to the government as a whole, while the chief concern of lower-Ievel counsel in the Department of Justice (who, while the y do not litigate the case, make appeal recommendations) should be the fate of the policy.

lOS Zorn (1997), 90-100.

40 Among his OSG sample, Zorn found that reviewability was the most important

factor, closely followed by the presence of an amicus curiae, and, more weakly, financial

cost. Surprisingly, the prospect ofwinning exercised no influence, by any measure, nor

did the non-amicus measures of salience. As expected, the DOJ was motivated most by

policy invalidations. Although reviewability and financial costs were also statistically

significant, they were less important to the DOJ than the OSG, as hypothesized. Zorn's

methodology and findings are discussed more thoroughly in Chapter Three.

IV. Conclusions

The literature on government litigation, while relatively sparse, yields a number of

important insights. Most basically, government litigation and the decision to appeal have

real policy consequences. For this reason, the degree ofindependence lawyers exercise in

litigation is profoundly important. By the same token, it is crucial to gain a better

understanding of what factors guide litigation behaviour. The literature also reveals that

government lawyers do not mechanically follow established guidelines, but enjoy

considerable discretion in their activities. Moreover, they exercise this discretion

rationally, with careful consideration of a variety of factors, including details of the case

and their relationship with other branches of the government. In the next chapter, l detail the structure of Department of Justice Canada, which is responsible for litigation on

behalf of the Government of Canada, with an eye to determining the precise locus of decision making authority. CHAPTERTWO

The Federal Government's· Appeals Process

The federal government appears frequently in court, and often against its will. As

the focus of this dissertation is on cases where the federal government must decide

whether to appeal to the Supreme Court of Canada following a loss in the lower appellate

courts, 1 am concerned with those cases where the government is a direct party (rather

than a third party intervener). Such cases are ofparticular interest for two additional

reasons. As Zorn argues, first, the government has a direct interest in the outcome of

such cases (in contrast to interventions), and second, "because of the discretionary nature

of these appeals, these cases allow for the most direct examination of the factors which

drive government appeal decision making."l

Cases involving the federal government as a direct party may begin in one of four ways. Most commonly, the government initiates litigation against an individual to enforce a federallaw or regulation. This may entail prosecution of offenses, or simply the application of federal regulations, such as occurs in immigration hearings. Charter issues may arise in these cases when the individual daims his or her (usually legal) rights were violated by the government during the investigatory process. Second, an individual or group may bring a civil action against the government or its officiaIs, for example, government employees suing their employer for wage discrimination on the basis of sex.

Christopher Zorn (1997) "U.S. Govemment Litigation Strategies in the Federal Appellate Courts," (Ph.D. dissertation, Ohio State University), 10-11.

42 Third, the government may bring a civil action against another party; Charter claims

would be rare in this instance, unless they arose via counter-litigation by the individual,

since the government cannot claim Charter rights against citizens. Finally, governments

may appear in court through the reference function, whereby the government requests an

"advisory opinion" from the court, typically on the constitutionality of proposed or

existing legislation.2 In Canada, the federal government may refer questions to the

Supreme Court of Canada, and provincial governments may refer to their respective

highest court of appeal, with governmental parties enjoying a right of appeal to the

Supreme Court. Notably, provinces may refer federallegislation or the legislation of

other provinces, and vice versa. Therefore, the federal government may be in the position

of defending its legislation against a provincial reference in a provincial court of appeal, where if Ottawa is unsuccessful, it must decide whether to appeal to the Supreme Court.

This chapter provides a descriptive overview of the federal government' s appeal decision process, beginning with a brief discussion of the Justice Department Canada's

litigating authority, which has been challenged on severai occasions by the provinces. l then detail the internaI organization of the federal Justice Department, with particular attention to establishing the degree of centralization in the decision making process regarding appeals, as weIl as of the conduct of such appeals. What emerges is a considerably decentralized Department with respect to the handling of day-to-day

2 The reference function is not provided by the constitution, but is a statutory instrument found in the federal Supreme Court Act and provinciallegislation regulating their respective appellate courts. Reference opinions, while neither "true" decisions or binding precedents, have nevertheless been treated as such by the courts, and respected by governments. See Barry Strayer (1988) The Canadian emstitl/tion and the Courts: The Function and Scope ofJudicial Review, 3rd edition (Toronto: Butterworths) for a fuller discussion of constitutional references.

43 litigation and even the conduct of appeals to the Supreme Court, yet profound centralization at the highest levels when determining whether to appeal and what will be argued.

1. Government Litigating Authority in Canada

Aceording to the tirst Department ofJustice Act, passed in 1868, the Attorney

General of Canada (who is also the Minister of Justice) "shall have the regulation and conduet of alllitigation for or against the Crown or any Public Department, in respect of any subjects within the authority or jurisdiction of the Dominion."J However, by the early 1960s, the Royal Commission on Govemment Organization [hereafter Glassco

Commission] found that the federal government' s legal practices bore little resemblance to the 1868 Act: "In spite of efforts at eentralization that dated back to the birth of the

Department of Justice in 1868"-including Prime Minister Sir John A. Macdonald's decision to act as his own Minister of Justice and Attorney General-"it became obvious, nearly a century later, that 'a signiticant amount oflegal work was being done for federal departments without direct involvement or participation of the Department of Justice. ",4

Over time, the various departments and agencies of the federal government had gradually set up their own legal services, such that by the early 1960s, "it was estimated that more

3 J.LU. Edwards (1980) Ministerial Responsibility for National Security as it relates to the Offices of the Prime Minister, Attorney General, and Solicitor General of Canada (Ottawa: Supply and Services Canada), 7. 4 Mélanie Brunet (2000) Out of'the Shadows: The Civil Law Tradition in the Department vfJustice Canada, 1868-2000 (Ottawa: Department of Justice Canada), 67.

44 than 85 percent of the government's lawyers ... did not come under the responsibility of the Department of Justice."s

The Glassco Commission's Report in 1963 recommended the integration of aU federal government legal services under the direction of the Justice Department, which was begun with the 1966 Government Organization Act. By 1970, after extensive negotiations between Justice and other departments and agencies and the transfer of nearly 200 federal government lawyers to the Justice Department, "the revamped

Department more closely reflected the organization mandated by the Department of

Justice Act. ,,6

Today, although the Justice Department's authority over litigation is extensive, it does not hold a true monopoly over the conduct of government litigation. This is reflected in the most recent (1999-2000) Federal Guidebook's description of the AG's responsibilities, as "regulating or conducting litigation for aU departments and agencies."7

Litigation involving some federal agencies (for example, the Canadian Labour Relations

Board), commissions (most notably the Canadian Human Rights Commission), and aU

Crown corporations are currently conducted by non-Justice Department house counsel, or external private counsel. However, the lion's share offederal government litigation involves the Justice Department, other Ministries ("line departments" such as Fisheries and Oceans, or Citizenship and Immigration), or Secretaries of State, aU of which are

5 Brunet (2000),67. 6 Brunet (2000),67. 7 Ferguson. Mary, ed. (1999-2000) Federal Guidebaak: A Guide ta the Canadian Federal Gavernment and its Decisia,!~~akers (Ottawa: J-K Carruthers), 43-1 [emphasis added].

-1-5 represented by Justice Department counse1. 8 For this reason, the dissertation focuses

exclusively on federallitigation involving Justice Department counsel.

A more important exception to the Justice Department's litigating authority

concerns the prosecution of Criminal Code offenses. The Constitution Act, 1867 grants

the federal Parliament exclusive jurisdiction over the criminallaw,9 and beginning in

1882, Parliament has periodically codified criminal offenses. However, the constitution

also grants provincial governments authority over "the Administration of Justice,"10

which historically has included prosecution of criminallaw. As section 2(a) of the

Criminal Code reads, "the' Attorney General,' with respect to proceedings to which this

Act applies, means the Attorney General or Solicitor General of the province in which

those proceedings are taken and includes his lawful deputy." [emphasis added]

Currently, prosecution ofthe Criminal Code is firmly in the hands of provincial AGs,

except in the Territories and, more sporadically, on lndian reservations where it falls to

the AG Canada.

Despite the periodic codification of criminallaw, many federal offenses remain

outside the Code, inc1uding drug possession and trafficking, II illegal fishing,12 tax fraud,

and violations of the Combines and Competition Acts. Under section 2(b) of the Criminal

Code,

proceedings commenced at the instance of the Government of Canada and conducted by or on behalf of that Government in respect of a contravention of, a

8 Only one exception was found in the datas et, Granovsky v. Canada (Minister of Emp/oyment and Immigration) [1998] 3 F.C. 175 (Federal Court of Appeal), where the Ministry was represented by in hou se counsel. 9 Section 91 (27), Constitution Act, 1867. 10 Section 92( 14), Constitution Act, 1867. Il Narcotics Control Act [repealed 1996]; Contl'ol/ed Drugs and Substances Act, 1996. 12 Fisheries Act (Canada).

46 these other possible grounds. In particular, they conc1uded that regulating drugs-even to the point of completely prohibiting them-did not rely on the criminallaw power in section 91(27), but could be justified under the federal residual power (the Court argued that narcotics were not a public policy issue in 1867, but a genuinely "new" issue, analogous to aeronautics or telecommunications). Notably, despite the Court's finding in

Hauser, the Justice Department has agreements with the governments of Quebec and

New Brunswick which permit provincial AGs to prosecute drug-related offenses, except where the RCMP has laid the charges. 17

II. "Canada's Largest Law Firm": The Department of Justice Canada18

A. InternaI Organization

The AG Canada does not, of course, personally conduct the government's litigation. This task falls to the many Justice Department lawyers who are widely dispersed geographically across Canada, and bureaucratically across Ministries

(Departments). Justice Department litigators faU into one ofthree general organization units: Regional Offices; Legal Services Units (LSUs); or headquarters in Ottawa. t9

17 Department of Justice Canada (2000) Federal Prosecution Service Deskbook, 1-2-7; Department of Justice Canada (2001) Serving Canadians: Report ofthe Federal Proseclltion Service Review, 3. 18 Much of the information below was obtained through Justice Department documents, interviews, and mail-back surveys of Justice Department counsel in Spring 2001. Seventy of the counsel of record in the database were contacted, of which 10 agreed to participate, although half of these requested anonymity. 19 Mary Ferguson, ed. (I 999-2000) Federal Guidebook: A Guide to the Canadian Federal Government and ifs Decision-Makers (Ottawa: J-K Carruthers); Government Electronic Directory (2001) http://direct.srv.gc.ca/cgi-bin/wgweng (Ottawa: Public Works and Govemment Services Canada); Department of Justice Canada (2001) Serving Canadians: Report of the Federal Prosecution Service Review; Graham Garton, Department of Justice Canada, personal interview (July 31, 2001, Ottawa). ln addition, the Minister of Justice appoints private members of the bar, or "agents", to conduct trial level federal prosecutions, primarily in narcotics cases. According to a 2001 report, the Federal Prosecution Service (FSP) engaged 233 legal tinns and 763 counsel as agents, compared to only 316

48 Regional Offices

In addition to advocating the integration of federal government legal services in the Justice Department, the Glassco Commission recommended the decentralization of such services within Justice, through the creation of regional offices. The first regional office was established in Montreal in 1965, and offices soon followed in Toronto,

Yellowknife (1966), Vancouver (1967), Winnipeg (1969), Whitehorse and Halifax

(1970), Edmonton (1972) and Saskatoon (1972).20 Today, there are eleven Regional

Offices, five of which are main offices (Halifax, Montreal, Toronto, Edmonton, and

Vancouver) headed by a Senior Regional Director answerable directly to the Deputy

Minister of Justice and Deputy Attorney General of Canada.21 Counsel in the Regional

Offices are by far the most numerous, and handle most litigation at the trial and even first appellate leve1. 22 Litigation within each office is subject to varying amounts of specialization, along such lines as aboriginal affairs, criminal prosecutions, civil litigation, and tax litigation; Figure 2.2 provides a summary of Regional Office sub-units.

Legal Services Units (LSUs)

As noted earlier, the Justice Department is responsible for litigation "in respect of any subjects within the authority or jurisdiction of the Dominion," or, in other words, in policy areas outside the purview ofthe Minister of Justice. Rather than acting as a

in-house counsel (Department of Justice Canada (2001),3). However, "ail counsel, whether in-house or agents, work under the direction of group heads and regional directors. They are in turn responsible to the Assistant Deputy Attorney General (Criminal Law)" (Department of Justice Canada (200\),3). 20 Brunet (2000), 71 (fn. k). 21 The other sub-offices are \ocated in Winnipeg, Calgary, Saskatoon, Yellowknife, Whitehorse and Iqaluit. 22 Graham .G._arton, Department of Justice Canada, personal interview (July 31,200 \, Ottawa).

49 Figure 2.1: Litigation Decision Making Authority, Department of Justice Canada

Minister & Attorney General

Deputy Minister & Deputy AG

r National Litigation Committee

Regional Civil Litigation and Offices Real Property Law, (Quebec) Section

1 Aboriginal Tax Affairs Law Portfolio Services

Business and Public Law & Criminal Regulatory Central Agencies Law Law Portfolio Branch

Citizenship Civil and Litigation Immigration Branch

Legal Services Units

50 Figure 2.2: Department of Justice Canada, Regional Offices

Regional Offices

1 1 1 1

Atlantic Quebec Ontario Prairie/ Arctic BClYukon

Aboriginal Affairs Civil and Tax Law Aboriginal Law Aboriginal Law Abonginal Law Civil Litigation Federal Prosecutions Business Law Cri minai Prosecutions Business Law Criminal Prosecutions Criminal Prosecutions Immigration Crim Prosecutions Tax Litigation Immigration Tax Litigation Immigration Public Law Regulatory La\\ Regulatory Law Tax Litigation Tax Litigation General Counsel Federal Prosecutions

discrete "clearing house" for legal matters originating outside the Justice Department, the

Justice Department attaches lawyers to line departments in Legal Services Units, or

LSUs. LSUs serve as in-house counsel to their client departments, providing legal advice and, less frequently, representing the client before the courts. Kelly argues that the

Justice Department has become a central agency thanks to LSUs. who advise client departments of possible Charter violations during the policy-drafting process (or

"Charter-proofing").23

However, the case for the Justice Department being a central agency is even stronger wh en one considers the Department's neur-monopoly over govemment litigation.

23 [(dly (1998: 1999a).

51 Horowitz, describing a similar situation in the United States federal government, terms the dependent relationship between client departments and the legal bureaucracy-as well as the "lateral relations among legal bureaucrats"-a "laterarchy".24 In Canada, LSUs are sub-units of several different central units (Aboriginal Affairs, Business and Regulatory

Law, Citizenship and Immigration, Public Law and Central Agencies, Tax Law Services), and liaise frequently with parallel Regional Office counsel who handle most litigation.

The Justice Department's web-like structure thus connects LSU counsel to central officiaIs and regional counsel in matching policy areas, while allowing the Department influence laterally across all other Ministries. It must be stressed, of course, that the influence between Justice and other departments in alllikelihood flows in both directions;

LSUs may import Justice's preferences (through policy advice and litigation) into other departments, but LSU counsel also risk "capture" by the client department. As Horowitz observes,

Exactly where a lawyer sits in government may be an important determinant of the accommodation that he reaches between his professional caUing and his bureaucratie employment. The lawyer who serves in a substantive department may develop one orientation toward his work; the lawyer who serves in the Department of Justice may develop another.25

Headquarters

The final group of Justice Department counsel are the relatively smaU number of central officiaIs located in Ottawa, housed primarily in legal-area-specific units, such as

Aboriginal Affairs, Tax Law, Criminal Law, Citizenship and Immigration, and Civil

24 Horowitz (1977b), 2. 25 Horowitz (l977b), 1.

52 Litigation. These are typically headed by an Assistant Deputy Attorney General

(ADAG), who reports directly to the Deputy MinisteriDeputy Attorney General.

According to a senior Department official, "sorne ofthese are just groupings of convenience for senior management purposes, so they can have one pers on overlooking the financial side, the staffing side of it. It doesn't make a lot of sense in terms of the legal subject matters that get grouped together.,,26 Nevertheless, the same official continued, these central units are responsible for trac king the development of the law in their respective policy areas. LSUs and regional counsel are "answerable up to the

ADAG on subject matter, the substantive law rat issue in] litigation," while "the day-today conduct of litigation, in matters of ethics, practice, and so on, would be under the

Senior Regional Director who is now answerable to the Deputy Minister."27

B. Internai Processes for Supreme Court Appeals

In stark contrast to the U.S. federal govemment, where Department of Justice lawyers are replaced by counsel from the Office of the Solicitor General for Supreme

Court appeals (Zorn 1997), the same Justice Department Canada counsel frequently shepherd a case from the lower courts to the Supreme Court. Indeed, the Department's guidelines for appointing Supreme Court counsel explicitly accept "the principle that regional counsel.. .involved in the appeal before the appeal court below should continue to participate in the case before the Supreme Court, whether as lead counsel, co-counsel or junior counsel. ,,28 Of the 47 appeals to the Supreme Court examined herein, the same

26 Graham Garton, Department of Justice Canada, personal interview (July 31,2001, Ottawa). 27 Graham Garton, Department of Justice Canada, personal interview (July 31,2001, Ottawa). 28 Department of Justice Canada (2000), V-23-9.

53 counsel from the Court of Appeal case (or fewer, where multiple counsel) handled the

Supreme Court appeal in 19 cases (40%) and counsel was added in 16 (34%), leaving only 12 cases (26%) where the original counsel was replaced.29 This fact, in conjunction with the dispersion of Justice Department counsel, gives the appearance of a decentralized appeal decision process; however, this appearance is misleading.

While the conduct of the government's Supreme Court litigation is decentralized-although specialization along the lines of the V.S. Solicitor General's

Office is now being considered30-the decision whether to appeal and the authoring of facta are far more centralized. As a senior Justice Department official put it,

[T]he decision making about whether to seek leave is tied up in a lot of bureaucratic process because we don't want lone rangers running offto the Supreme Court with unmeritorious leave applications. When someone gets the approval of the Ministry to takeone up [to the Supreme Court], their written argument. .. has to be processed and it's the same thing with facta on appeal. There's a real pro cess ofreview and commentary, and ... we've sent facta back to be entirely rewritten because they're just not up to what we feel our standard is and what the Court expects ofus. It's something that plays right through the process, from proposaI to seek leave to the end offiling a factum. 31

As the above comment suggests, appeals to the Supreme Court of Canada require formaI Ministry approval, a pro cess with at least three stages: Regional Office recommendation; National Litigation Committee recommendation; and final decision by the Minister of Justice. 32 Since almost alllower-levellitigation is handled by Regional

29 Source: calculated from original data. 30 Graham Garton, Department of Justice Canada, personal interview (July 31, 200 \, Ottawa). 3\ Graham Garton, Department of Justice Canada, personal interview (July 3\, 200 \, Ottawa). 32 Kelly (1998; 1999a) also mentions the Charter Litigation Co-ordinator, a position no longer in existence. In the words of the former holder ofthat position, "My job got created when no one knew what the Charter meant, and because it was going to be high-profile. Over the years we've developed a lot more expertise, and it doesn't make much sense for one person to try to be involved in ail of these cases." (Graham Garton, Department of Justice Canada, personal interview (July 31, 2001, Ottawa)).

54 Office counsel, the first stage in the appeal pro cess is for the Regional Office to make a recommendation to Justice Department Headquarters, in sorne cases via the central official in Ottawa responsible for the policy area (for example, the ADAG (Criminal Law

Branch) for prosecutions of narcotics offenses). In the five main Regional Offices, these recommendations are produced by a Regional Litigation Committee (RLC), which "are a fairly recent phenomenon."33 Prior to the creation ofRLCs, or where RLCs do not exist, recommendations are made by the regional counse1 and the Regional Director. The sheer volume of cases handled by the Regional Offices precludes the National Litigation

Committee and the Minister from reviewing every case. Thus, in practice, the only cases which are referred to Headquarters are those where the Regional Office wishes to appeal a loss, or believes the case is significant enough to warrant review by central officiaIs. In this sense, the decision not to appeal is in many cases a highly decentralized one. That said, a senior Department official and member of the National Litigation Committee recalled that in his experience, "there are usually enough checks and balances that major cases were in sorne fashion brought to someone's attention [at the nationallevel] and someone made a decision about whether it would go on."34 The same official confirrned that there have been occasions, albeit rarely, where the National Litigation Committee and the Minister decided in favour of appealing against the recommendation of the

Regional Office.

ProposaIs to appeal from the Regional Office are next reviewed by the National

Litigation Committee (NLC), which makes recommendations to the Deputy Attorney

33 Robert Frater, Department of Justice Canada, personal interview (April 17,2002, Ottawa). 34 Robert Frater, Department of Justice ~a~ada, personal interview (April 17,2002, Ottawa).

55 General and ultimately to the Minister of Justice. The NLC is composed of eighteen

"core" members (senior litigation managers in issue-specifie portfolios at Headquarters)

and several "invited" members who "have a strong interest in specifie areas covered in

the Committee's agenda," including senior Regional Directors and the counsellitigating

the case. 35 According to Justice Department guidelines, the NLC is responsible for

"reviewing aIl recommendations to appeal or not to appeal significant cases," where

"significant cases" typicaUy involve:

(i) cases likely to attract public attention and require the Attorney General (or other member of the government) to act or make sorne public statement. For instance, every case in the Supreme Court of Canada falls into this category; (ii) cases involving significant constitutional questions; (iii) cases involving federal-provincial or international relations; (iv) cases where it is important to ensure that positions being taken in court on behalf of the Attorney General are consistent across the country. [emphasis added]36

Given that this dissertation's focus is on Charter appeals to the Supreme Court, aU cases

in the database, in theory, qualify as "significant." In practice, however, criminal process

cases, where legal rights are routinely claimed (often without merit) and legislation has

not been chaIlenged, are less significant, and therefore less likely to be reviewed by the

NLC.

The final stage in the appeal decision process rests with the Minister of

Justice/Attorney General of Canada. Ministerial approval for (or rejection of) appeals to the Supreme Court is no mère formality. As a CUITent member of the NLC stated, "the

[National Litigation] Committee always treats its work as a recommendation to the

35 Department of Justice Canada (2000), IX-46-1. 36 Department of Justice Canada (2000), IX-46-4, fn. 4.

56 Deputy and to the Minister. ... There is further review, and the committee is very

cognizant that the Minister has the last word.,,37 Similarly, another member of the NLC

recalled instances where the Committee concluded there was an issue of public

importance, but left the decision entirely with the Minister due to the presence of "other,

more political concems where we have no expertise. ,,38 The Minister or Deputy Minister

also might refer it back to the NLC to consider other factors that may have not been

considered.39

III. Conclusion

The federal govemment' s appellate process is marked by a number of contradictions. The first is that although the Department of Justice Canada largely monopolizes the govemment' s litigation, the Department itself is intemally decentralized, across both geographic regions and other Ministries. The second is that the process for appealing adverse rulings to the Supreme Court in Charter cases can be described, in general, as highly centralized; this is definitely so where the decision is made to proceed with an appeal. However, the participation of central officiaIs is less certain when an appeal is not pursued, with Regional Offices acting as a potential veto point. The third and final contradiction-or perhaps more accurately, source oftension-is that despite a centralized procedure for approving appeals and facta, the Department remains committed to collaboration with Regional Offices, LSUs, and initial counsel in the

37 Robert Frater, Department of Justice Canada. persona! interview (Apri! 17, 2002, Ottawa). 38 Graham Garton, Department of Justice Canada, persona! interview (July 31, 2001, Ottawa). 39 Robert Frater, Department of Justice Canada, personal interview (April 17,2002, Ottawa).

57 appeals process. This is evident in the inclusion of Regional Directors and representatives of affected Departments on the NLC, as weIl as in the retenti on of original counsel for Supreme Court appeals. Thus, although the appeal process is sequential in nature, passing from Regional counsel and Offices through the NLC to the Minister, it is also collegial, particularly at the intermediate stage.

With the federal government' s appeals process now established, the following chapter details the dissertation's analytical framework, core hypotheses and methodology.

As we will see, the complex and somewhat contradictory organization of the Justice

Department imposes sorne important limitations on the study' s scope and possible methods of inquiry.

58 CHAPTER THREE

Modelling the Decision to Appeal: Factors and Hypotheses

The dissertation takes as its central assumption that the decision to appeal is a rational one, made by government lawyers weighing the costs and benefits of appealing, and the competing interests ofthose parties directly involved. This is consistent with findings from the V.S., discussed in Chapter One, that governments are "'procedurally rational' when they decide to interact with the Court.'" As weU, Justice Department officiaIs and documents speak of considering a variety of factors in their decision making, including the likelihood the Supreme Court will grant leave, maintaining the government's reputation with the Court, and resolvingjurisprudential conflict in the lower courts.2

The broader conceptual approach that informs the dissertation is grounded in the neo-institutional rational choice theory literature. Institutions, by this account, provide incentives and disincentives to sorne types of political behaviour, forming a context within which rational agents act strategically to achieve their goals as fully as possible.3

Institutional features influence the government lawyer's rational calculus, as limited

Waltenburg and Swinford (1999),255. 2 Dawson (1992); MacPherson (1986); Department of Justice Canada (2000), V-22-3-6; Graham Garton, Department of Justice Canada, personal interview (July 31, 2001, Ottawa); Robert Frater, Department of Justice Canada, personal interview (April 17, 2002, Ottawa). 3 George Tsebelis (1990) Nested Games: Rational Choice in Comparative Politics (Berkeley: University ofCalifomia Press); Peter Hall and Rosemary Taylor (1996) "Political Science and the Three New Institutionalisms: Review Article," Political Studies 44: 936-957; Ellen Immergut (1998) "The Theoretical Core of the New Institutionalism," Politics & Society 26: 5-34.

59 institutional capacity necessitates a selective appeal strategy; as the Justice Department

concedes, "not every unfavourable ruling or error in law should be appealed. Neither the

courts nor the Crown has the resources to review every judgment that appears to be

wrong.,,4 This is certainly the case for the U.S. Solicitor General, whose office of only 18

lawyers faces thousands ofpotential appeals per year. 5 The AG Canada is less

constrained in this regard, as Regional Office and (rarely) LSU counsel are permitted to

handle litigation, including before the Supreme Court of Canada. Nevertheless, the

federal government is party to large amounts of litigation per year and the fact of

decentralization may actually contribute to overload, at the level of individual

government lawyers. For example, a Regional Office lawyer is responsible for handling

a large volume oftrial-Ievellitigation, as weIl as appeals to multiple appeal courts.

Under these conditions, it is likely that individual government lawyers may need to limit

their commitments to the appeals process.

A second and more important institutional constraint on the behaviour of

government litigators is the Supreme Court of Canada. The Court hears only

approximately 100 cases per year, and, with the few exceptions noted later this chapter,

enjoys complete docket control. Justice Department lawyers must, therefore, be highly

selective about which cases ta appeal to the Court, not only to gain access to the Court,

but to maximize the odds of winning once there.

A clarification before proceeding: one possible criticism of this approach is that

treating appeal decisions as the product of rational goal-maximization ignores the

4 Department of Justice Canada (2000), V-22-3 [emphasis added]. 5 Zorn (1997; 2001).

60 possibility that decision makers may be guided by princip1e, rather than a "weighing of factors." There are two possible responses to this criticism. The first is that the distinction between acting on princip1e and on goa1s--or, in technica1 par1ance,

"preference"-is specious. This is not to belitt1e princip1e or to equate it with acting on a whim; rather, to say one is driven by a sense of right and wrong is indistinguishab1e, for ana1ytic purposes, from saying one is motivated by a preference for a particu1ar course of action or pattern ofbehaviour.

While there is sorne merit to this view, it is not the position l take. It is poor po1itica1 science to reject out ofhand the motivation ofpo1itica1 actors; simp1y "b1ack- boxing" the content of actor preferences falls into what Green and Shapiro label a

"pathology" of rational choice theory.6 Rather, following the second possible response, l point to existing research and the fact that the 1awyers themse1ves admit to this "weighing of factors"; by the same token, however, l must take them at their word when they contend sorne decisions were made-for examp1e, accepting a 10ss in the 10wer courts- because "[y]ou don't continue beating away at something if you fee1 the judgment was probably right.,,7

U1timately, l approach the possibi1ity ofprincip1e-driven behaviour as an empirica1 issue: "doing what we felt was right" cannot be quantified or otherwise measured given the information avai1ab1e. Conversely, many non-principled explanatory factors can be measured, which provide a methodological starting point for systematically

6 Donald Green and Ian Shapiro (1994) Pathologies of Rational Choice Theory: A Critique of Applications in Po/itical Science (New Haven: Yale University Press). 7 Graham Garton, Department of Justice Canada, personal interview (July 31,200 1, Ottawa).

61 testing hypotheses. ·Ifthese hypotheses are strongly confirrned, principles are, therefore, a poor explanation for most decisions; if many appeal decisions remain unexplained, then principled behaviour may weIl (but, it must be stressed, not necessarily) provide an explanation. It is, to repeat, an empirical question. The first step in answering that question is to establish what factors are likely to motivate government actors when they make appeal decisions.

I. Factors in Appeal Decision Making

Zorn's analysis of the V.S. federal government's decisions to appeal to the

Supreme Court hypothesizes four categories of explanatory factors: "matters relating to the cost of the lower court's decision, the salience [or importance] ofthat outcome, the likelihood that the case will be accepted for appellate review, and the likelihood that, conditional on that acceptance, the reviewing court will mIe in favour of the government."g AIthough Zorn's factors influence my thinking about probable factors in the Canadian context, l operationalize these factors differently. As Zorn suggests, these categories do not exhaust possible motivations, nor are they watertight compartments.

Rather, they should be considered as "heuristic categories" which help organize the

"diverse set of elements which enter into the appeal calculus."9 Each category is discussed in general terrns below, and operationalized in the subsequent section.

8 Zorn (1997), 73. 9 Zorn (1997), 75.

62 A. Costs

Alliosses in litigation entail sorne cost to the losing party, and most litigants (or

other political actors, for that matter) seek to minimize loss. For private parties in civil

litigation, this may be financialliability, or incarceration in criminal cases. While

governments which lose in court cannot be imprisoned, they can face a number of costs,

which can be sorted into two categories: financial and policy costs. Like private litigants,

governments may face financial costs if found liable in civil suits, as weil as the resources

required to prepare and present the case. In addition, however, the state may bear fiscal

costs associated with policy change; for example, a decision extending state-funded

"maternity" leave benefits to fathers would, left unaltered, significantly increase the

government' s expenditures in that pro gram. ID

Similarly, policy costs are more specific to governmentallitigants. Policy costs include the loss of a particular statute, but also lost power or authority in the given policy area. A court decision invalidating, for example, government legislation requiring a common day ofrest (such as a "Sunday closing law") because it violates religious freedom effectively bars the government from imposing any truly universal day of rest.

The government in this example has lost a statute, as weil as sorne of its ability to regulate business and labour. A related policy cost is the opportunity cost and resources

(in time, personnel, and political capital) associated with attempting to draft replacement legislation.

10 R. v. Schachter [1990] 2 F.C. 129; [1992] 2 S.C.R. 679.

63 B. Salien ce

In the simplest terms, salience refers to the importance of the case. Costs are obviously part ofthis calculation, but where the previous category dealt with the immediate financial and policy costs of a loss, salience includes the broader consequences of the court's ruling. In other words, what is the scope ofthe decision's impact? Following Galanter's seminal distinction between "one-shot" and "repeat player" litigants, Il Zorn contends that "whereas a one-time litigant may care very little about the impact of his or her case on the law over time, the govemment' s pervasive presence in the ... courts means that cases which effect fundamental shifts in the le gal

'playing field' will be of special interest to it."l2

There are two dimensions to scope that may provoke an appeal. First, does the court's decision create a rule which will apply to a broad spectrum of govemment departments and policies? For example, the Justice Department instructs its prosecutors to consider, "Is the issue raised by the case ofwidespread importance for the effective enforcement of the criminallaw, or is its impact confined largely to the immediate case?"l3 The second dimension of sc ope pertains to the ruling's impact on the government's power relationships to other state institutions. As Salokar phrases it in her study of the U.S. Solicitor General, "Has the executive branch, as a result of the lower- court decision, been weakened vis-à-vis the power of other branches, and to what degree has this occurred? If the holding results in a significant decrease in executive power ... the

Il Marc Galanter (1974) "Why the 'Haves' Come Out Ahead: Speculations on the Limits of Social Change," Law & Society Review 9: 95-160. 12 Zorn (1997), 80. 13 Department of Justice Canada (2000), V-22-·t

64 case will receive serious consideration for submission [to appeal]."14 A good example is

R. v. Schachter,15 where the lower appeal court ruled that the Charter's remedy clause

(s.24(1» empowersjudicial "reading in" to address rights violations. Zorn correctly observes that

the govemment's status as a 'classic repeat player' (Galanter 1974) suggests that cases which have the potential to significantly change existing precedents, and thus alter the long-term landscape of litigation in an area of law in which the government is involved, will be of special relevance to the govemment. 16

Other factors unrelated to costs can enhance the importance of a case. Zorn identifies two such factors: (i) the extent to which the decision harms the interests of a department' s clients; and (ii) the policy significance of the case. As discussed in Chapter

One, when the Justice Department goes to court, its "client" is often another line department or govemment agency, but these bureaucracies-as weIl as Justice-have their own clientele in turn. Horowitz's study of the V.S. federal govemment's litigation found that "If there are political constituencies with a strong interest in the pro gram under attack, the defense may be pushed on even in the face ofimpendingjudicial disaster.,,17

Canada's Senior Counsel for Civil Litigation echoes this comment, admitting that in tax law, where "we've been losing with relative consistency for at least a decade, we'd be less inclined to support a leave application there unless it was truly necessary or demanded by the client.,,18 There are certainly opportunities for the Justice Department to

14 Salokar (1992), Ill. 15 R. v. Schachter [1990] 2 F.C. 129; [1992] 2 S.C.R. 679. 16 Zorn (2002), 150. 17 Horowitz (l977b), 84. IS Graham Garton, Department of Justice Canada, personal interview (July 31, 2001, Ottawa) [emphasis added].

65 hear the interests of these political constituencies. Justice Department personnel surveyed indicated that litigation involves extensive consultation with client departments and their

LSUs-recall that they are represented on the National Litigation Committee-and

Department guidelines require consultation with "responsibility centres" within Justice, such as the Aboriginal Issues Committee, the Ruman Rights Law Section, the Criminal

Law Policy Section, and the Diversity and Gender Equality Office. 19 Indeed, the

Department identifies "situations where consultation is particularly important-for example, aboriginallitigation, equality issues, appeals ta the Supreme Court of Canada, environmental prosecutions, and constitutional interventions.,,2o

The final dimension of salience is whether the case in question is particularly important in the context of the day. While the entire category of"salience" refers to a case's importance, the idea here is that sorne cases are simply more important than others, because of the policy area involved, the type of law, public opinion, media attention, and so forth. These factors can of course change over time, and the Department of Justice willlikely change its behaviour accordingly. For example, since securing virtually complete docket control in 1975 (and especially since the adoption of the Charter in

1982), the Supreme Court of Canada has allocated a dramatically higher proportion of its time for constitutionallaw cases, and correspondingly less on traditional private law disputes. Mirroring this trend, the Justice Department has shifted its focus toward constitutionallaw. As one senior official recounted, "1 found certainly throughout my 28 years that more of the work we do now is really important stuff, with the introduction of

19 Department of Justice Canada (2000), Chapter 25; Chapter 45 [emphasis added]. 20 Department of Justice Canada (2000), IX-45-2.

66 the Charter. Constitutional cases used to be a very small number, but now ... instead of

doing smaIl c1aims court trials or personal injury c1aims ... we're doing public law

cases.,,21

C. Reviewability

The Department of Justice's shift toward constitutionallaw issues suggests a third

factor influencing appeal decisions: whether the Supreme Court is Iikely to hear the case at aIl, or the case's "reviewability." As noted, the Court has enjoyed virtuaIly full docket control since 1975, with almost aIl appeals to the Supreme Court requiring the Court's permission, or "leave to appeal." There are three major exceptions, when AGs may appeal "as ofright": from ajudgment of the Federal Court of Appeal in the case ofa controversy between Canada and a province;22 from a provincial court of appeal

"opinion" in reference cases23 ; and in criminallaw cases involving indictable offenses/4 when the Court of Appeal dismisses an appeal by the AG or sets aside a conviction, and there is dissent on the Court of Appeal on any pure question of law.25

In aIl other cases, the appeal process is govemed by Section 40(1) of the Supreme

Court Act, which Flemming and Krutz characterize as "elastically worded and vaguely

21 Graham Garton, Department of Justice Canada, personal interview (July 31, 2001, Ottawa). 22 Section 35.1, Supreme Court Act, R.S.C. 1985, c. S-26. 23 Section 36, Supreme Court Act, R.S.C. 1985, c. 8-26. 24 Indictable offenses in Canada are typically more serious offenses, with a more extensive trial process; they are analogous to felonies in the V.S. Less-serious offenses are 'summary' offenses, analogous to misdemeanors. The federal Parliament can create either, but provinciallegislatures may only create summary offenses. 25 Section 693 (1)(a), Crimina/ Code, R.S.C. 1985, c. C-46.

67 defined" to maximize judicial discretion.26 Section 40(1) instructs the Court to grant

leave where

the Supreme Court is of the opinion that any question involved therein is, by reason of its public importance or the importance of any issue of law or any issue of mixed law and fact involved in such question, one that ought to be decided by the Supreme Court or is, for any other reason, of such a nature or significance as to warrant decision by it.

Similarly, Crane and Brown conclude, "It is not enough to show that a legal error may

have been made by the court of appeal, the question is why is the case of national or

public importance.,,27 However, the Court has steadfastly resisted clarifying what

constitutes "public importance" in its jurisprudence, and reasons for granting or denying

leave are usually not given. Nor have individual Justices provided much direction in their

off-bench remarks. For example, Justice John Sopinka, in a 1997 speech intended to

address the issue of public importance, only indicated that the Court would usually not

hear cases involving isolated noncompliance by lower courts, issues recently addressed

by the Court, or an issue recently or about to be dealt with by legislation.28

Notwithstanding this ambiguity, Justice Department documents and personnel

emphasize considering the odds of securing leave when making appeal decisions. For

example, one senior official suggested this factor outweighed concems about the financial

cost of an adverse lower court ruling:

26 Supreme Court Act, R.S.C. 1985, c. S-26; Roy B. Flemming and Glen S. Krutz (2002) "Selecting Appeals for ludicial Review in Canada: a Replication and Multivariate Test of American Hypotheses," Journal of Po/itics 64, 233. Flemming and Krutz also provide a fuller description of the Supreme Court's leave to appeal process. 27 Brian A. Crane and Henry S. Brown (1999) Supreme Court of Canada Practice 2000 (Toronto: Carswell), 26. 28 Cited in Crane and Brown (1999), 26-27; however, the Court has heard sorne cases rendered moot by legislative amendments, such as R. v. Schachter [1992] 2 S.C.R. 679.

68 Occasionally, there are cases-such as tax cases, for example-where ... the Court of Appeal' s judgment means millions of dollars in 10st revenue, but we say, "but where's your legal issue ofpublic importance? ... We're not going to get over the leave threshold." ... .1 think there's a sensitivity in the Department that "public importance" me ans something substantial, so you better have a substantial proposaI to make.29

While the simple fact of the Court's docket control is reason enough for

government 1itigators to worry about obtaining leave when deciding whether to appeal,

the Justice Department is also concerned with maintaining its professional reputation with

the Supreme Court. When queried directly about whether institutional credibility

concerns Justice Department decision-makers, one senior official replied,

Definitely. . .. [I]n civil, tax, immigration, aboriginal [cases] we get leave in about 15 cases a year, so we don't want to wear out our welcome. They [the Supreme Court] grant leave on average in about 100 cases a year. . .. Something in the order of 15 to 20 per cent is going to be our share, and if we're there 40 and 50 per cent of the time, l think we're going to create an impression in their minds that we're not looking at these things critically.30

Another senior official concurred:

We do not want to be seen as bringing frivolous leave applications, so ... we're employing the same test that they employ and we think it' s the role of the Attorney General to sort the wheat from the chaff, and only take cases where we legitimately think we have a national issue that they ought to decide. They don't always agree with us, but we pride ourselves on taking a rigorous look at these cases before launching a leave application.31

To summarize, challenging an adverse ruling from the lower appellate courts first requires gaining access to the Supreme Court of Canada, access which is tightly

29 Graham Garton, Department of Justice Canada, personal interview (July 31, 2001, Ottawa). 30 Graham Garton, Department of Justice Canada, personal interview (July 31,200 l, Ottawa). 31 Robert Frater, Department of Justice Canada, personal interview (April 17,2002, Ottawa).

69 controlled by the judges ofthat Court. As a matter of strategy, then, government litigators must be able to identify factors in a case-most importantly, its "national" or

"public importance"-which render it a compelling case for Supreme Court Justices to hear.

D. The Prospect of Winning on Appeal

Zorn considers the positive prospect ofwinning on appeal (or, as l term it,

"winnability") as a necessary condition for proceeding with an appeal, on the grounds that governrnents will not waste resources or risk their credibility with the Supreme Court litigating "lost causes.,,32 Winnability is also important for jurisprudential reasons, namely, that losing before the country's highest court of appeal may establish an unfavourable legal rule with the widest possible application. As Waltenburg and

Swinford found (as reported in Chapter One), U.S. state govemment lawyers were unwilling to "put good precedent at risk. ,,33 The risks of appealing are even greater for the Canadian federal governrnent, in that their losses are limited when they occur in provincial courts of appeal, while a loss before the Supreme Court has consequences for the application of federallaw at the nationallevel. This is because rulings by provincial appeal courts (as opposed to the Federal Court of Appeal) apply only in the jurisdiction of that court (for example, Ontario for the Ontario Court of Appeal), but the Supreme

Court' s decisions apply in aH jurisdictions. In view of this risk, governments have an extra incentive to avoid losing in the Supreme Court, although this factor should be less

32 Zorn (1997),85. 33 Waltenburg and Swinford (1999), 254.

70 prevalent in Federal Court of Appeal cases.

While these arguments suggest that winnability influences the decision to appeal, it is incorrect to conclude that it is a necessary factor, as this assumes a govermnent can never benefit from proceeding "against the odds." There is ample evidence that interest groups litigate in the face of almost certain failure. For example, Smith observes that

"[e ]quality-seeking, including litigation, was undertaken in the [Canadian lesbian and gay liberation] movement in the seventies, despite the fact that the chances of success were dismal.,,34 As she argues, "[r]ights claims function as claims to politicallegitimacy, as a means of politicizing grievances, and thus as mechanisms for the creation of a sense of political identity," regardless ofwhether such claims are successful; indeed, a rejection by the court may be even more effective.35 Similarly, govermnents may reap political gains with clientele or partisan supporters by appearing to "go down fighting." For example, the Parti Québécois appealed the invalidation of its mandatory French education regime

("Bill 101"),36 even though there was virtually no chance the federally-appointed

Supreme Court would reverse a nullification under the new Charter's language rights

(which had been drafted specifically to counter Bill 101 !). Govermnents and their lawyers operate within "nested games,'m in which the apparently irrational decision to squander resources and credibility litigating a lost cause becomes rational in light of the larger political context. Therefore, although winnability is an important factor, it cannot, by itself, prevent (or provoke) an appeal.

34 Smith (1999), 42. 35 Smith (1999), 41. 36 Attorney General ofQuebec v. Association ofQuebec Protestant School Boards [1984] 2 S.C.R. 66. 37 Tsebelis (1990).

71 II. Operationalizing the Factors

Having identified the general types of factors which likely motivate government lawyers, it is necessary to establish how to measure these factors empirically, in view of the constraints imposed by the available data. Each factor is operationalized in turn and the core hypotheses to be tested summarized below.

A. Costs

1 largely concur with Zorn's treatment of the costs associated with the lower court' s ruling. The actual financial costs of an adverse ruling usually defy measurement, as expenditures associated with policy changes may be influenced by a wide variety of factors. The exception is those civilliability cases-rare, in Charter litigation-where the court assigns financial damages against the government. In any case, it is unclear where to draw the line, in terms of dollars, between "significant" and "insignificant" financial costs. However, Zorn provides a simple yet compelling solution: following Galanter's conclusion that civillitigation is typically more costly than criminal prosecution,38 Zorn hypothesizes that the former should therefore be appealed more frequently. Furthermore, central decision-makers should be less concerned with criminal case losses "because the costs of a court of appeals reversaI in a criminal case is [sic] borne by someone other than the individual(s) responsible for making the final appeal decision."39 Accordingly, 1 code cases "0" where the federal government initiated prosecution of an offense and "1" if otherwise.

38 Galanter (1974). 39 Zorn (1997),109.

72 As to the policy authority dimension of costs, Zorn again uses a simple measure: whether the court invalidated a federa11aw or order. Based on the assumption that governments wish to proteet existing policies, this is a logical measure, and is consistent with aneedotal accounts from government lawyers,reported in Chapter One, that "if at aU possible, the impugned statute will be defended."40 However, Zorn's focus on invalidation is incomplete; a more accurate measure would inc1ude aU forms of constitutional remedy: invalidation (or "nuUification"), constitutiona1 exemption, severance, andjudicial extension. Section 52(1) of the Canada Act, 1982, which formaUy establishes constitutional supremacy in Canada, expressly authorizes only nullification:

52. (1) The Constitution of Canada is the supreme law of Canada, and any law that is inconsistent with the provisions of the Constitution is, to the extent of the ineonsistency, of no force or effect.

However, the Charter provides greater judicial discretion to remedy violations of its

4 rights, as explained in the s.24(1) enforcement c1ause: !

24. (l) Anyone whose rights or freedoms, as guaranteed in this Charter, have been infringed or denied may app1y to a court of competent jurisdiction to obtain such remedy as the court eonsiders appropriate and just in the circumstances.

40 Mitchell (1993), 81. 41 Another possible remedy, for violations of the Charter's legal rights only, is provided by section 24(2): "Where, in proceedings under subsection (1), a court concludes that evidence was obtained in a manner that infringed or denied any rights or freedoms guaranteed by this Charter, the evidence shaH be excluded if it is established that, having regard to ail the circumstances, the admission of it in the proceedings would bring the administration of justice into disrepute." Unlike s.52(1) or s.24( 1) remedies, s.24(2) remedies are limited to the immediate case and do not have a policy co st. ln those instances where evidence is excluded (under s.24(2)) for novel reasons, this is captured by the "novel interpretation" measure for salience, discussed below.

73 The Supreme Court of Canada c1arified its remedial options under s.24( 1) in

Schachter. 42 In addition to nullification and a finding of "temporary validity" (a suspended nullification to allow legislators time to enact a constitutionally valid law),43 the Court affirmed its power to "read in" and to "read down." According to the Court,

"In the case of reading in, the inconsistency is defined as what the statute wrongly excludes rather than what it wrongly includes. Where the inconsistency is defined as what the statute exc1udes, the logical result of dec1aring inoperative that inconsistency may be to inc1ude the exc1uded group within the statutory scheme. The reach of the statute is effectively extended by way ofreading in rather than reading down.,,44 The

Supreme Court of Canada' s controversial decision in Vriend v. Alberta45 provides the best-known example of "reading in." In Vriend, the Court ruled that the Alberta Human

Rights Code' s failure to explicitly ban discrimination on the basis of sexual orientation violated the Charter' s equality rights, and issued the remedy of "reading in" sexual orientation to the Code.

"Reading down", or severance, entails nullifying only part ofthe law in question-a given phrase, but not an entire subsection-which may profoundly alter the meaning and scope of the legislation. While the Court in Schachter appeared to recognize the novel and activist nature of reading in, the majority of the Justices characterized severance as "an ordinary and everyday part of constitutional adjudication," arguing that "courts have always struck down laws only to the extent of the inconsistency

42 [1992] 2 S.C.R. 679. 43 For purposes here, findings oftemporarily suspended nullification were coded the same as nullification. 44 R. v. Schachter [1992] 2 S.C.R. 679, at 681 per Lamer C.J. [emphasis in original]. 45 [1998] 1 S.C.R. 493.

74 using of the doctrine of severance or 'reading down. ",46 However, there is an important difference between nullifying a section of a law and "reading down" (or more accurately,

"reading out") a particular ward or phrase. As one senior Justice Department official observes:

Usually reading in or reading down is going to be one of the most intrusive remedies. When you just strike down, usually you're giving Parliament a little leeway to devise a solution. You may be saying this one solution is bad, but you've still got all the others to choose from, whereas when they read in [or down] they're saying this is what the law shall be henceforth. It may not be legally impossible ... to come along with a different solution, but ifs going to be politically very difficult.47

Although not discussed in Schachter, the courts also have the option of issuing a constitutional exemption. This occurs when the courts exempt individuallitigants from the application of legislation on constitutional grounds, rather than invalidating or rewriting the law. For example, when Robert Latimer appealed his minimum ten year sentence for the mercy-killing of his disabled daughter, claiming the punishment was cruel and unusual, he did not ask the Justices ta strike down the mandatory sentencing provision, only that they grant a constitutional exemption (they did not).48 Constitutional exemption becomes judicial extension when the exemption (or an entitlement) is granted ta a class of individuals, regardless of whether they appear before the court.

With the foregoing in mind, it is possible to refine the policy cost measure by distinguishing between types ofremedy, based on the degree ofjudicial intrusiveness.

46 R. v. Schachter [1992] 2 S.C.R. 679, at 696 per Lamer C.J. 47 Graham Garton, Department of Justice Canada, personal interview (July 31, 2001, Ottawa). 48 R. v. Latimer [2001] 1 S.C.R. 3. The mandatory sentencing provision for second-degree murder is found in the Criminal Code, R.S.C. 1985, c. C-46, 5S. 235, 745(c).

75 Severance and reading-in both entail outrightjudicial amendment of the law; cases in which either occurs are coded "1" on the ')udicial amendment" variable, with all other cases coded "0". On the other hand, constitutional exemption leaves the law essentially intact, while invalidation leaves rewriting the legislation (or not) to the government. If either of these remedies are employed, l score the case "1" on the "less-intrusive remedy" variable, with all other cases coded "0". The more-intrusive remedies ofjudicial amendment should be more likely to provoke a government appeal than less-intrusive remedies.

B. Salience

Although Zorn identifies three dimensions of salience-the long-term policy impact or "scope" of the lower court' s decision, department clientele interests, and policy significance-he only operationalizes the last factor, and even then his focus is more on the importance of the case than the policy area. In his analysis, a case is more important if a constitutional issue is raised, an amicus curiae (third-party intervener) is present, or if the decision is published. F ollowing this logic, only the presence of an intervener is applicable to my study, as every case will involve a constitutional (Charter) question, and, for the pragmatic reasons cited below, only published decisions are used. Although interventions by interest groups and govemments are common in the Supreme Court's

Charter cases,49 they are relatively rare in Canada's lower courts. Therefore, their

49 Ian Brodie (2001) "Interest Group Litigation and the Embedded State: Canada's Court Challenges Program:' Canadian Journal of Po/itical Science 34: 357-376.

76 presence is a good indicator that the case involves an issue of particular importance.

Cases with interveners are coded "1 ", those without "0".

Unlike Zorn, 1 also measure the scope of a lower court decision, by focusing on novellegai mIe development. As discussed in Chapter One, the courts reguiarly engage in micro-constitutional politics, by interpreting and enforcing "first-order mIes" enunciated in constitutional rights. Examples abound: does "freedom of expression"

(section 2(b)) prote ct only political speech and the press, or commercial expression, lies, hate speech, and visual portrayals such as pornography as well?; do es the right "to retain and instmct counsel" (section 1O(b)) entail a right to Legal Aid or pro bono counsel?; does the "right to life, liberty and security of the person" (section 7) extend to fetuses?

When a lower appeal court crafts a new legal mIe that will erode the government's authority in other policy fields, or vis-à-vis other state actors or individuals, the government should be more likely to appeal the decision. Examples of such mIes include the application of the Charter to new issues and jurisdictions, the creation of new judicial remedial powers, the augmentation of judicial discretion over the application of the

Charter, and expanding the scope of justiciable rights. A legal interpretation is "novel" if there is no Canadian precedent cited for the mIe. Cases where a novellegal mIe is developed are coded as "1 ", and aU others are coded "0".

C. Reviewability

Zorn's measure for reviewability, citation of intercourt conflict over the issue in question by the court of appeal deciding the case, is adopted here. Flemming and Krutz' s recent study of Supreme Court of Canada leave to appeal decisions found that "a daim of

77 conflicting lower court decisions, as the American literature amply suggests, is more likely to be heeded by Canada's Court than an argument lacking this trump card."so This conclusion is as expected, not only because it reflects the American courts' behaviour, but because, "as the court of final appeal in a centralized national court structure, the resolution of conflicting lower court decisions is the basis ofthe Supreme Court's responsibilities. "SI

Mirroring the Court's agenda, the Department of Justice Canada's guidelines explicitly cite jurisprudential conflict between appeal courts as a matter of "national importance" warranting appeal. S2 The Justice Department also has its own reasons to appeal because of intercourt conflict. As a "repeat player" with a national focus, the federal government seeks legal coherence and consistency across regional jurisdictions.

As one senior department official explained, "we consider it generally not a very good state of affairs if [federal government] lawyers in one province are precluded from taking a position that they might be able to take in another province.,,53 1 code cases "1" where the appellate court cites intercourt conflict, or where a precedent is cited by the court but not followed, and "0" if not.

D. Winnability

Zorn operationalizes winnability in three ways: whether there were reversaIs among the lower courts as the case was appealed; whether there was dissent on the

50 Flemming and Krutz (2002), 241. 51 Flemming and Krutz (2002), 241. 52 Department of Justice Canada (2000), V-23-1. 53 Robert Frater, Department of Justice Canada, personal interview (April 17,2002, Ottawa).

78 highest court of appeal before the Supreme Court; and whether the ideological orientation of the lower appeal court ruling was inconsistent with the ideology of the Supreme Court.

Cases in which the appeal court reversed the trial or lower appellate decision54 indicates that the federal government won at sorne stage. This should encourage the belief that another victory is possible, in contrast to scenarios where the government has ne ver prevailed. As Horowitz contends, "a prediction of an unfavourable result on appeal, based on an already unfavourable result in the court below, is more credible than the same prediction not based on an already adverse decision."55 The government should be more likely to appeal cases involving a reversaI (coded "1") than those without (coded

"0"). Similarly, a dissent on the court of appeal also indicates that the government' s position has merit, and that sorne expectation of success is not unreasonable. Cases where the government loses by a split vote are coded as "1 ", and unanimous rulings coded "0".

Although aIl ofZorn's measures for winnability are sound, the Canadian context presents a major difficulty for the ideological measure, as there is no scholarly agreement on the Supreme Court of Canada' s ideological orientation or level of activism. 56 The

Court has certainly been more activist since 1982 in terms of nullifications and crafting

"positive" remedies such as "reading in", and its decisions have tended to exp and the definitions of Charter rights, particularly with respect to equality rights and legal rights.

54 Sorne jurisdictions, such as Ontario and British Columbia, have multiple levels of appeal courts. For example, in B.C., cases typically progress from the (confusingly titled) B.C. Supreme Court, through the B.e. County Court, to the B.e. Court of Appeal. 55 Horowitz (1977b), 94. 56 See, for example, Mandel (1994); Russell, Morton & Riddell (1994); Hiebert (1996); Bakan (1997); Hogg and Bushell (1997); Hogg and Thonon (1999); Kelly (1999b); Manfredi and Kelly (1999; 2001); Morton and Knopff (2000); Roach (2002).

79 However, rights c1aimants have been successful only 30% of the time overall, with the

Court usually deferring to governments in section one "reasonable limits" analysis,S7 and iIlegally obtained evidence has been admitted through the section 24(2) discretionary exclusionary ruie over haif of the time (55% of cases).58

An additional problem for the ideology measure is created by this dissertation's focus on Charter cases. Zorn examined aIl types of cases, and coded the ideological direction of the iower court decision based on Spaeth's widely-used criteria, one ofwhich is whether the ruling is "pro-rights claimant."S9 Since every Charter case the govemment loses is necessarily "pro-c1aimant", the ideological direction of the decisions are constant, and therefore of no analytical value. For these reasons, the possible effect of a case' s ideological dimension must be bracketed.

To summarize, the hypothesized explanatory factors are as follows:

Costs Xl: the immediate tinancial costs of the lower court' s ruling (non-criminallitigation) X2: the immediate policy costs of the lower court's ruling 1 Gudicial amendment) X3: the immediate policy costs of the lower court's ruling II (less-intrusive judicial remedy)

Salience X4 : the presence of an intervener at the lower appeal court X5 : the scope of the lower court's ruling (novel ruie development)

Reviewability x6 : the citation of intercourt conflict by the lower court

Winnability X7 Iower court reversaIs x8 : dissent on the lower appeal court

57 Hiebert (1996). 58 Kelly (1 999b). 59 Harold Spaeth (1994) United States Supreme Court Judicial Database: 1953-1992 Terms (Ann Arbor: ICPSR).

80 where x 1-8 are positively related to the government choosing to challenge the lower court's ruling by appealing to the Supreme Court (y). As detailed above, all explanatory factors as well as the dependent variable (not appeal/appeal), are coded as durnrny variables, scored as "1" for the presence of the factor and "0" for its absence.

In summary, l expect that government lawyers will take these sundry factors into account when deciding whether to appeal an adverse decision to the Supreme Court of

Canada. That said, the existing literature on government appeal decision-making suggests it is unlikely that these factors are given equal weight. A clear conclusion from the neo-institutionalist literature is that "where you stand depends on where you sir'; that individuals within an organization are motivated by different factors depending on their position within that organization. For exarnple, the Cabinet Minister with the Finance portfolio is likely to be more concemed with government-wide fiscal matters such as the national debt than the Minister of Health, who is probably more concemed with the state ofhealth care and her relationships with clientele in the health care industry. Horowitz observed a similar phenomenon in his surveys ofU.S. Justice Department and executive agency counsel, and Zom's later statistical analysis confirrned it. 60 Specifically, the

Solicitor General, at the top of Justice's hierarchy, is less motivated by irnrnediate policy costs than lower-level counsel who deal regularly with those agencies responsible for the policy in question. Conversely, the factors that most concem the SG-jurisprudential development, the governrnent's reputation with the Supreme Court, and the related issues of reviewability and winnability-are of less concem to lower-Ievel counsel.

60 Horowitz (l977b); Zorn (1997; 2001).

81 A similar pattern might be expected for the Department of Justice Canada. Where the decision to appeal is fairly centralized, the appeal decision ca1culus should be somewhat biased against factors oflesser concern to central decision-makers: namely, immediate policy costs, although judicial amendments should be more likely to prompt an appeal than less-intrusive remedies. That said, this hypothesis flies in the face of anecdotalaccounts that govemment lawyers typically try to protect legislation.

Furthermore, the significant elements of decentralization in the Justice Department's appeal decision pro cess-Regional Office and LSU representation on the NLC, and the fact that Regional Offices can be a "veto point" for appeals-muddy the conceptual waters, and may weaken any statistical relationships. In short, it is difficult to make strong hypotheses about the relative weight of explanatory factors in the Canadian case, and is perhaps better approached inductively.

Admittedly, this is a less than optimal solution, but a briefnote ofjustification before concluding: the unit of analysis in this dissertation is final appeal decisions by the

Department of Justice Canada. In contrast, Zorn's study analysed appeal decisions of the

Solicitor General's Office and appeal recommendations by the Department of Justice, based on their internaI records. This allows him to compare appeal decisions by two different institutions within the Justice Department hierarchy, and therefore to test (and confirm) empirically the foregoing hypotheses about institutional position. Unfortunately,

1 am unable to replicate this anaiysis in Canada, because internai data from the Canadian

Justice Department about appeal recommendations by Regional Offices and LSUs are not publicly available; indeed, senior Department officiaIs indicated they were unsure

82 Regional Office records even existe d, and, in any case, such records would be protected by solicitor-c1ient privilege.61

In the following chapter, l discuss the case selection criteria used to generate the database of appeal decisions, and subject the explanatory factors and hypotheses to statistical testing. In addition to these findings, l provide a statistical overview of the cases, with attention to possible patterns over time, and across region, policy area and party.

61 Graham Garton, Department of Justice Canada, personal interview (July 31, 2001, Ottawa); Robert Frater, Department of Justice Canada, personal interview (April 17,2002, Ottawa). In a tclling statement about the Department's opinion oftransparency, one official said: "My guess is that ifwe published that kind of information here, then there would be cases where the media and others would be ail over [he Minister, saying 'why didn't you overrule? How can you let [for example] the child pomography law go to the Supreme Court wh en you should be legislating? What's wrong with you?'"

83 CHAPTER FOUR

Analysing the Decision to Appeal: Data and Statistical Results

Despite suffering relatively few defeats in the lower appellate courts, the Canadian

federal government appeals only a fraction of these to the Supreme Court of Canada,

effectively conceding the rest. These concessions carry costs-in the form oflost policies,

increased policy expenditures, financial damages, and lost authority-as do decisions to

proceed with an appeal, which use departmental resources and entail opportunity costs. In

short, these decisionsmatter, but, to date, no one has sought to explain the government's

appeal choices. In this chapter, 1 test the model of appeal decision making outlined in

Chapter Three. First, however, it is necessary to identify and explain the case selection

criteria employed, and to provide an overview of the federal government's Charter

litigation before the country' s highest appeal courts.

I. Data Selection

Before proceeding, it is necessary to clarify two key terms used herein. First, the

"federal government" refers only to those institutions identified earlier which are

represented by Justice Department lawyers (hereafter referred to collectively as "AG

Canada") in litigation proceedings. Second, the dissertation promises to analyse "appeals" to the Supreme Court of Canada. With the few exceptions discussed in Chapter Two, the federal government does not enjoy a right of appeal to the Supreme Court, but must apply

84 for leave to appeal, which may be denied. The body of appellate rulings issued by the

Supreme Court of Canada is not, therefore, an accurate reflection of the government's

decision to appeal; attention must focus instead on leave to appeal applications, which

signal the government' s intention to contest rather than concede. Thus, the government' s

decision to "appeal" means its decision to seek leave to appeal or to appeal as of right.

The following case selection criteria were employed to construct the database of cases

requiring a decision whether to appeal:

1. The case was a reported decision of a PCA or the FCA (see below),

2. involving a daim under the Canadian Charter ofRights and Freedoms

3. decided between 1982 and 2000 inclusive,

4. where the federal government was a party (appellant or respondent) to the case, and

5. the federal government lost the dispute (i.e., their appeal was dismissed, or the appeal was allowed when the federal government was the respondent) in the penuItimate court of appeal. 1

Appeals to the Supreme Court of Canada involving the federal government come from the penuItimate courts of appeal, ofwhich there are two major categories: "section

96" and "section lOI" courts, so-named for the sections of the Constitution Act, 1867 under whose authority they were created (see Figure 4.1). There is one section 96

"superior" court, or "Court of Appeal," in every province and territ ory, which 1 refer to

Notably, a party (and sometimes, even an intervener) may appeal to the sec on a point of law even though it technically won the appe,ù. However, for the sake of simplicity and to prevent selection bias, only cases where the federal govemment was a direct party and lost on the disposition of the appeal are included.

85 collectively as "provincial Courts of Appeal" or PCAs.2 They hear appeals from the lower

section 96 trial and intermediate appeal courts, as weIl as trial courts under purely

provincial authority ("section 92 courts"). Unlike section 96, section 101 does not

establish courts per se, but rather empowers the federal government to create "any

additional Courts for the better Administration of the Laws of Canada." In 1971, the

Federal Court of Canada was established, consisting of a trial division (FCTD) and

appellate division (Federal Court of Appeal, or FCA).3

Reported decisions are used because the detailed case information required for the

study is typically unavailable for unreported decisions. The Canadian Abridgment, a

comprehensive electronic list of aIl Court of Appeal and SCC decisions, was used in

conjunction with major law reports (Supreme Court Reports (S.C.R), National Reporter

(N.R), the Dominion Law Reports (D.L.R), Federal Court Reports (F.C.), Canadian

Criminal Cases (C.C.C.), Western Weekly Reports (W.W.R), Atlantic Provinces Reports

(A.P.R)) and several provincial reports (Ontario Reports (O.R.), Alberta Reports (A.R), etc.) to compile the list of reported cases. The study' s start date of 1982 marks the adoption of the Charter ofRights and Freedoms. Notably, this time frame means that the

2 Although organised by province and subject to provincial administration, section 96 court judges are appointed, paid, and (rarely) disciplined by the federal government. Because ofthis shared responsibility, section 96 courts are the linchpin of Canada's unified court system, in marked contrast to a divided or "dual" court system, such as in the u.s. 3 The FCA's appellatejurisdiction was extended in 1983 to include appeals from the newly-created Tax Court of Canada. In addition, the federal government created the Court Martial Appeal Court (CMAC) to handle cases involving prosecution of military personnel. The FCA has exclusive jurisdiction only over civillitigation initially brought against the federal government and cases involving federal administrative law (immigration, Employment Insurance, et al.), while the PCAs have exclusive jurisdiction to hear most criminal appeals (in broad terms, i.e., including non-Code offenses, although not tax fraud or combines violations). For aIl other types of cases, the FCA and PCA _l~ve concurrent jurisdiction, with the plaintiff choosing the venue.

86 cases used for the study are not a sample, but the actual universe of Charter cases involving the federal government that me et the other criteria.

Information regarding leave to appeal applications from 1986-2000 was obtained from Quicklaw's Supreme Court of Canada Appeals database (SCCA), and from the

S.C.R. and the N.R. for 1982-1986. Determining wh ether the AG Canada filed an appeal as of right required cross-referencing the Court of Appeal decision with the leave to appeal records and S.C.R.; if the Supreme Court ruled on an appeal by the federal government but no leave was sought, it must have been brought by right.

Figure 4.1: The Canadian Court System, General Structure

Supreme Court of Canada (1875- ) Section 96 Courts Section 101 Courts

Provincial Court Martial Federal Courts of Appeal Appeal Court Court of Appeal (1867- ) (1971- )

S pecialized Section 96 Trial Courts Tax Court (e.g., Family Court) of Canada (1983- ) Section 96 (General) Trial Court Federal Trial Court Section 92 (Provincial) (1971- ) Trial Court

Source: compiled by author.

87 II. Data and Findings

The first four case selection criteria generated 593 cases, ofwhich the federal

government lost only 159-a remarkable 73.2% overall success rate on case disposition. 4

The government was only slightly less successful in civillitigation (168 wins of 23 7 cases,

or 70.9%) than in criminal prosecutions (268 wins of356 cases, or 75.3%). Wheeler5

suggests an alternative measure of success that factors in the role of the litigant-that is,

appellant or respondent-to address the phenomena that "appeal courts tend to affirm

rather than to reverse the lower court" and that governments appeal much less often than

they are appealed against. 6 Wheeler's measure, "net advantage," is calculated by

subtracting a litigant's loss rate when respondent from that litigant's success rate when

appellant. The results for the government of Canada, as Crown ("R.") in criminal

prosecutions, in civillitigation, and overall, are presented in Table 4.l. As the table

reveals, the federal government has a tremendous net advantage of +46.5% in Charter

cases, and this advantage is only slightly higher in criminal cases than in civillitigation.

For comparison, consider that the net advantage of the U.S. government, widely

renowned for its phenomenal success, is +45.1% in the U.S. Courts of Appea1. 7 These

findings are consistent with Galanter's "party capability theory," according to which

4 As F.L. Morton and Avril Allen illustrate, case disposition is only one way ofmeasuring litigant success. Other, more substantive measures include the effect of the case on the "policy status quo" and the creation offavourable or unfavourable legal resources (precedents). See Morton and Allen (2001) "Feminists and the Courts: Measuring Success in Interest Group Litigation in Canada." Canadian Journal ofPolitical Science 34: 55-84. However, to compile the database of appealable losses, case disposition is the appropriate measure of success. 5 Stanton Wheeler, et al. (1987) "Do the 'Haves' Come Out Ahead? Winning and Losing in State Supreme Courts, 1870-1970," Law and Society Review 21: 403-45. 6 Peter McCormick (1994), 159. 7 McCormick (1994), 160, calculated from data in Donald Songer and Reginald Sheehan (1992) "Who Wins on Appeal? Upperdogs and Underdogs in the United States Courts of Appeals," American Journal ofPolitical Science 36: 235-58.

88 governments, as the ultimate "repeat players" with unmatched experience and vast

financial and legal personnel resources, are more likely to win than individuallitigants. 8

However, as Zorn observes, party capability theory is only one possible explanation for

this pattern of government success; others include "institutional deference" by the

Supreme Court,9 and Zorn's own argument, that because the government appeals oruy the

strongest cases on a variety of criteria, "the odds of achieving a successful outcome on the

merits will naturally be quite high."lo While Zorn's argument has intuitive appeal, it does

not explain why the Canadian government's success rate as respondent roughly equals that

when appellant.

Table 4.1: Federal Government Success Rates in Charter Cases (1982-2000), by Litigant Roles

: ...... ::::: :Revers~l(wiÏ1) ·R~ve ..sal(loss) •.•.• Govermri~rit

Crown ("R.") 71.7% 23.9% 75.3% +47.8% n=l13 n=243 n=356

Civil Litigant 75.9% 31.6% 70.9% + 44.3% n=79 n=158 n=237

Total 73.4% 26.9% 73.2 + 46.5% n=192 n=401 n=593

The cases involved a wide range of federal policy areas, but the most common were drug enforcement (Narcotics Control Act prosecutions alone represented almost half

8 Galanter (1974) and KeYÏn T. McGuire (1998) "Explaining Executive Success in the V.S. Suprcme Court." Political Research Quarterzv 51: 505-26. 9 Perry (1991), cited in Zorn (2002). 10 ZoOm (2002). 163.

89 of the government's caseload), immigration, and taxation (Income Tax Act, Excise Tax

Act). The government's success varies among the lower courts of appeal, ranging from a low of 44.4% in the Prince Edward Island Court of Appeal to a high of88.2% in the

Manitoba Court of Appeal. Table 4.2 reveals the surprising finding that the federal government is more successful as one moves west, with the lowest success rate in Atlantic

Canada and the highest in BC and the Prairies. There is also a small difference between the government's success rate in the federal section 101 courts (FCA, CMAC) (70%) and the section 96 PCAs (including Territorial courts of appeal) (75%). This discrepancy is somewhat surprising, since the judges ofboth types of courts, the labels "federal" and

"provincial" notwithstanding, are appointed by the federal government.

Table 4.2: Federal Government Losses by Court of Appeal

Coutj:Nùmbërof Total LO~~es~~.%<.%ofTotal . ofAppeal . < Losses ·Nufubè.;at)ofCasêsin '.. . Losses ...... <êases/ '. ·····\ê.6rA~··.· ....•. .'...... <. ..•...... NFCA 4 10 40.0 2.5 NSCA 10 27 37.0 6.3 PEICA 5 9 55.6 3.1 NBCA 4 11 36.4 2.5 QCA 11 34 32.4 6.9 OCA 29 118 24.6 18.2 MCA 2 17 11.8 1.3 SCA 11 27 40.7 6.9 ACA 7 35 20.0 4.4 BCCA 18 114 15.8 11.3 FCA 52 174 29.9 32.7 CMAC 3 8 37.5 1.9 YTCA 1 5 20.0 .6 NWTCA 2 4 50.0 1.3 Total 159 593 26.8 100.0

90 Despite the relatively few appeal decisions the government has to make in Charter

cases due to its remarkable success rate, appeals to the SCC are not undertaken routinely.

Indeed, the federal government appealed only a fraction of its losses, a total of 47 times

(29.6%),42 by leave and five by right. Perhaps because ofits selectivity, the federal

government enjoys a high level of success at securing leave to appeal from the Supreme

Court: of 42 applications, aIl but seven were allowed (83.3% success rate). Surprisingly,

the existence of a right to appeal did not increase the likelihood of appealing. Of the 17

cases where the right existed, only five were appealed (29.4%), compared to 42 of 142

potential applications for leave (29.6%).

Table 4.3: Federal Government Appeals to the Supreme Court of Canada

ByLélve.·· BYRight.·. Action ..... ••·•··• •• Total . .. ···(n) (n) •.•.•...... •• (n) Not Appealed 100 12 112

Appealed 42 5 47

Totals 142 17 159

SCC Granted 35 nia nia Leave

The data reveal the absence of a clear trend in either losses or appeals with respect to time, although, as might be expected, both were relatively high in the first year of

Charter litigation (see Figure 4.2).11 This said, governments have been getting steadily

Il No cases meeting the selection criteria emerged before 1983. Hence the absence of 1982 in Figure ·U.

91 more successful. The 1982-1984 Liberal government under Trudeau (and Turner) lost

32.8% (19/58) ofits Charter cases, the 1984-93 Progressive Conservatives under

Mulroney (and Campbell) 27.7% (86/310), and the 1993-2000 LiberaIs under Chrétien

24.0% (154/225). This trend may be explained by any ofseveral factors: the courts may have become less activist over time; the govern~ent has improved its quality of argument before the courts; the government has gotten better at drafting laws and administrative policies, thereby avoiding rights violations; or older legislation, which is more likely to run afoul of the Charter, is progressively rarer over time, as it is amende d, replaced, repealed or judicially remedied. Notably, governments have also appealed los ses at a progressively greater rate. The 1982-84 LiberaIs appealed 26.3% (5/19) and the subsequent Tories

27.9% (24/86), while the Chrétien government has been the least likely to concede defeat, appealing 33.3% of the time (18/54).

As Table 4.4 indicates, policy areas were not treated equally in the decision to appeal. Cases involving the two most frequently litigated policy issues-narcotics and immigration-were appealed at the lowest rate proportionately (as expected, in the case of the former), while losses involving competition (combines) were appealed, and aIl but one in each of the extradition and prisons categories. Regarding the lower court origin of the case, the government was slightly more likely to appeal a decision of the FCA

(19/55=34.5%) than a PCA (28/104=26.9%). Within the PCAs, Ottawa appealed decisions from Atlantic Canada (13.0%) and the Prairies (Manitoba, Saskatchewan,

Alberta) (21.1%) at a rate below the 29.6% average, while decisions from the Courts of

Appeal in Ontario (31.0%), Quebec (45.5%), and British Columbia (31.6%) were

92 Figure 4.2: Federal Government Losses in Courts of Appeal,

and Appeals to the sec in Charter Cases, 1982-2000

16 - C.A. Losses 14 F --- Appeals to SC C r 12 e q 10 u e 8 n 6 ,\ C /\. , \ /\ """".... \ \ / \ y 4 , , / \. """"...... 1 \ / \ n=47 ,\ ..... / "" "" \. 1 \ , '....' ..... , 2 ...."" .... , \. , \ .j / '...... , \. 1 \ ", / . ...., \ 1 \ ...... / \ 1 \ , ..... / 0 \ ..... '1i" 1983 1985 1987 1989 1991 1993 1995 1997 1999 1984 1986 1988 1990 1992 1994 1996 1998 2000

Year

93 Table 4.4: Government Appeals to the Supreme Court, by Policy Area

...< . ., . Policy Area < ····

94 appealed at an aboye-average rate. However, one should not infer too much from these regional figures, due to the small number of cases and appeals in sorne provinces.

Finally, the Justice Department' s appeal decisions vary according to the Charter right c1aimed. Among the cases lost by the federal government at the penultimate court level, legal rights c1aims (sections 7-14) are by far the most common (n=129), but are appealed at one of the lowest rates (27.1 %). In contrast, Ottawa appealed half of its losses in equality rights cases (n=18) and 42% ofthose involving fundamental freedoms

(n=12), the second- and third-largest-sized categories respectively. This is not to say that the government' s appeal decisions were driven by the Charter right in question; rather, a more compelling interpretation is that equality rights cases tend to contain sorne of the factors, discussed above, which encourage appeals, while legal rights cases (arising in criminal prosecution of individuals) do not.

Table 4.5: Federal Government Appeals to the Supreme Court of Canada, by Charter Right

... >Numberof·. Numberof Charter Right Percentage ••••••••• Lossës*· . Appeals Appealed Fundamental Freedoms 12 5 41.7 Democratic Rights 6 2 33.3 Mobility Rights 4 2 50.0 Legal Rights 129 35 27.1 Equality Rights 18 9 50.0 Language Rights 1 0 0.0 Remedy (s.24.1) 4 1 25.0 Aboriginal Rights (s.25) 1 1 100.0 Notwithstanding Clause 1 1 100.0 Other (generaVstanding) 3 0 0.0

* Totals exceed 159 losses and 47 appeals due to multiple rights issues in several individual cases.

95 Turning to the explanatory factors hypothesized in Chapter Three, summary statistics for the variables employed are presented in Table 4.6. To restate, 1 hypothesize a

"procedurally rational" model of appeal decision making, in which government lawyers are influenced by a variety of factors related to costs, case importance, and the prospect of gaining entry to and winning in the Supreme Court of Canada. The immediate financial costs associated with losing in the lower court are represented by the type of case, with civil cases typically entailing greater costs than criminal prosecutions. A judicial remedy of legislation captures the immediate policy cost, with a distinction drawn between judicial amendment (reading in, severance) and the less-intrusive remedies of invalidation and granting constitutional exemptions. A case's importance or "salience," is heightened by the presence of third party interveners, or a novel constitutional interpretation that expands judicial power or otherwise limits governmental authority. The citation of intercourt jurisprudential conflict increases the odds the Supreme Court will hear the case, or what 1 refer to as the "reviewability" factor. Finally, reversaIs as the case proceeded from trial through appeals and dissent on the lower appeal court encourage the government' s belief that it can win on appeal to the Supreme Court.

As discussed in the last chapter, all factors (or independent variables) as weIl as the dependent variable are coded dichotomously, "1" if present and "0" if not; for this reason, the mean of each variable represents the probability that the factor is present. As Table

4.6 reveals, aIl variables except "Lower Court ReversaI" are highly skewed, with the factor being relatively rare. The preliminary data from bivariate correlation and logistic regression are summarised in Table 4.7, and trom multivariate regression in Table 4.8.

Because factors are dichotomous and highly skewed, estimation is by maximum likelihood using binary logistic regression.

96 Table 4.6: Summary Statistics for Explanatory Factors

.. ."',"" ·S 'D .. , ." Variable < .' ..,>. »>, .'>.'/. .,Mean td. ~v. R.angè> •.•. N (factor-l) Dependent Variable Appeal Decision 0.30 0.46 1,0 47 Independent Variables Cost Factors Non-Criminal Case 0.33 0.47 1,0 52 Less-Intrusive Remedy 0.21 0.41 1,0 34 Judicial Amendment 0.08 0.27 1,0 13 Salience Factors Novel Interpretation 0.15 0.36 1,0 24 Intervener in Lower Court 0.15 0.36 1,0 24 Reviewability Factors Intercourt Conflict 0.17 0.38 1,0 27 Winnability Factors Lower Court ReversaI 0.69 0.47 1,0 109 Lower Court Dissent 0.23 0.42 1,0 37

Note: N=159.

Table 4.7: Explanations for the Decision to Appeal, Bivariate Regression and Correlations

'." ,.... Regression' . .. Explallatory Factor " ,"C 0rrelatioIl, .,.". , CoeffiCient ·JPears~ri's rl "., Non-criminal case 0.62 (.36)* .14* Less-Intrusive remedy 0.50 (.41) .10 Judicial amendment 1.86 (.63)** .26** Novel interpretation 1.46 (.46)** .27** Intervener 1.26 (.46)** .23** Intercourt conflict 0.61 (.44) .11 Lower court reversaI -0.17 (.37) -.04 Court of Appeal dissent 1.11 (.39)** .23** N=159

Note: Unstandardised regression coefficients, with standard errors shown in parentheses. Estimation is by maximum likelihood. * p < .10 ** p < .01 (regression one-tailed, correlation two-tailed)12

12 Where the dissertation uses the population of cases rather than sample data, p values are technically ,,_ irrclevant. Nevertheless, they provide a uscful and 'widely-understood metric for evaluating rcsults.

97 Table 4.8: Explanations for the Decision to Appeal, Multivariate Regression

Explanat()ryFactor ... Coefficients Non-criminal case 0.59 (.42) Less-Intrusive remedy 0.59 (.47) Judicial amendment 1.70 (.74)* Novel interpretation 1.29 (.54)* Intervener 0.16 (.56) Intercourt conflict 0.77 (.49) Lower court reversaI -0.18 (.43) Court of Appeal dissent 1.31 (.44)** Constant -2.00 (.44)** -2 Log Likelihood 160.97 Pseudo R2 .260 (Nagelkerke) N=159

Note: The column entries are unstandardised regression coefficients, with standard errors shown in parentheses. Estimation is by maximum likelihood. * p < .05 (one-tailed) ** p < .01 (one-tailed)

III. Discussion

The negative constant in Table 4.8 is consistent with the federal government's tendency not to appeallosses. Overall, the model represents a good statistical fit to the data, with signs for all but one factor in the expected direction, and several factors achieving accepted levels of statistical significance. The predicted probability of a positive decision to appeal, holding all independent variables at their mean values, is .255, compared to the observed probability of .296 (see Appendix at end of chapter for an explanation ofhow to convert logistic regression coefficients into probabilities). With respect to the individual variables, no correlations are stronger than r=.27, although the se figures are likely depressed by the small number of cases. Multivariate regression using aIl eight factors largely confirms the preliminary results, although two of the promising factors (non-criminal case and intervention) are no longer significant. l discuss each category of factors in turn below.

98 A. Costs

Cross-tabulations reveal that only 25.2% (27/107) of cases involving federal

prosecution ("criminal" cases, for lack of a better term) were appealed, compared to

38.5% (20/52) ofnon-criminal cases. Similarly, the correlation and bivariate regression

coefficient for the criminal/non-criminal variable is positive, confirming the hypothesis that

cri minai cases are less likely to be appealed. The factor is only marginally significant

(p=.16) in multivariate regression, perhaps because it is weakly correlated with

intervention (r=.19, p<.05) and novel interpretations (r=.16, p<.05).

The hypotheses regarding policy costs fares somewhat better, especially that

concerning judicial amendment versus less-intrusive remedies. Although judicial remedies

as a whole are positively correlated with appealing (1=.23,p<.01), as predicted there is a

marked difference between the two categories of remedies. Judicial amendment

(severance, reading-in) is a much stronger factor than the presence of a less-intrusive

remedy (invalidation, exemption). Nine of 13 cases involving judicial amendment by the lower appeal court (69.2%) were appealed to the sec, compared to 13 of33 invalidations

(39.4%). Furthermore, while judicial amendment is among the strongest predictors of a government appeal, there is no statistically significant relationship between less-intrusive remedies (or even invalidation by itself) and the dependent variable. This evidence refutes the claim of sorne interest groups13-and, more interestingly, sorne government lawyers14-that the government typically defends impugned legislation, and suggests a more nuanced governmental response to judicial activism. In light of the comments of

13 Shilton (1993). l-l Mitchell (1993): Gallagher (1993).

99 government officiaIs presented in Chapter Three, the government' s greater dissatisfaction withjudicial amendment than with invalidation is probably due to the degree offreedom the remedy gives the government to respond legislatively. Invalidation offers the most freedom in this respect, while either form of amendment ("reading down" or "reading in") offers the least, and as such, appealing a judicial amendment may offer an easier way to

"undo the damage" to government policies. The data therefore suggests that government appeals are, to sorne extent, the product of calculated decision-making by central officiaIs concerned with the level ofjudicial (and governmental) power.

B. Salien ce

Both factors in this category-the presence of an intervener, and scope (novel legal interpretation)-perform weIl in initial bivariate analysis, with novel interpretation emerging as the strongest factor overall. The latter finding has particularly strong intuitive appeal, as government decision-makers at thehighest levels should be concerned with jurisprudential developments that weaken the government vis-à-vis the judiciary. Novel interpretation continues to be among the strongest factors in multivariate regression, but, as Table 4.8 shows, the intervener variable weakens considerably. In addition, the presence of an intervener correlates somewhat strongly with novel interpretation (r=. 3 7, p<.O 1) as weIl as judicial amendment (r=. 3 2, p<. 0 1), and, to a lesser extent, judicial dissent (r= .14, p<.1 0). On the surface, this suggests intervention may exercise an indirect effect, by influencingjudges-an intuitively-appealing interpretation, since this is the purpose of intervention. Specifically, interveners may emboldenjudges on the lower court to greater activism and creativity, which in turn encourage the government to appeal.

100 Novel interpretations are indeed more likely when an intervener is present: 42%

(10/24) of cases with interveners involved novel interpretation, compared with only 10%

(14/135) of cases with no intervener. However, cross-tabulations between novel interpretation and the dependent variable, while controlling for the presence of an intervener, reveal an even more complex relationship than that suggested above.

According to the intuitive explanation, novel interpretations, while more likely given an intervener, should provoke appeals regardless of whether an intervener is present. This is not the case. When no intervener was present, only 35.7% of the instances of novel interpretation were appealed, compared to 90% of instances appealed when an intervener was present. The measures of association tell a similar story. Novel interpretation bears

Table 4.9: Cross-Tabulations, Federal Government Appeals by Novel Interpretation by Intervener

Novel Interpretation Intervener no yes Total No SCC Appeal No Count 92 9 101 % within Novel 76.0% 64.3% Interpretation 74.8% Yes Count 29 5 34 % within Novel 24.0% 35.7% Interpretation 25.2% Total Count 121 14 135 % within Novel 100.0% 100.0% 100.0% 1nterpretation Yes SCC Appeal No Count 10 1 11 % within Novel 71.4% 10.0% 45.8% 1nterpretation Yes Count 4 9 13 % within Novel 28.6% 90.0% 54.2% 1nterpretation Total Count 14 10 24 % within Novel 100.0% 100.0% 100.0% 1nte rpretati 0 n

101 no statisticaIly significant relationship to the appeal decision with no intervener present

(tau-b=.083, sig.=.389), but is strongly related when an intervener is present (tau-b=.608, sig.=.OOO). To repeat, novel interpretations are significantly related to appealing only in conjunction with the appearance of an intervener.

While there is clearly a strong relationship between these two factors, how to interpret the data is less obvious. One interpretation is that the presence of an intervener signaIs to government lawyers that a case involving a novel interpretation is of particular importance, and should therefore be appealed. A second is that the government is inclined to appeal in cases involving particularly important novel interpretations, and that these are the types of cases which also attract interveners. In other words, the same factor which encourages the government to appeal-case importance, tied to nov el interpretation-also draws interveners, but intervention is not causally related to appealing. It is not possible, using statistical methods, to determine which interpretation is more accurate, but one conclusion is consistent with both-intervention and novel interpretation are cumulative indicators of case importance, as an appeal is much more likely in the presence of both factors than of either factor by itself On the other hand, the presence of an intervener did not increase the probability of appealing when there was judicial amendment (4 of 6 with no intervener, 5 of7 with intervener) or judicial dissent (14 of28 versus 4 of9).

In light of these findings, an alternative regression model was tested, including an interaction term for novel interpretation and the presence of an intervener. As Table 4.10

2 shows, the overall goodness offit measure (Nagelkerke pseudo-R ) improves to .295 with the inclusion of the term, and the new variable ("Novel Interpretation*Intervener") is statisticaIly significant as weIl as the strongest explanatory factor. As expected, novel

102 Table 4.10: Explanations for the Decision to Appeal (with Interaction Term), l\'Iultivariate Regression

...... ExplanatoryFactor ...... Coefficients . Non-criminal case 0.57 (.43) Less-Intrusive remedy 0.53 (.48) Judicial amendment 1.56 (.79)** Novel interpretation 0.56 (.66) Intervener -0.62 (.71) Intercourt conflict 0.81 (.49)* Lower court reversaI -0.19 (.44) Court of Appeal dissent l.46 (.45)*** Novel Interpretation*Intervener 2.92 (1.45)** Constant -l.92 (.45)*** -2 Log Likelihood 156.03 Pseudo R2 (Nagelkerke) .295 N=159

Note: The column entries are unstandardised regression coefficients, with standard errors shown in parentheses. Estimation is by maximum likelihood. * p <.10 ** p <.05 *** P <.01

interpretation is no longer significant by itself, but the regression coefficients for lower court dissent and judicial amendment remain strong.

The divergent findings for novel interpretation and judicial amendment are, however, somewhat puzzling. Both measures tap a similar motivation for government appeals: to defend policy-making authority from judicial intrusion. This being the case, it is surprising that novel interpretations are not a significant influence on their own, while judicial amendments are. The explanation may lie with the profile or "visibility" of each with government decision makers. Judicial amendments obtain a high profile, as

Iegisiation is necessariIy altered, thereby drawing the attention of politicai officiaIs responsible for the law or policy, as well as bureaucrats who administer the Iaw and those who are directly affected by it. These vested interests are almost certain to bring the matter to the attention of senior Justice Department officiaIs, either through the Minister,

103 Legal Services Unit s, issue-specifie units in Justice Headquarters, or when consulted as

clientele by the National Litigation Committee. In contrast, novel interpretations may not

decide the ultimate out come of the case, and as a consequence, they can be difficult to track for vested interests. Thus, perhaps only the most important novel interpretations achieve the high profile necessary to provoke an appeal. Notably, this explanation is consistent with the argument, offered above, that there is no causal relationship between intervention and novel interpretation, but that particularly important cases-that is, those in which novel interpretations are likely-both attract interveners and spur appeals.

C. Reviewability

There are several reasons to expect that Justice Department decision makers would be concerned with gaining access to the Supreme Court. Evidence from abroad suggests that this factor exercises a powerful influence on appeal decisions by government lawyers. 15 As noted in Chapter Three, Department officiaIs are cognisant of the Court's docket constraints, and characterise intercourt jurisprudential conflict as a matter of

"national importance." For the same reason, this factor taps not only a tactical dimension of appeal decision making, but also a more legalistic motivation (that is, resolving legal indeterminacy). It is surprising, then, that there is little statistical evidence in support of reviewability as an explanatory factor. There is a very weak correlation between intercourt conflict and the decision to appeal, but neither the correlation nor the bivariate regression are statistically significant. The same was true in

15 Zorn (2002).

104 the initial multivariate regression model, but intercourt conflict obtains marginal statistical

significance (p=.10) in the revised model summarized in Table 4.10. While surprising,

these findings are consistent with Flemming and Krutz' s conclusion that litigants

(including governments) seeking leave to appeal to the Supreme Court of Canada cite

conflicting lower court decisions very rarely.16

One possible conclusion from this evidence is that, although the federal

government considers jurisprudential conflict important, it does not track such legal

developments effectively. Another possible explanation is that the method 1 use of

measuring conflict-citation by the lower appeal court of precedents not followed-may underestimate the actual effect of this factor; that is, if a court simply ignored conflicting jurisprudence altogether, the case would be coded "0," even if a conflict existed and the

government knew it. A better measure for conflict would be one that was independent of the court rendering the decision; unfortunateIy, tracking precedents in so many courts for

so many legai issues is methodologically unfeasible at this time. Finally, the measure might be sound, and the federal government simply is 1101 very concerned with gaining access to the Court. That is, Ottawa' s excellent track record in securing leave to appeai may lead decision makers to conclude that the Court will usually grant leave requests, regardless of whether intercourt conflict exists.

D. Winnability

The resuIts for winnability are mixed, as the two measures-Iower court reversaI and judicial dissent-display markedly different levels of success. There is no statistical

16 Flemming and Krutz (2()()2), 238, 2·H.

105 support for the hypothesis that winning at trial (or on appeal below the higher Court of

Appeal) encourages the government to appeal to the SCC. Both the correlations and the

regression coefficients indicate that the relationship for lower court reversaI is the opposite

ofthat predicted, and is not significant in any case. Zorn's study of US government

lawyers also found that reversaI exerted an influence in the direction opposite of that

hypothesized, although in his case this effect was statistically significant. 17 My findings

regarding reversaI are likely explained by the fact that alone of the hypothesized factors,

reversaIs are quite common (69% of cases), while appeals are not. 18

In contrast, appeal court dissent is one of the strongest explanatory factors. It

does not correlate with other independent variables, and it remains strong in multiple regression, suggesting that judicial disagreement on the lower appeal court exercises an

significant influence on government decision makers. Notably, dissent is more significant for losses in the PCAs than in the FCA-in fact, it is not significant at all for the latter. 19

This confirms the hypothesis, set out in Chapter Three, that winnability will be a stronger factor when a loss would exp and the jurisdictionai application of an unfavourable precedent. Notwithstanding the findings regarding reversaIs, the dissent variable confirms the importance ofwinnability, although the positive prospect ofwinning on appeal is cIearly not "necessary." As weIl, the data strongly indicate that the government's appeal

17 Zorn (2002), 158-9. Zorn appears to misreport his findings regarding reversaI. According to his coding (which is used here), reversaIs should encourage appeals. He concludes that "at least one factor of each type [including winnability] is influential," but he finds no significant relationship for judicial dissent or ideology, and he does not acknowledge that, although statistically significant, the relationship for reversaI is negative (158). 18 Notably, Zorn (2002) also found that reversaIs were typical. 19 For PCA cases, dissent correlates with the appeal decision at r=.27, p < .01, and is the strongest factor in multivariate regression (coefficient (B)=1.96, p < .01). In FCA cases, the corresponding figures are r=.15, p=.28 (correlation) and B=.94, p=.26 (regression).

106 decisions are calculated.

Figure 4.3: Statistically Significant Factors in the Decision to Appeal

Judicial Amendment

Decision Lower Court Dissent to Appeal

Novel Interpretation + Intervener

To further illustrate the impact of individu al explanatory factors, 1 calculated the

probability that the federal government would appeal (y=I) given the presence of

particular independent variables. Probabilities were generated by running statistical

simulations, in which the actual factor values and coefficients were ca1culated while

holding the particular independent variable at the desired value (that is, "1" if simulating its presence, "0" if simulating its absence). Table 4.11 presents simulations focusing on the three most statistically significant independent variables from Table 4.1O-judicial amendment, the novel interpretation-intervener interaction term, and dissent on the lower appeal court-both individually and in conjunction with one another. The first column

lO7 describes the factors used in each simulation, and the second column the corresponding probability that the federal government will appeal; for comparison, 1 provide the actual

(observed) probability of appeals in the first row. The "net difference," in the third column, represents the change in the probability that the government will appeal between when the factor(s) is absent, and when it is present. As expected, aH three factors increase

Table 4.11: Simulated Probabilities of Appeal Decisions, by Explanatory Factors

CaseCharaderistiC(whenother factors ..... Probabili~yof NetDitTerel1ce in at adüalvalues) ... .. Appeal (P(y:::l» P(y=!) Actual cases .296 -- Judicial Amendment present .579 +.303 Judicial Amendment absent .276 Novel Interpretation*lntervener present .824 +.558 Novel Interpretation*lntervener absent .266 Dissent present .507 +.273 Dissent absent .234 Judicial Amendment + Novel .951 Interpretation *Intervener present +.710 Judicial Amendment + Novel Interpretation *Intervener absent .241 Judicial Amendment + Dissent present .800 +.586 Judicial Amendment + Dissent absent .214 Novel Interpretation*lntervener + Dissent .938 present +.737 Novel Interpretation*lntervener + Dissent absent .201 Judicial Amendment + Dissent + Novel .985 Interpretation *Intervener present +.809 Judicial Amendment + Dissent + Novel Interpretation*lntervener absent .176

108 the probability ofan appeal, the interaction term the most (+.558), thenjudicial

amendment (+.303) and dissent (+.273). As weIl, the factors appear to have a cumulative

effect. Cases containing two or three of these factors have considerably higher appeal

probabilities than those containing only one (or two), or than actual cases as a whole.

IV. Conclusion

The research presented here confirms the initial assumption of the dissertation, that

appeals by the Canadian federal government are the product, to a large extent, of

calculated decision making. In stark contrast to the intuition and anecdotal evidence that

governments routinely appeal aIl unfavourable rulings, the federal govemment is a highly

selective appellant. This is true even among cases where the lower court has remedied

legislation or crafted a new interpretive rule that is unfavourable to legislative authority

generaIly. The evidence clearly indicates that Ottawa usually appeals only the most

important cases-those involving novel interpretation and interveners-and those

decisions constituting the deepest incursions into legislative jurisdiction, namely, judicial amendments. The strong statistical results for the "lower court dissent" variable further

support the conclusion that govemment decision makers actively consider strategic factors

such as minimizing losses. The only category of explanatory factors not strongly

supported by the data was reviewability, as measured by the intercourt conflict variable.

However, this do es not necessarily mean that reviewability is of no import. Recall that the categories of factors are not watertight, and the measures for winnability and salience, in particular, may also help convince the Supreme Court of Canada to hear a case.

109 As note d, one especially interesting conclusion the evidence permits is that the

Justice Department actively defends the authority of the government, and,

correspondingly, appears to challenge attempts by the lower appellate courts to expand judicial power. The chapters that follow examine this dimension of government appeals

more closely. The next chapter provides greater qualitative detail ofthose cases involving

the two factors most associated with this issue, novel interpretation (in conjunction with

intervention) and judicial amendment. In addition, 1 consider those cases with unexpected

outcomes, such as los ses containing these factors which were not appealed. The

penultimate chapter considers the degree to which the government' s appeal decisions can

be interpreted as part of a "dialogue" with the judiciary over constitutional meaning.

110 Appendix 4.1: Converting Logits into Odds and Probabilities20

Binary logistic regression produces a logit and corresponding regression coefficients (a, b, c, etc.) for each explanatory factor (Xl, X2, X3, etc.) and the constant (a) for the equation:

Log(PI1-P) = a + aXI + bX2 + CX3 + ... nxn

where P is the probability that the dependent variable is present, or pey = 1).

This logit (or "logged odds"), while having the mathematical virtue of ranging from -00 to 21 +00, has little or no intuitive meaning. However the logit can be converted to a more intuitive odds-that is, the ratio of the probability that y = 1 to the probability that y :f. 1 (or (y = 0) when y is binary)-by taking the antilog of (i.e., exponentiating) the right side of the logit equation:

PI1-P = e(u + axl + bx2 + cx3 + ... nxn)

For example, if the odds of appealing to the sec were 3.5, this means that the probability of appealing (y = 1) is 3.5 times greater than the probability of not appealing (y = 0), given the constellation of explanatory factors.

While this information is useful and easy to grasp, fully understanding odds ultimately requires knowing the probabilities involved. In algebraic terms, one can convert odds into probability p (in this case, that y= 1), by the following formula: pey = 1) = e(u + axl + bx2 + cx3 +... IL"ill) 1 + e(u + axl + bx2 + cx3 +... nxn)

It is then possible to test the impact of individual independent variables upon the probability that y = 1, by substituting the actual regression coefficients (and constant) and the simulated value of the variable, and performing the algebraic computations above.

20 Scott Menard (2002) Applied Logistic Regression Analysis, 2nd ed., Quantitative Applications in the Social Sciences Series. Number 106 (Thou sand Oaks: Sage Publications), 13. Menard provides an excellent explanation of current methods associated with 10gistic regression; see also Tim Futing Liao (1997) Interpreting Probability Uodels: Logit. Probit. and Other Generalized Linear Mode/s, Quantitative Applications in the Social Sciences Series, Number 10 1 (Thousand Oaks: Sage Publications). 21 Fred C. Pampel (2000) Logistic Regression: A Primer, Quantitative Applications in the Social Sciences Series. Number 132 (Thou sand Oaks: Sage Publications), 20; James Jaccard (2001), Interaction EfJects in Logistic Regression, Quantitative Applications in the Social Sciences Series. No. 135 (Thousand Oaks: Sage Publications), 4.

111 Appendix 4.2: Government of Canada Losses in Charter Cases in the Penultimate Courts of Appeal

[Cases marked with an asterisk indicate that the govemment sought leave to the SCC].

1. Alford v. Canada (Attorney General), [1998] B.C.J. no. 2965 (B.C.C.A.). 2. Armadale Communications Ltd. v. Adjudicator (Immigration Aclj, [1991] 3 F.C. 242 (F.C.A.). 3. Anval v. Canada, [1988] 1 F.C. 107 (F.C.A.). 4. Bains v. Canada (lv/inister ofEmployment and Immigration), [1989] 3 F.C. 487 (F.C.A.). 5. Baron v. Canada, [1991] 1 FC. 688 (F.C.A.).* 6. Batchewana Indian Band v. Batchewana Indian Band and Attorney General ofCanada [sub nom Corbiere v. Canada (Department ofIndian A.fJairs and Northern Development], [1997] 1 F.C. 689 (F.C.A.).* 7. Belczowski v. Canada, [1992] 2 F.C. 440 (F.C.A.).* 8. Canada (Attorney General) v. Goguen [sub nom Goguen v. Shannon], [1989] 50 C.C.C. (3d) 45 (N.B. C.A.). 9. Canada (Attorney General) v. Hydro-Quebec, [1995]67 Q.A.C. 161 (Q.C.A.). * 10. Canada (Attorney General) v. Sander, [1994] 114 D.L.R. (4th) 455 (B.C.c.A.). Il. Canada (Human Rights Commission) v. Lane & Chief Electoral Officer ofCanada (A-1150-88), [1990] 2 F.C. 327 (F.C.A.). 12. Canada (Minisler ofEmploymenl and Immigration) v. Clnmg, [1993] 2 F.C. 42 (F.C.A.). 13. Canada v. Amway ofCanada Ltd., [1987] 2 F.C. 131 (F.C.A.). * 14. Canada v. O'Neill Motors Lld., [1998] 4 F.C. 180 (F.C.A.). 15. Canadian Counci/ ofChurches v. Canada, [1990] 2 F.C. 534 (F.C.A.). * 16. Canadian Newspapers Co. Ltd. v. Attorney General ofCanada, [1985]16 D.L.R. (4th) 642 (O.C.A.).* 17. CanadianOxy Chemicals Lld. v. Canada (Attorney General), [1997] 145 D.L.R. (4th) 427 (B.C.C.A.).* 18. Chiarelli v. Canada (Minister ofEmployment and Immigration), [1990] 2 F.C. 299 (F.C.A.). * 19. Chief Electoral Officer (Canada) v. Canadian Human Rights Commission, [1990] 2 F.C. 327 (A- 1155-88) (F.C.A.). 20. Committeefor the Commonwealth of Canada v. Canada, [1987] 2 F.C. 68 (F.C.A.).* 21. Cushnie v. Canada (Minister ofEmployment and Immigration), [1989] 54 D.L.R. (4th) 420 (Q.C.A.) 22. Davidson v. Canada (Solicitor General), [1989] 2 F.C. 341 (F.C.A.). 23. Del Zotto v. Canada, [1997] 3 F.C. 40 (FC.A.).* 2-J.. Donovan v. Canada, [2000] 188 D.L.R. (4th) 42 (F.C.A.). 25. Dulude v. Canada, [2000] 192 D.L.R. (4th) 714 (F.C.A.). 26. Dywidag Systems v. Zutphen Brothers Construction [sub nom The Queen v. Dywidag and Zutphen], [1987] 35 D.L.R. (4th) 433 (N.S.C.A.).* 27. El Zein v. The Queen, [1987] 29 C.c.c. (3d) 560 (Q.C.A.).* 28. Energy Probe v. Canada (Attorney General), [1989] 58 D.L.R. (4th) 513 (O.C.A.). 29. Clayton Group Ltd v. Minister ofNational Revenue, [1988] 2 F.C. 467 (F.C.A.). 30. Gernhart v. Can(1da, [2000] 181 D.L.R. (4th) 506 (F.C.A.). 31. Haig v. Canada (Chief Electoral Officer), [1992] 3 F.C. 611 (F.C.A.). 32. Haig v. Canada, [1992] 94 D.L.R. (4th) 1 (O.C.A.). 33. Howard v. Stony MOllntainInstitution, [1984] 2 F.C. 642 (FC.A.). * 3.. 1-, Imperial Tobacco Ltd. v. Canada (Attorney General), [1989] 59 D.L.R. (4th) 743 (Q.C.A.). 35. International Fundfor Animal Welfare v. Canada, [1989] 1 F.C. 335 (F.C.A.). 36. Jamieson v. Canada (Minister ofJustice), [1995] 93 c.C.C. (3d) 265 (Q.C.A.). * 37. Kaurv. Canada (Uinister ofEmployme11l and Immigration). [1990] 2 F.C. 209 (F.C.A.). 38. Kig01l'a v. Canada. [1990] 1 F.C. 80.J. (F.C.A.). 39. Lagiorgia v. Canada. [1987] 3 F.C. 28 (A-.J.56-85) (F.C.A.). *

112 40. Lavallee, Rackel and Heintz v. Canada (Attorney General), [2000] 184 D.L.R. (4th) 25 (AC.A). * 41. Law v. Solicitor General ofCanada and Jvlinister ofEmployment and Immigration, [1985] 1 F.C. 62 (F.C.A.). 42. Li v. Canada (Minister ofCitizenship and Immigration), [1997] 1 F.C. 235 (F.C.A.). 43. Luscher v. Deputy Minister, Revenue Canada, Customs & Excise, [1985] 1 F.C. 85 (F.C.A). 44. Mileva v. Canada (Minister ofEmployment and Immigration), [1991] 3 F.C. 398 (F.C.A.). 45. Minister ofNational Revenue v. Kruger Inc., [1984] 2 F.C. 535 (F.c.A.). 46. Mooring v. Canada (National Parole Board), [1995] 93 C.C.C. (3d) 415 (B.C.C.A). * 47. (D.J.) v. Canada (So/icitor-Genera/), [1996] 25 B.C.L.R. (3d) 389 (B.C.C.A). 48. Native Women's Association ofCanada v. Canada, [1992] 3 F.C. 192 (F.C.A.). * 49. Nie/sen v. Canada (Employment and Immigration Commission), [1997] 3 F.C. 920 (F.C.A.). * 50. Osborne v. Canada (freasury Board), [1988] 3 F.C. 219 (F.C.A.). * 51. Pacifie Press Ltd. v. Canada (Jvlinister ofEmployment and Immigration), [1991] 2 F.C. 327 (F.C.A.). 52. Pacifie Press Ltd. v. Canada (Minister ofEmployment and Immigration), [1990] 1 F.C. 419 (F.C.A.). 53. Perera v. Canada (R.), [1998] 3 F.C. 381, (F.C.A.). 54. R. v. Abolmolouk, [1987] 23 O.A.C. 144, (O.C.A). 55. R. v. Aeciavatti, [1993] 80 C.C.c. (3d) 109 (O.C.A.). 56. R. v. Alainga, [1985] NWTR 263 (N.W.T.C.A.). 57. R. v. Aseneios, [1987] 34 C.C.C. (3d) 168 (Q.C.A.). 58. R. v. Baylis, [1989] 43 C.C.C. (3d) 514 (S. C.A.). 59. R. v. Bergeron, [1999] 136 C.C.C. (3d) 327 (C.M.A.C.). 60. R. v. Bisko, [1998] 123 C.C.C. (3d) 432 (O.C.A.). 61. R. v. Blomert, [1993] 79 C.C.C. (3d) 541 (P.E.LC.A.). 62. R. v. Bohn, [2000] 145 C.C.C. (3d) 320 (B.C.C.A.). 63. R. v. Brownridge, [1999] B.C.C.A 27 (B.C.C.A.). 64. R. v. Calloechia and D'Angelo, [2000] 149 C.C.C. (3d) 215 (Q.C.A.). 65. R. v. Campbell, [1990] 53 C.C.C. (3d) 93 (P.E.LC.A.). 66. R. v. Carroll, [1983] 147 D.L.R. (3d) 92 (P.E.LC.A.). 67. R. v. Cheecham, [1990] 51 C.C.C. (3d) 498 (S. c.A.). 68. R. v. Collins and Collins, [1999] 133 C.C.C. (3d) 8 (Nf. C.A.). 69. R. v. Cook, [1983] 147 D.L.R. (3d) 687 (N.S.C.A.). 70. R. v. Cox, [1999] 170 D.L.R. (4th) 101 (N.B. C.A.). 71. R. v. Croueher, [1996] 112 c.c.c. (3d) 428 (Nf c.A.). 72. R. v. Dennis, Kubin, and Frank, [1985] (N.W.T.C.A.) D.L.R. (4th) 205 (N.W.T.C.A.). 73. R. v. Doyer, [1993] 77 C.c.c. (3d) 203 (Q.C.A.).* 74. R. v. Finta, [1992] 92 D.L.R. (4th) 1 (O.c.A.).* 75. R. v. Flewwelling, [1983] 63 N.S.R. (2d) 264 (N.S.C.A.). 76. R. v. Fridleifson, [1999] B.C.C.A. 351 (B.C.C.A.). 77. R. v. Fry, [2000] 142 C.C.C. (3d) 166 (Nf c.A.). 78. R. v. Gillingham, [2000] 149 C.C.C. (3d) 261 (B.C.C.A.). 79. R. v. Goldhart, [1996] 25 O.R. (3d) 72 (O.C.A.). * 80. R. v. Grant, [1992] 73 C.C.c. (3d) 315 (B.C.C.A.).* 81. R. v. Guiller, Anderson and Manninen, [1987] 65 CR (3d) xxx (O.c.A.). 82. R. v. Hayden, [1984] 3 D.L.R. (4th) 361 (M.c.A.).* 83. R. v. Higgins; R. v Beare, [1987] 34 C.C.C. (3d) 193 (S.C.A.).* 84. R. v. Hosie, [1996] 107 C.C.c. (3d) 385 (O.C.A.). 85. R. v. Ingebrigtson, [1991] 76 D.L.R. (4th) 481 (C.M.A.C.). 86. R. v. Innocente, [1992] 113 N.S.R. (2d) 256 (N.S.C.A.). 87. R. v. Ironeagle, [1989] 49 C.C.C. (3d) 339 (S.C.A.). 88. R. v. Jamieson, [1989] 48 C.C.C. (3d) 287 (N.S.C.A.). 89. R. v. Jewitt, [1983] 5 C.C.c. (3d) 234 (B.C.C.A.).* 90. R. v. Lamy, [1993] 80 C.C.C. (3d) 558 (M.C.A.). 91. R. \J. Landry, [1984] 2 D.L.R. (.tth) 518 (Q.C.A.). 92. R. v. LaPlante, [1988] 48 D.L.R. (4th) 615 (S.C.A.).

113 93. R. v. Lauda, [1999] 136 C.C.C. (3d) 358 (O. c.A.). 94. R. v. Lauzon, [1999] 129 C.c.c. (3d) 399 (C.M.AC.). 95. R. v. Laws, [1999] 165 D.L.R. (4th) 301 (O.C.A.). 96. R. v. Lerke, [1986] 25 D.L.R. (4th) 403 (A. c.A.). 97. R. v. Light [sub nomR. v. Denbigh], [1993] 78 C.C.C. (3d) 221 (B.C.C.A.). 98. R. v. MacDonald, [1989] 44 c.C.C. (3d) 134 (P.E.I.C.A.). 99. R. v. Marstar Trading International Inc., [1999] 138 C.C.C. (3d) 87 (O.C.A.). 100.R. v. McCormick and Lingenfelter, [1993] 81 C.C.C. (3d) 209 (B.C.c.A.). 101.R. v. Misra, [1987] 32 c.C.C. (3d) 97 (S.C.A.). 102.R. v. Monney, [1998] 153 D.L.R. (4th) 617 (O.C.A).* 103.R. v. Monloute, [1991] 62 C.C.C. (3d) 481 (A.C.A.). 104.R. v. Nielsen, [1989] 43 C.C.C. (3d) 548 (S.C.A.). 105.R. v. Noble, [1985] (N.W.T.C.A.) D.L.R. (4th) 216 (O.C.A). 106.R. v. Oakes, [1983] 145 D.L.R. (3d) 123 (O.C.A.).* 107.R. v. O'Day, [1983] 148 D.L.R. (3d) 371 (N.B. C.A.). 108.R. v. Parker, [2000] 188 D.L.R. (4th) 385 (O. c.A.). 109.R. v. Pastro, [1988] 42 C.C.c. (3d) 485 (S. C.A.). 11O.R. v. Pellerin, [1999] 132 C.C.C. (3d) 434 (N.B.C.A.). l11.R. v. Peterson, [1994] 50 B.C.A.C. 24 (B.C.C.A.).* 112.R. v. Pippin, [1994] 4 W.W.R. 1 (S.C.A.). 113.R. v. Poirier, [1997] 116 C.C.C. (3d) 551 (Q.C.A.). 114.R. v. Polashek, [1999] 172 D.L.R. (4th) 350 (O.C.A.). 115.R. v. Randall and Weir, [1984] 1 D.L.R. (4th) 722 (N.S.C.A.). 116.R. v. Rao, [1984] 9 D.L.R. (4th) 542 (O.C.A.).* 117.R. v. Richard, [1995] 99 C.C.C. (3d) 441 (N.S.C.A).* 118.R. v. Ricketts, [2000] 144 C.C.C. (3d) 152 (O. C.A.). 119.R. v. Rodenbush and Rodenbush, [1985] 21 c.C.C. (3d) 423 (B.C.C.A.). 120.R. v. Rowbowtham, [1988] 41 C.C.C. (3d) 1 (O. c.A.). 121.R. v. Ruzic, [1998] 128 C.C.C. (3d) 97 (O.C.A).* 122.R. v. Silverslrone, [1991] 66 C.C.C. (3d) 125 (B.C.C.A.). 123.R. v. Silvini, [1992] 68 C.C.c. (3d) 251 (0.c.A.). 124.R. v. Smith, [1991] 108 N.S.R. (2d) 311 (N.S.C.A.). 125.R. v. Stanger (R. v. Bramwell; R. v. Kerr; R. v. Leskosek), [1984] 2 D.L.R. (4th) 121 (AC.A). 126.R. v. Stensrud, [1988] 42 C.C.c. (3d) 569 (S.C.A.).* 127.R. v. Stevens, [1984] 1 D.L.R. (4th) 465 (N.S.C.A.). 128.R. v. Stock, [1984] 10 C.C.C. (3d) 319 (B.C.C.A). 129.R. v. Sutherland, [2000] 150 C.C.C. (3d) 231 (O.C.A.). 130.R. v. Tanner, [1989] 46 C.C.C. (3d) 513 (A.C.A.). 131.R. v. Turcotte, [1988] 39 C.C.C. (3d) 193 (S.C.A). 132.R. v. Valiquette; R. v. Rioux, [1988] 62 C.R. (3d) 363 (Q.C.A). 133.R. v. Viger, [1983] 132 A.P.R. 10 (P.E.I.C.A). 134.R. v. Walsh, [1983] N.S.l. no. 208 (N.S.C.A.). 135.R. v. Wholesale Travel Group Inc., [1990] 63 D.L.R. (4th) 325 (O.C.A). * 136.R. v. Williams, [1988] 38 C.C.C. (3d) 319 (Y.T.C.A). 137.R. v. Young, [1993] 79 C.C.c. (3d) 559 (O. C.A.). 138.R. v. Zammit, [1993] 81 c.e.e. (3d) 112 (o.e.A.). 139.Rosenberg v. Canada (Attorney General), [1998] 158 D.L.R. (4th) 664 (O.C.A). l-lO.Rothmans, Benson & Hedges v. Canada (Attorney General) (A-301-89), [1990] 1 F.C. 90 (F.C.A). l-ll.Ruby v. Canada (So/icitor-General), [2000] 3 F.C. 589 (F.c.A). * 142.Sauvé v. Canada (Attorney General), [1992] 89 D.L.R. (4th) 644 (O.C.A.).* 1-l3.Schachter v. Canada, [1990] 2 F.C. 129 (F.C.A.).* IH.Schreiber v. Canada (.4ttorney General), [1997] 2 F.C. 176 (F.C.A.). * 1-l5.Sivakumar v. Canada (Minisrer ojCiti=enship & Immigration), [1996] 2 F.C. 872 (F.C.A.).

114 146.Skis Rossignol Canada v. HunIer and Attorney General of Canada [cited as Lagiorgia v. Canada], [1987] 3 F.C. 28 (A-245-85) (F.C.A). * 147.Somerville v. Canada (Attorney General), [1996] 136 D.L.R. (4th) 205 (AC.A). 148.Southam Inc. v. Hunter. et al., [1983] 147 D.L.R. (3d) 420 (AC.A).* . 149.Steele v. Mountain Institution, [1990] 54 c.c.c. (3d) 334 (B.C.C.A).* IS0.St-Onge v. Canada (Commissioner of Official Languages), [1992] 93 D.L.R. (4th) 649 (F.C.A). ISl.Suresh v. Canada (lvfinister of Citizenship & Immigration), [1998] 4 F. C. 192 (F. C.A). 152. Tétrault-Gadoury v. Canada (Employment and Immigration Commission), [1989] 2 F.C. 245 (F.C.A.).* 153. Thibaudeau v. Minister ofNational Revenue, [1994] 2 F.C. 189 (F.C.A). * 154.~vler v. Minister ofNational Revenue, [1991] 2 F.C. 68 (F.C.A). 155.v. Cazzetta, [1996] 108 C.C.C. (3d) 536 (Q.C.A.).* 156. Vanguard Coatings and Chemicals Ltd v. Minister ofNational Revenue, [1988] 3 F.C. 560 (F.C.A). 157. Vespoli v. The Queen, [1984] 55 N.R. 269 (F.C.A). . 158. White. Ottenheimer and Baker v. Canada (Attorney General), [2000] 187 D.L.R. (4th) 581 (Nf.C.A.).* 159. Williams v. Canada (Regional Transfer Board), [1993] 1 F.C. 710 (F.C.A).

115 CHAPTER FIVE

Factors in the Decision to Appeal: A Second Look

Statistical analysis of the federal government' s appeal decisions supports the hypothesis that the Justice Department chooses its cases in a rational and relatively predictable manner. Much remains to be explained, however. What, for example, is really at stake in cases involving novel interpretations and judicial amendments?

Although the numbers strongly indicate the importance of the novel interpretation in conjunction with an intervener, is it because of the presence of allies in court, or opponents, or neither? Are there particular types of judicial amendment that are provoking a response, or is there sorne other discernible pattern? This chapter attempts to answer these questions by examining in greater detail three categories of cases: (i) those with both novel interpretation and intervention; (ii) those witnessing a judicial amendment; and (iii) those with unexpected outcomes.\ The first two categories de al with the most statistically significant explanatory factors (the nature of judicial dissent is largely unambiguous, and therefore not discussed here), while the last category attempts to expIai n, on the one hand, government appeals where the reason is not immediately obvious, and, on the other, government concessions (not appealing) in the face of apparently compelling reasons to appeal.

A final note before proceeding: ideally, this qualitative analysis would have inc1uded feedback from the Justice Department lawyers who handled the cases discussed. Unfortunately, it became clear early in this project that concems about solicitor-client privilege and Cabinet secrecy prevented current and former govemment lawyers from commenting on individual cases. As such, my analysis relies on the courts' decisions, govemment facta, and secondary sources.

116 I. Novel Interpretations with Interveners

The strongest indicator that the federal government will appeal a loss in a Charter case to the Supreme Court of Canada (SCC) is the joint presence of a novel interpretation and an intervener. Often such cases, ni ne were appealed: Southam v. Hunter, R. v.

Oakes, R. v. Rao, Canadian Newspapers v. A.G. Canada, Schachter v. Canada, Native

Women 's Association ofCanada v. Canada, Thibaudeau v. Minister ofNational Revenue,

Corbiere v. Canada (Minister ofIndian and Northern Affairs) [sub nom Batchewana

Indian Band (non-residents) v. Batchewana Indian Band], and Nielsen v. Canada. 2 Haig v. Canada3 was the sole exception. 1 discuss each case in turn, with special attention to the nature of the novel interpretation, the identity and allegiance of the intervener, and, where applicable, the government' s argument on appeal to the Supreme Court.

The first case in this group, chronologically speaking, was the Alberta Court of

Appeal's 1983 ruling in Southam v. Hunter, et al. The central issue was whether warrantless searches, as part of a combines investigation, violated the Charter's section 8 right against unreasonable search and seizure. Section 10 of the Combines Investigation

Act empowered the Director of Investigations and Research to authorize searches without a warrant from ajudge or magistrate. In one of the earliest cases dealing with s.8, the

Court of Appeal unanimously decided in favour of the c1aimant, Southam, and invalidated s.l 0 of the Act. In so ruling, the court established that searches must be

2 [1983] 147 D.L.R. (3) 420 (A.C.A.); [1983] 145 D.L.R. (3d) 123 (o. c.A.); [1984] 9 D.L.R. (4th) 542 (O.C.A.); [1985] 16 D.L.R. (4th) 642 (O.C.A.); [1990] 2 F.C. 129 (F.C.A.); [1990] 2 F.C. 129 (F.C.A.); [1990] 2 F.C. 129 (F.C.A.); [1992] 3 F.C. 192 (F.C.A.); [1994] 2 F.C. 189 (F.C.A.); [1997] 1 F.C. 689 (F.C.A.); [1997] 3 F.C. 920 (F.C.A.). 3 [1992] 94 D.L.R. (4th) 1 (O.C.A.).

Il ï authorized by a qualified, impartial arbiter to pass constitutional muster under s. 8, a ruling with widespread implications for criminal investigations. On appeal to the

Supreme Court, which upheld the Alberta Court of Appeal, the federal government argued unsuccessfully that the Director was a neutral and impartial arbiter, despite the fact that he or she is part of the investigating team that decides whether to prosecute.

Section 10 of the Act could be saved, Ottawa insisted, by procedural safeguards ensuring the Director's impartiality, which could be "read in." Appearing as an intervener, the

Attorney General of Alberta's defense of the Act was even more forceful, contending that there was no primafacie Charter violation. Alberta argued that s.8 only prohibits

"unreasonable search or seizure," and in this instance, the agents' conduct was reasonable, as it was authorized by the proper authorities, was strictly limited in scope and time, no force was used, the accused were duly notified and given an opportunity to consult with a lawyer, and the materials seized could be recovered.

The government's next loss in this category came in R. v. Oakes, from the Ontario

Court of Appeal. Although Oakes is now primarily associated with the Supreme Court's elaboration of the section l "reasonable limits" test, the lower court case focused on section Il (d)' s presumption of innocence and the "reverse onus" provision regarding trafficking in the Narcotics Control Act. Under section 8 of the Act, if one was found to be in possession of a narcotic, she must also be convicted of trafficking unless she can establish otherwise. For the first time at the penultimate court of appeallevel, the Ontario

Court of Appeal held that this "reverse onus" provision violated the accused' s ne ar absolute right to be presumed innocent, and struck down section 8, sending a clear message that other reverse onus offenses were vul~erable to a Charter challenge. The

118 sole intervener in the case, the Canadian Civil Liberties Association, sided with Oakes, arguing "that it would 'make nonsense' of the presumption of innocence if Parliament could shift to the accused the burden of proof of the essential element of the offense on proofby the Crown of an insignificant fact [i.e., mere possession of a narcotic]." On appeal, Ottawa urged the Court to adhere to the Bill ofRights, 1960 jurisprudence regarding reverse onus, specifically, the decisions in R. v. Sharpe4 and R. v. Applebl permitting "the existence of statutory exceptions reversing the onus of proof with respect to one or more ingredients of an offense in cases where certain specific facts [i.e., possession] have been proved by the Crown in relation to such ingredients,"6 provided "it is rationally open to the accused to prove or disprove" his or her intent. In affirming the lower appeal court, the Supreme Court, as it usually has since 1982, rejected its earlier

Bill of Rights ruling in favour of a more expansive interpretation of rights.

R. v. Rao concemed the constitutionality of warrantless searches of places other than dwelling-houses by police during narcotics investigations. Section 10(1)(a) of the

Narcotic Control Act required a warrant or writ of assistance for searches of dwelling- houses (homes), but did not for other places, such as offices and vehicles. Unlike Oakes, in which reverse onus provisions were declared prima fade unreasonable and therefore inconsistent with the section 8 of the Charter, the Ontario Court of Appeal held in Rao that warrantless searches "are not on their face necessarily unreasonable and do not necessarily coll ide with the Charter."7 Instead ofnullifying s.10(1)(a), the court read in

4 (1961), 131 C.C.C. 75 (O.C.A.). 5 [1972] S.C.R. 303 (S.C.C.). 6 R. . v Appleby [1972] S.C.R. 303, at 315-16 per Ritchie J. 7 R. v. Rao [1984] 9 D.L.R. (4th) 542 at 571 per Martin J. (O.C.A.)

119 the conditions that "s.l 0(1)( a) is inoperative to the extent that it authorizes the search of a person's office without a warrant, in the absence ofcircumstances which make obtaining a warrant impracticable."g The "novel interpretation" component ofthe ruling was the court's finding that searches of places other than homes would usually require a warrant, despite the accused's lower expectation ofprivacy in such places; this conclusion was supporte d, not surprisingly, by the sole intervener, the Canadian Civil Liberties

Association.9 The federal government sought leave to appeal this ruling, but the Supreme

Court refused without providing reasons. However, the government's argument before the Ontario Court of Appeal hints at its rationale for appealing. As the Court of Appeal noted,

Mr. Dambrot, on behalf of the Crown, in support of his submission that warrantless searches [where the investigating officer has reasonable grounds to search the premises] have gained common acceptance in Canada, referred us to a number offederal statutes which confer power on designated officiaIs to enter, search, inspect or audit at business premises without a warrant. 10

The government cited twelve statutes, covering such diverse policy are as as food and drugs, explosives, textiles, the Canada Pension Plan, income tax, and Unemployment

Insurance. Given the widespread impact of a decision effectively outlawing warrantless

8 R. v. Rao [1984] 9 D.L.R. (4th) 542 at 571 per Martin J [emphasis added]. (O.c.A.) 9 Interestingly, the O.C.A. did not cite the Alberta Court of Appeal's decision in Southam v. Hunter only six months earlier, which, while not dealing with an identical issue, similarly dealt with the matter of properly authorized searches. This, in conjunction with the federal government's argument that warrantless searches "have gained corn mon acceptance in Canada" (see below), suggest the hearing and opinion-writing in Rao pre-dated the Southam ruling (unfortunately, the law reports ofthese cases did not provide hearing dates. only ruling dates). 10 R. v. Rao [1984] 9 D.L.R. (4th) 542 at 557-558 per Martin 1. [emphasis added]. (O.C.A.)

120 searches, the government' s desire to appeal is not surprising. More will be said of Rao near the end of the chapter.

In Canadian Newspapers v. A. G. Canada, the government defended section

442(3) of the Criminal Code, which protected the identity of sexual assault complainants, from a freedom of expression (section 2(b)) challenge brought by a newspaper publisher.

Rooted in a desire to encourage victims of sexual assault to come forward, section 442(3) required the trial judge to issue a publication ban on the identity of the complainant if she or the prosecution so requested. Although the Ontario Court of Appeal ruled that the ban constituted a prima facie infringement of freedom of the press, it accepted the rationale behind the provision in the context of section 1. However, the court ultimately rejected section 442(3)'s denial ofjudicial discretion over issuing the publication ban as

"umeasonable," and severed the mandatory aspect of the provision. In short, the court amended the section rather than invalidate it, and in so doing, crafted a new discretionary rule to replace a categorical rule. As with all changes from statutory or common law categorical rules to discretionary rules, the effect is to enhance the influence of judges over legal interpretation, while decreasing the influence of the legislative branches. Il

Ottawa was supported in Canadian Newspapers by the Attorney General of

Ontario, who handled the prosecution of the initial criminal case, but appeared as an intervener in the Charter-based civil action brought against the federal government. In its

II Other examples include the Supreme Court ofCanada's high-profile rulings in R. v. Seaboyer [1991] 2 S.C.R. 577, R. v. Daviau/t [1994] 3 S.C.R. 93, R. v. O'Connor [1995] 4 S.C.R. 411, and R. v. Carosella[1997] 142 D.L.R. (4th) 595, ail conceming sexual assault. See Christopher P. Manfredi and Scott Lemieux (1999) "Judicial Discretion and Fundamental Justice: Sexual Assault in the Supreme Court of Canada," TIJ.~ American Journal of Comparative Law 47: 489-514.

1:21 appeal to the SCC, Ottawa conceded the Charter violation but reiterated the need for a

mandatory publication ban; the Court unanimously agreed, concluding "a discretionary

provision under which a judge would retain the power to decide whether to grant or

refuse the ban on publication would not be effective in attaining Parliament's pressing

goal, since it would deprive the complainants of the assurance that, upon request, their

identity would not be disclosed, an important factor in their decision to report the

crime.,,12

One of the government's most high-profile and important losses came in R. v.

Schachter. Schachter concemed whether the then-Unemployment Insurance Act's refusaI of "matemity" benefits to natural fathers but not adoptive fathers violated Mr.

Schachter's section 15(1) equality rights under the Charter. Section 32 of the Act allowed adoptive parents of either sex 15 weeks of benefits, to be divided as the parents wished, but denied natural fathers any benefits except in the case of the mother' s death or disability. Schachter's claim rested on the argument that "natural fathers" were an analogous ground under s.15(1), which the Federal Court of Appeal (FCA) accepted; this was the first of two novel interpretations issued by the court. 13

As the federal government conceded the rights violation at the FCA, the case focused on the appropriate remedy. At trial, the judge invoked the Charter's section

24( 1) remedial clause to extend benefits to natural parents so as to mirror those for

12 Canadian Newspapers v. Canada (Attorney General) [1988] 2 S.C.R. 122 at 124 per Lamer 1. (S.C.C.) 13 On appeal to the SCC, Schachter added a sexual discrimination cIaim, based on the distinction drawn between natural mothers and tàthers-a cIaim curiously absent in the Federal Court of Appeal case, which was only raised in passing by Mahoney 1. in dissent (R. v. Schachter [1990] 2 F.C. 129 at 153 CF. C.A.)).

122 adoptive parents. 14 Ottawa submitted on appeal to the FCA that "where a law is found to be inconsistent with the provisions of the Constitution the only recourse open to a court is to dec1are it of no force or effect pursuant to subsection 52(1) of the Constitution Act,

1982.,,15 When complex social policy is in question, Ottawa maintained that it is up to

Parliament to consider what-if any-replacement legislation ought to be enacted.

Furthermore, the AG Canada insisted that section 24(1) did not displace Parliament's constitutional authority over the public purse, and judges could not, therefore, extend financial benefits associated with legislated public policy.

The feminist advocacy organization LEAF (W omen' s Legal Education and

Action Fund) intervened in support of Schachter and an activist approach to s.24(1 ).16

LEAF' s counsel argued that "while a dec1aration of invalidity in this case would place natural and adoptive parents in the same position," such a result would amount to

"equality with a vengeance" because of the punitive aspect of the result:

Ensuring that groups or individuals have the same entitlement to no benefits is contrary to the purpose of the equality guarantee in section 15 and, produces only sameness, not equality ... it [is] not possible to achieve the equality envisioned by section 15 simply by taking a benefit away from others. 17

14 Section 24(1) reads: "Anyone whose rights or freedoms, as guaranteed by this Charter, have been infringed or denied may apply to a court of competent jurisdiction to obtain such remedy as the court considers appropriate and just the circumstances." 15 R. v. Schachter [1990] 2 F.C. 129 at 130 per Heald J. (F.C.A.) 16 Although LEAF's official status was "intervenor" in Schachter, they were granted the unprecedented privileges of leave to cali and cross-examine witnesses (Sherene Razack (1991) Canadian Feminism and the Law (Second Story Press), 66. 17 Cited in R. v. Schachter [1990] 2 F.C. 129 at 146 per Heald 1. (F.C.A.)

123 The FCA sided with LEAF and Schachter, in the process enunciating a judieially aetivist

novel interpretation of the remedial power in s.24( 1) that permitted "reading in" and judicial extension ofbenefits.

On appeal to the Supreme Court, the federal government reiterated its arguments

before the FCA, that s.24(1) did not permit "reading in"; the Court disagreed, but chose

not to invoke this power in Schachter's case. That this appeal was motivated primarily

by micro-constitutional strategy is evidenced by the fact that Parliament had already

amended the Unemployment Insurance Act to grant all new parents benefits, but reduced

the eoverage period to ten weeks. Ottawa's goal on appeal, then, was not to proteet a

partieular social poliey but to challenge the FCA's legal interpretation regarding remedy

that could erode the poliey authority of the elected government in virtually any area

vulnerable to a Charter daim.

In Haig v. Canada, the Ontario Court of Appeal ruled that section 3(1) of the

Canadian Human Rights Act was inconsistent with the Charter' s equality rights, as it did not list sexual orientation as a protected ground. This was the first time a Court of

Appeal explieitly identified sexual orientation as an analogous ground under s.15. The federal government actually eonceded this important novel interpretation, but contended that there was no s.15 violation in this case: "in failing to indude [sexual orientation] as a proscribed basis of discrimination in s. 3(1) of the Canadian Human Rights Act,

Parliament has not itse1f acted in a discriminatory fashion." 18 Furthermore, Ottawa

\8 Haig v. Canada [1992] 94 D.L.R. (4th) 1 ut \8 per Krever J. (O.C.A.)

124 argued that it "could have chosen not to have a Human Rights Act and, having chosen to enact one, was free to legislate in respect of sorne social problems and not others.,,19

The Court of Appeal rejected these arguments, and because the federal government had, strangely, offered no section 1 defense, tumed immediately to the question ofremedy. Rather than striking down the law, the court "read in" sexual orientation to s.3(1). They did so with the encouragement of the Canadian Human Rights

Commission, which intervened to ask the court to arder the Commission to deal with complaints on the basis of sexual orientation. In a surprising move, then-Justice Minister

Kim Campbell opted not to appeal to the Supreme Court, the only instance of concession in this category of cases. More will be said ofthis decision in Section III below.

The government's next loss in this category came during the midst of the constitutional negotiations regarding the 1992 Charlottetown Accord, which included provisions for aboriginal self-government. The Native Women's Association of Canada

(NWAC) charged that the major aboriginal organizations involved in constitutional negotiations-the Assembly of First Nations (AFN), the Native Council of Canada

(NCC), the Métis National Council (MTC), and the Inuit Tapirisat of Canada (ITC)-did not represent the interests of aboriginal women. Specifically, women were numerically under-represented in the executives of these groups, and NWAC criticized the organizations' perceived failure to address the particular problems faced by aboriginal women. NWAC also strongly opposed the position ofthese groups that aboriginal self government should be exempt from the Charter, given past sexual discrimination within

19 Haigv. Canada [1992] 94 D.L.R. (4th) 1 at 18 per Kre.v.:r 1. (O.C.A.)

125 bands, as evidenced by the Lavell and Bédard case.20 In short, NWAC contended that aboriginal women were likely to suffer a loss of rights if the position taken by the

Assembly of First Nations and others prevailed.

In Native Women 's Association of Canada v. Canada, NWAC convinced the

Federal Court of Appeal that the federal government's under-funding ofNWAC (5% of a total $10 million to aboriginal groups, with most going to the AFN, NCC, MNC, and

ITC), violated NWAC' s freedom of expression in section 2(b) and offended the guarantee of sexual equality in section 28 of the Charter. The court found that "by inviting and funding the participation of those organizations in the CUITent constitutional review process and excluding the equal participation of NWAC, the Canadian government has accorded the advocates of male-dominated aboriginal self-governments a preferred position in the exercise of an expressive activity.,,21 In other words, the FCA asserted that courts can determine whether government funding of interest groups is discriminatory under the Charter, and, more profoundly, that a failure to fund a group which advocates values consistent with the Charter violates freedom of expression. This conclusion was opposed by the aboriginal groups criticized by NWAC, who intervened (with the exception of the AFN) in support of the federal government, although it is inaccurate to describe them as true "third parties": they had a direct stake in the decision, since NWAC

20 In Attorney General ofCanada v. Lavell and Bédard [1974] S.C.R. 1349, two aboriginal women unsuccessfully challenged the Indian Act provision that Status rndian women who married non-Status men lost their Status, while Status Indian men did not in the reverse situation, as a violation of equality rights under the statutory 1960 Bill of Rights. They were opposed by several aboriginal organizations, including the National Indian Brotherhood and Indian associations from every province and territory (save Newfoundland). The case is now regarded as one of the Supreme Court's more infamous decisions, and is credited with spurring aboriginal (and non-native) women's support for a constitutionally-entrenched Charter of Rights. 21 Native Women's Association of Canada v. Canada [1992] 3 F.C. 192 at 200 per Mahoney J. (F.C.A.)

126 had petitioned the court to freeze federal funding to these groups until its own funding was increased. The FCA denied this request, but ordered the government to increase funding to NWAC by an unspecified amount.

The government won on appeal to the Supreme Court, but its technical arguments, criticizing the FCA's remedy when NWAC had only asked to prohibit funding to other groups, were not those adopted by the Court. Rather, the majority of the Court rejected the FCA's interpretation that s.2(b) of the Charter provided a positive right to publicly- funded expression, and found no infringement by the government ofNWAC's freedom of expreSSIOn:

There is no question here of the Government of Canada attempting to suppress NW AC's expression of its point of view with respect to the Constitution. The s. 2(b) argument advanced is dependent on a finding that the funding of and participation by NWAC were essential to provide an equal voice for the rights of women. A corollary to this submission is that the funded groups are not representative of Native women because they advocate a male-dominated aboriginal self-government. This is the submission that was accepted by the Court of Appeal and is the foundation of its judgment.22

However, the Court found

no evidence to support the contention that the funded groups were less representative of the viewpoint ofwomen with respect to the Constitution. Nor was there any evidence with respect to the level of support ofNWAC by women as compared to the funded groups. As well, the evidence does not support the contention that the funded groups advocate a male-dominated form of self government."3

12 Native Women 's Association of Canada v. Canada [1994] 3 S.C.R. 627 at 657 per Sopinka 1. (S.C.C.) . 23 Native Women 's Association of Canada v. Canada [1994] 3 S.C.R. 627 at 657-8 per Sopinka 1. (s.c.e.).

127 Ottawa a1so successfully challenged a 10ss in the FCA in the high-profile

Thibaudeau v. Minister a/National Revenue. The FCA nullified section 56(1)(b) of the

Incarne Tax Act, which required custodia1 parents to pay income tax on chi1d support

payments received from the other parent. A corollary section of the ITA. (s.60(b)),

allowed the parent making support payments to deduct them from his or her taxable

income. The intent of this "inclusionldeduction" scheme was to increase disposable

income for "post-divorce fami1y units" (specifically, to combat child poverty), by

transferring the tax burden to the parent with the lower marginal tax rate. However,

Suzanne Thibaudeau, like 30% of custodial parents, was actually economically worse off

under the system. She argued, with the support of the intervener SCOPE (Support and

Custody Orders for Priority Enforcement),24 that because many custodial parents were

disadvantaged by the scheme, andthat most were women, s.56(1)(b) discriminated on the

basis of sex, contrary to s.15(1) of the Charter. The FCA actually rejected this argument, on the grounds that the system disadvantaged sorne men in exactly the same way; while it acknowledged that men were disadvantaged far less frequently than women, it stressed the focus should be on the "nature of effect," not numbers. Rather, the FCA crafted a new analogous ground under s.15, that of "fami1y status," arguing that separated custodial parents are a discrete and insular minority historically suffering prejudice and in need of protection.

In its appeal to the Supreme Court, the federal government argued that situations such as Thibaudeau' s were anticipated by other parts of the ITA, including provisions instructing family courts to "gross-up" support payments to offset ta'\: disadvantages, and

24 Whose counsel, Mary-Eberts, was a founding and active member of LEAF.

128 tax credits for custodial parents. The problem, therefore, was not one of legislative design, but of execution by the family courts. The Court agreed, holding that, overall, the inclusionldeduction system was intended to-and did-work to the advantage of most custodial parents, and people such as Thibaudeau could seek redress in the family court.

A majority of the court concluded that separated custodial parents were therefore not discriminated against by s.56(1 )(b), but did not address whether the category could survive as an analogous ground in other circumstances. The court's female members dissented, explicitly accepting the new analogous ground. Notably, despite winning in the Supreme Court, Ottawa scrapped the inclusionldeduction system in 1997, replacing it with a system oftax credits.25 This suggests Ottawa appealed to protect its legislative discretion, especially in an area as complex and laden with potentially-disadvantageous distinctions as the Incarne Tax Act.

ln Corbiere v. Canada, the government defended section 77(1) of the lndian Act, which required that band members be "ordinarily resident on the reserve" to vote in Band

Council elections, from an equality rights challenge brought by several non-resident members of the Batchewana Indian Band. The government was supported by LSLIRC

(the Lesser Slave Lake Indian Regional Council), one of the three interveners, with the

Native Women's Association of Canada and the Congress of Aboriginal Peoples supporting the non-resident members. The Federal Court of Appeal ruled in favour of the non-residents, in the process recognizing "residency off-reserve" as a new analogous

25 See Brent Kraus (1996) "Disentangling the Tax Unit: the Subsidization ofChild Support after Thibaudeau," Saskatchewan Law Review 60: 295-314, and Mélanie Beaulieu (2001) "Traitement Fiscal Des Pensions Alimentaires Pour Enfants: Les Modifications De 1997 Étaient-Elles Une Réponse Adéquate à L'arrêt Thibaudeau?," University ofT.o.!:onto Facu/ty of Law Review 59: 319-347.

129 ground. The court displayed confusion, however, with respect to the remedy, both

severing from s. 77(1) the phrase regarding residency and, redundantly, granting the claimants a constitutional exemption from the residency requirement. Only the latter was appropriate given the court's des ire to limit the ruling to the Batchewana Band.

The federal government' s appeal to the Supreme Court relied on section 1, that the restriction was designed "to give voice in affairs ofthe reserve only to the persons most directly affected by the band council," and "the administrative difficulties and costs involved in setting up, for example, a two-tiered council where one tier would deal with local issues and the other with issues affecting aU band members, or in maintaining a voter's list and conducting eleetions where the eleetorate may be widely dispersed."26

While the Court aecepted the rationale behind s. 77(1), it rejected the means used, criticizing the government for not attempting, or even considering in its factum, any less- restrictive alternative.27 The Court did, however, grant the government and band's request to vary the FCA's remedy, by reading out the residency requirement for ail bands but suspending this declaration for 18 months. This is somewhat curious, as the government, in effeet, requested a more intrus ive remedy-severanee-than eonstitutional exemption. Possibly, this refleeted the government's desire to reform Band governance under the lndian Act, as it recently announeed with Bill C-61, the proposed

First Nations Act. Even more curious, in this context, is the Batchewana Band's support for the same remedy preferred by Ottawa.

26 Corbiere v. Canada [1999] 2 S.C.R. 203 at paragraphs 21 and 104 per L'Heureux-Dubé J. 27 Notably, the SCC refined the analogous ground to the hybrid category of "Aboriginality-residence," but this did not differ substantially from the FCA' s analysis.

130 In 1989, Carol Nielsen, a federal public servant, filed a complaint with the

Canadian Human Rights Commission after the government refused to pro vide dental coverage to her same-sex partner. Her case was held in abeyance while the courts considered whether the Canadian Human Rights Act protected sexual orientation, either on its own terms (Haig v. Canada) or in the context of "family status" (Canada (Attorney

General) v. MOSSOp28). As noted above, the Ontario Court of Appeal's 1992 ruling in

Haig extended Charter equality rights protection to sexual orientation. The issue facing the Federal Court of Appeal in Nielsen v. Canada, then, was whether Haig could be applied to a complaint from 1989. Over the objections of the federal government and the intervenant Canadian Human Rights Commission, the FCA rendered the novel interpretation that Haig-and by extension, other "read in" remedies-were retroactive:

the basis of the judgment is that the provision must be understood, interpreted and read as the legislature would have written it, or rewritten it upon the advent of the Charter, had it been properly informed as to the limits of its powers. This does not allow any other conclusion than that the Court me ans to go back to the time of enactment or the coming into force of the Charter in 1982. The "reading in" technique is retro active , while an amendment by a legislature is, in principle, prospective only.29

The government's attempt to appeal this decision was refused by the Supreme Court.

Statistical analysis clearly indicates that novel interpretations combined with the presence of an intervener provoke a federal government appeal. But why? The cases exhibit no immediately obvious patterns that might explain the government' s tende ne y to appeal. The cases are not concentrated around a particular time or era (they span from

28 [1991] 1 F.C. 18 (F.C.A.); affd [1993] 1 S.C.R. 554 (S.C.R.). 29 Nie/sen v. Canada [1997] 3 Fe 920 at 922 per Marceau 1.

131 1983 to 1997), and although they all involve either freedom of expression (s.2(b )), unreasonable search and seizure (s.8), or equality rights (s.15), this is hardly surprising given the numerical preponderance of such cases in Charter litigation as a whole. One intuitive factor is the direction of intervener allegiance; that is, are the interveners supporting or opposing the government? There was no clear pattern on this issue, however: of ni ne appeals, Ottawa was supported by an intervener in five (Southam v.

Hunter, Canadian Newspapers, Corbiere, Nie/sen and NWAC), and opposed in five

(Corbiere, Oakes, Rao, Schachter, and Thibaudeau), with Corbiere witnessing both.

There was sorne indication, however, that the type ofintervener matters: all of the supporting interveners were governmental actors (AO Ontario, Canadian Human Rights

Commission (CHRC), and aboriginal organizations composed of Band representatives), while all but one of the opposing interveners were non-governmental organizations

(LEAF, SCOPE, CCLA, NWAC, and CAP; the exception was the CHRC in Oakes). To rephrase, the government appealed novel interpretations when allied with other governments, or when opposed by NOOs. In Haig, the sole case where Ottawa did not appeal given an intervention and a novel interpretation, the intervener was a federal commission (CRHC) opposing the government. These findings are actually quite surprising in the Canadian context, given that the federal and provincial governments frequently oppose one another (usually via intervention) in Charter litigation,30 and all of the NOOs cited here receive signiticant funding from the federal government, and on several occasions have co-operated with Ottawa in the drafting of legislation.

30 Morton, Hennig,!! and Ho (1996). Notwithstanding these findings, it may be the case that interveners do not exercise a direct influence (that is, as "allies" or "opponents") on government decision-makers, but serve, rather, as additional indicators of a case's ab ove-average importance.

Finally, the government's success rate on appeal to the Supreme Court suggests either that the Justice Department is a po or judge of its prospect of winning, or that winnability was not a strong factor in these decisions: of ni ne appeals, Ottawa was successful on the disposition of the case only three times (Canadian New.spapers, NWAC,

Thibaudeau), and only secured a clear reversaI of the lower court's novel interpretation in two (Canadian Newspapers and NWAC).

II. Judicial Amendments

The penultimate appeal courts amended federallegislation (through severance or reading in) in thirteen cases, nine of which were appealed to the Supreme Court: R. v.

Rao; Canadian Newspapers, El Zein v. The Queen, R. v. Wholesale Travel Group Inc.,

Schachter, Corbiere, R. v. Monney, R. v. Ruzic, and White, Ottenheimer and Baker v.

A. G. Canada. 3 1 As with the novel interpretation-intervener cases, judicial amendment cases spanned the time period studied, ranging from one of the earliest cases (Rao) to one of the most recent (White, et al.). Several ofthese cases have already been discussed; the remaining cases are reviewed below, with emphasis on whether and how the federal government addressed the judicial remedy in its appeal to the SCC.

31 [1987] 29 c.e.e. (3d) 560 (Q.C.A.); [1990] 63 D.L.R. (4th) 325 (O. C.A.); [1998] 153 D.L.R. (4th) 617 (O.C.A.); [1998] 128 e.c.e. (3d) 97 (O.C.A.); [2000]187 D.L.R. (4th) 581 (Nr.ç,,A.).

1.),)"" El Zein concerned whether extraditing Canadians for crimes committed in Canada

(in this case, international drug trafficking) violates the Charter's mobility rights (section

6(1 )). The Quebec Court of Appeal unanimously ruled that because s.6(1) explicitly grants Canadians the right to "remain in" Canada, the Extradition Act is of no force or effect, but only to the extent that it authorizes the extradition of Canadians for crimes committed while in Canada. That is, the court "read in" these conditions. Ottawa's appeal to the Supreme Court challenged this remedy, on the grounds that mobility rights are not absolute, and that extradition for transnational crimes is both consistent with tradition and a deterrent to such crimes. A majority of the SCC (Justices Wilson and

Sopinka dissenting) concurred with Ottawa, and reversed the Quebec court.

In R. v. Wholesale Travel Group Inc., the federal government defended the provisions of the Competition Act dealing with false or misleading advertising. Sections

36 and 37.3 of the Act required the accused business to prove it did not knowingly advertise falsely, a "reverse onus" which the Ontario Court of Appeal ruled violated sections 7 and 8 of the Charter. Further complicating matters, the Act required the accused to publish a "timely retraction" in order to claim the defense of "due diligence."

In a split decision, the OCA used a mixture of nullification and reading down to eliminate the requirement that the accused "establish" that he or she acted with due diligence via a retraction, with the result that "the statutory due diligence defense is available to an accused who raises a reasonable doubt on that issue.,,32

On appeal, the Crown simply contended that although the due diligence defense

32 [1990] 63 D.L.R. (4th) 325 at 326 per Tamopolsky lA.(O.C.A.) [emphasis added].

134 provided in the Act was more restrictive than the traditional common law defense, the provisions were nonetheless constitutional. Ottawa also made the corollary argument, which the SCC rejected, that the courts should treat economic regulation with greater deference than criminal offenses. The SCC's highly fractured decision provided Ottawa with a partial victory. The Court ruled 5-4 that the reverse onus provision requiring the accused to establish "on the balance of probabilities" that he had acted with due diligence was constitutional. However, the Justices were unanimous in finding the "timely retraction" requirement unconstitutional. In short, it was acceptable to put the onus of proving due diligence on the accused, but making retraction mandatory as part of this defense was not.

At issue in R. v. Monney was whether section 98 of the Customs Act empowered customs officers, based only a "reasonable suspicion," to detain travellers for several hours in a "drug 100" facility so as to obtain any drugs secreted intemally. Section 98 authorizes customs officers to search travellers on the basis of "reasonable suspicion," within a "reasonable period" after the traveller's arrivaI in Canada. The Ontario Court of

Appeal ruled that the vague criteria laid out in s.98 could not authorize such an extensive detention or invasive search, and found the search unconstitutional. Rather than strike down s.98, however, the OCA "read in" restrictions on the length of detention and invasiveness of the search. The federal govemment successfully appealed to the SCC, which concurred with the AG Canada that there is a lower expectation of privacy at the border. The Court reversed the read-in conditions, and actually expanded the meaning of the phrase "on or about his person" in s.98, to authorize "passive" searches for drugs concealed within the traveller's body (for exal1).Qle, "bedpan vigils," but not involuntary

135 body cavity searches). As weIl, the Court ruled that a long detention pursuant to s.98 is not, in itself, unreasonable.

The government contested the Ontario Court of Appeal' s revision of the Criminal

Code's defense of duress in R. v. Ruzic. Section 17 of the Code allowed the accused to claim the defense of duress-that is, that the accused committed the crime against her will, at the behest of another--only when the accused faced a threat of immediate death or bodily harm, and the threatener was present when the offense was committed.

Notably, this was significantly more restrictive than the traditional common law defense of duress. The OCA ruled that the restrictions violated the Charter's guarantee of fundamentaljustice in section 7, and read out (severed) the conditions. At the Supreme

Court, Ottawa submitted that the courts "should assume a posture of deference" when reviewing the constitutionality of statutory defenses, as "the legislature is best placed to determine what constitutes 'morally involuntary' conduct."33 No section 1 justification was even offered, as Ottawa instead argued that "properly construed, s.17 would not even infringe the Charter.,,34 Not surprisingly, the Court emphatically rejected the governrnent' s position, and upheld the OCA' s decision.

The final instance of judicial amendment appealed by the federal government occurred in the recent White, Ottenheimer and Baker v. A. G. Canada, which concemed the sc ope of solicitor-client privilege. The Criminal Code empowers police and the

Crown to seize and view material from lawyers, even though it may be protected by solicitor-client privilege. After such material has been seized, counsel may apply under

33 Cited in [2001] 1 S.C.R. 687 at paragraph 20, per Lebel 1. (S. c.e.). 34 Cited in [2001] 1 S.e.R. 687 at paragraph 20. per Lebel J. (S.C.e.).

136 section 488.1 of the Criminal Code for a determination of whether it is privileged; if so,

the materials are sealed and placed in the custody of the court. In White, et al., the

Newfoundland Court of Appeal severed the requirement in s.488.1 that counsel claim

privilege in order for seized records to be sealed, and read in a requirement that questions

of privilege be considered by a judge in aU cases. In other words, any documents seized

from lawyers were to be "automaticaUy packaged and placed in custody pending a

hearing before a judge," and the Attorney General must acquire judicial approval to

access the documents. Ottawa's appeal to the Supreme Court specificaUy cited the

remedy issue, with a mixed result: the Court opted for invalidation over amendment, but

then outlined ten conditions to govern search and seizure of materials in law offices until

new legislation is adopted.35

OveraU, Ottawa enjoyed only partial success in reversing the lower courts'

amendments. Of the nine appeals, the Supreme Court overturned the remedy outright

four times (Canadian Newspapers, El Zein, Schachter and Monney), and modified another for a partial government victory (Wholesale Travel). While these results may raise questions about the government's ability to predict success accurately, several of the cases contained multiple factors encouraging an appeal. As noted, Rao, Canadian

Newspapers, Schachter and Corbiere aiso witnessed novel interpretations and interveners, there was dissent on the Iower appeai courts in Schachter, Monney, and

Wholesale Travel, and the Attorney General of Ontario intervened in support of Ottawa in

Ruzic. In short, there were compeUing reasons to appeal these cases.

35 Lavallee, Rackel & Heintz v. Canada (Attorney Generai); White, Ottenheimer & Baker v. Canada (Attorney Gener~I!: R. v. Fink (S.c.c.) [rendered September 16,2002, not yet reported].

137 The same could be said of the four cases of judicial amendment not appealed to

the SCC-R. v. LaPlante,36 R. v. Ingebrigtson,37 Haig v. Canada, and Rosenberg v.

Canada (Attorney Generall8-but here, concerns about losing on appeal may have been

the deciding factor. It is notable that aIl of the cases were decided unanimously by the

lower court of appeal, suggesting Ottawa may have doubted its ability to win in the

Supreme Court. Jurisprudence certainly supports this conclusion in at least two of the

cases. In LaPlante, the Saskatchewan Court of Appeal severed the phrase in s.37(l)(a) of

the Food and Drug Act authorizing warrantless searches of places other than dwelling-

houses. The issue was therefore virtually identical to that in Southam v. Hunter and Rao,

in which the courts-most importantly, the Supreme Court in Southam-had already

found warrantless searches unconstitutional. In Ingebrigtson, the Court Martial Appeal

Court of Canada ruled that Standing Courts Martial can only be used in times of emergency, since they are not formally independent in the terms laid down by the

Supreme Court five years earlier in R. v. Valente (security of tenure and financial security for judges).39 However, as 1 detail in the next section, jurisprudence alone should not have dissuaded the government from appealing either Haig or Rosenberg, both dealing with discrimination on the basis of sexual orientation. The fact that both cases came from

36 [1988] 48 D.L.R. (4th) 615 (Sk.C.A.). 37 [1990] 76 D.L.R. (4th) 481 (C.M.A.C.). 3& [199&] 15& D.L.R. (4th) 664 (O.C.A.). 39 [1985] 2 S.C.R. 673; Va/ente (and Ingebrigtson) focused on section 11(d) of the Charter's right to a fair trial by "independent and impartial tribunal." Standing Courts Martial are govemed by section 177 of the National Defense Act. In contrast to General Courts Martial, which are composed of five officers and a Judge Advocate and deal with more-serious crimes (and which the SCC also found th violated judicial independence in R. v. Généreux [1992] 88 D.L.R. (4 ) 110), Standing Courts Martial have a single ad hoc presiding office ("president") and deal with offenses carrying punishments of 1ess than two years. The CMAC's decision effectively restored Standing Courts Martial to the original "emergency" function they served when creat~(tduring World War II.

138 a provincial court of appeal-specificaUy, the Ontario Court of Appeal, which the

Supreme Court tends to uphold-may be important here. As discussed in Chapter Three

(and confirmed in Chapter Four), the government should be less likely to appeal cases

from the provincial courts because of the greater jurisprudential risks, relative to the

federal appeal court, associated with a loss at the Supreme Court. The government's

concessions in Haig and Rosenberg may reflect calculated decisions to "cut its losses."

III. Unexpected Outcomes

The final category of cases are what can be characterised as statistically

"perverse" or unexpected outcomes. These are two types ofthese: unexpected appeals,

and unexpected concessions. We can distinguish two scenarios within the first category:

"lost causes," and, for lack of a better term, "why bother?" appeals. "Lost causes" are

appeals when there is virtuaUy no chance of winning, and are most likely motivated by

political considerations. The quintessential example is the Parti Québécois' s 1984 appeal

in A. G. Quebec v. Association of Quebec Protestant School Boards.40 In Protestant

Schools, the Quebec Court of Appeal invoked the new Charter's section 23 language

rights to invalidate the provision of Bill 101 41 requiring French-language instruction for

aU children except those whose parents were educated in English within Quebec, or in

another province and had moved to Quebec before 1977 (the year of the Bill's adoption).

In other words, Canadian children (or children of Canadian parents) who immigrated to

Quebec from another province after 1977 had to be educated entirely in French, even if

40 [1984] 2 S.C.R. 66. 41 Section 73 of the Charte de la langue française.

139 both parents were anglophones (this also applied to immigrants from outside Canada).

Bill 101 was the centerpiece of the separatist PQ government' s programme of protecting

and promoting French language and culture, while s.23 of the Charter embodied the

Trudeau federalists' vision ofpan-Canadian bilingualism. As Russell, Knopff, and

Morton write, the case "represented nothing less than the final clash between the two old

rivaIs, Trudeau and [PQ founder and leader René] Lévesque, and their different visions of

the future of the French in Canada.,,42 Because s.23 had been drafted specifically to

counter Bill 101, the Quebec government had little hope ofwinning on appeal to the

Supreme Court of Canada. However, the appeal would force the national institution to

issue a highly publicized ruling against "Quebec," thereby nicely serving the political

interests of the separatists by eroding the Court's legitimacy within the province.

No clear examples of lost cause appeals were found in the database of federal governrnent Charter appeals. Part of the reason may be that such cases are difficult to identifY on a strictly statistical basis, and trac king the jurisprudential and political context of every appeal is not feasible at this point. That said, 1 was able to identifY nine appeals which the government should have had little expectation of winning, as there was no dissent on the court of appeal nor any reversaI in the lower courts: Corbiere, Lagiorgia v.

Canada, Lavallee, Rackel and Heintz v. Canada, R. v. Doyer, R. v. Hayden, R. v. Oakes,

R. v. Rao, R. v. Ruzic, and Steele v. Mountain Institution. 43 These are not properly called

"lost causes," however, as Ottawa won on appeal in two ofthese (Corbiere and Doyer).

42 Russell, Knopff, and Morton (1990), 619. 43 [1987] 3 F.C. 28 (F.C.A.); [2000] 184 D.L.R. (4th) 25 (A.C.A.); [1993] 77 c.c.c. (3d) 203 (Q.c.A.); [1984] 3 D.LR. (4th) 361 (M.C.A.); [1990] 54 c.c.e. (3d) 334 (B.C.C.A.).

140 Nonetheless, the success rate in these cases, at roughly 25%, is significantly lower than the federal government's 54% overall success rate in appeals to the Supreme Court in

Charter cases.44 As well, three of the seven failed applications for leave appear in this category (Lagiorgia, Hayden, and Rao).

The second type of unexpected appeal, the "why bother?" cases, refer to appeals taken when none of the significant explanatory factors-novel interpretations, interveners, judicial dissent, or even less-intrusive remedies and reversal-were present.

They are, in effect, a sub-set of the nine cases noted above. Four cases fall into this category: Lagiorgia, Doyer, Hayden, and Steele. While the government's efforts to appeal these losses met with little success-the Court refused leave in two cases and dismissed the appeal in another-there were, on doser inspection, compelling reasons to appeal. This is most obviously so in Doyer, where the Ontario Court of Appeal stayed the extradition of a fugitive to the U.S. Not only would there have been sorne pressure from the U.S. government on the Justice Department of Canada to appeal, the OCA's ruling rejected a recent precedent on the issue from the British Columbia Court of Appeal in Ottawa's favour;45 a ruling the Supreme Court of Canada had effectively upheld by denying the losing rights daimant leave to appeal. The government assessed, correctly, that it could win in the Supreme Court.

While the remaining appeals did not me et with success, there were good reasons

44 Source: ca\culations by author from original data. Of 47 appeals, Ottawa won on the disposition of the case in 25, lost on appeal in 14, leave was refused in 7 (which effectively upholds the loss at the lower court), and 1 decision is pending; 54% represents 25 of 46, the number of cases whose outcome is known. 45 U.S.A .. , et al. v. Hagerman [1990] 60 C.c.c. (3d) 144 (B.C.C.A); leave refused by. S.. c.c.

141 to try. Following the Federal Court of Appeal's decision in Lagiorgia, there were four recent and conflicting precedents on whether illegally-seized documents had to be retumed immediately pending re-investigation or the Crown obtaining a new search warrant, an issue that would significantly affect the speed of investigations.46 Hayden concerned the constitutionality of the lndian Act provision making it an offense to be intoxicated on a reserve. The government actually won on the Charter issue, but appealed anyway, to contest the provision's invalidation under the 1960 Bill ofRights's right to equality (the Charter's equality rights were not in force until 1985). Moreover, the

Manitoba Court of Appeal's decision overruled two Saskatchewan lower court decisions on the issue.47 Finally, Steele concerned whether the courts had jurisdiction (over the

Parole Board) to consider the constitutionality of indeterminate sentences handed down before the adoption of the Charter. While the British Columbia Court of Appeal's decision defending the court's jurisdiction was not a nove! interpretation-that had been established generally in R. v. Gamble48-the case raised sorne interpretive issues on which it disagreed (to varying extents) with other recent cases.49

The third and final type of "perverse outcome" are those cases where the govemment does not appeal in spite of strong incentives to challenge the lower court ruling. Two cases in the database meet this description, both dealing with equality rights claims on the basis of sexual orientation: Haig v. Canada and Rosenberg v. Canada. As

46 The other cases were: Re Dobney Foundry Ltd. v. A. G. Canada [1985] 19 C.C.c. (3d) 465 (B.C.C.A.) [not followed in Lagiorgia]; Re Chapman and the Queen, 9 D.L.R. (4th) 244 (O.C.A.) [followed]; and Lefebvre v. Morin, [I985], No. 200- 10-000-174-83, digested at lE. 85-366 (Q.C.A.) [followed]. 47 R. v. Whiteman (No. 1), [1971] 2 W.W.R. 316 (Sk.C.Q.B.), and King v. The Queen, [1982] 2 W.W.R. 367 (Sk.C.Q.B.). 48 [1988] 2 S.C.R. 595. 49 R. v. Galbraith, [1988] 46 c.C.C. (3d) 82 (B.C.S.C.) and R. v. Milne, [1987] 2 S.C.R. 512.

142 already discussed, the concession in Haig is unique among the cases with both a novel

interpretation and an intervener, and the case also witnessed a judicial amendment; in

short, there were factors strongly encouraging an appeal. As well, although "there was no direct budgetary consequence of any significance" in adding sexual orientation to the

Canadian Human Rights Act,s° there were very substantial indirect consequences for all federallaws that distinguished on the basis of sexual orientation, most notably, the pervasive heterosexual definition of "spouse" in the Income Tax Act, pension programs,

Employment Insurance, and laws governing marri age and divorce.

Although none of the statistical indicators for winnability (dissent, reversaI) were present in Haig, the government should have had sorne reason to believe an appeal to the

Supreme Court could be successful. In 1992, no other penultimate court of appeal, let alone the Supreme Court, had endorsed the extension of s.15 protection to sexual orientation.51 Moreover, a year earlier, in Canada (Attorney General) v. MOSSOp,52 the

Federal Court of Appeal had ruled against a sexual orientation daim. In Mossop, the appeal court refused to extend the Canadian Human Rights Act' s protection of "family status" to same-sex couples, a decision upheld by the Supreme Court in 1993.53

Morton and Knopff openly accuse the Justice Department, and specifically Justice

Minister Kim Campbell (and Anne McLellan in Rosenberg), of intentionally losing the

50 Nitya Duclos (1992) "A Remedy for the Nineties: Schachter v. R. and Haig & Birch v. Canada," Constitutional Forum 4: 26. 51 Smith (1999),157-163. 52 [1991] 1 F.C. 18 (F.C.A.). 53 Canada (Attorney General) v. Mossop, [1993] IS,Ç.R. 554 (S.c.c.).

143 case in the lower courts and then refusing to appeal, to advance her agenda of supporting

gay rights, over the objections of caucus.54 There is considerable support for the

accusation, at least in the case of Campbell in Haig. Support within the Justice

Department for sexual orientation claims during this period is evident from the

government's factum in Veysey v. Canada (Commissioner ofCorrectional ServicesJ,ss

which stated: "it is the position of the Attorney-General of Canada that sexual orientation

is a ground covered by section 15 of the Charter.,,56 As weIl, Judge Krever of the Ontario

Court of Appeal supported his decision to "read in" by citing "the commitment of

successive Ministers of Justice on behalf oftheir governments to amend the legislation to

add sexual orientation to the list ofprohibited grounds of discrimination."S7 Dissent

within the Conservative caucus on the issue is also well-documented. As Duclos

observes, Krever's comments significantly ignored the fact that successive govemments

were unable to pass such legislation. s8 In his autobiography, John Crosbie, former

Minister of Justice in the Mulroney government, recalled that the attempt in 1986 to

amend the Canadian Human Rights Act to prote ct sexual orientation "caused an unholy row in the Conservative caucus," with one back-bench MP publicly announcing that a majority of caucus opposed the reforms. 59

54 Morton and Knopff (2000), 117-118. Their primary basis for this claim is the Justice Department's concession of the rights violation, rather than defending the legislative history, in which attempts to add sexual orientation to s. 15's enumerated grounds were "soundly defeated." (117) 55 [1990] 109 N.R. 300 (F.C.A.). 56 Veysey v. Canada (Commissioner ofCorrectional Services), [1990] 109 N.R. 300 at 304 (F.C.A.). 57 [1992] 94 D.L.R. (4th) 1 at 12 per Krever 1. (O.C.A.) 58 Duclos (1992), 26. 59 John C. Crosbie (1997) No Ho/ds Barred: My Life in Politics (Toronto: McClelland & Stewart), 271; the MP in question was Dan McKenzie (Manitoba), who opposed the reforms. For details of the Justice Department' s proposed reforms, see Department of Justice Canada (1986) Toward Eqllality: The Response to the Report of the Parliamentary Committee on Eqllality Rights (Minister of Supply and Services).

144 These facts notwithstanding, there were other factors which argued in favour of not appealing to the Supreme Court. Foremost among them was the Department's belief that the prospect ofwinning was low, regardless of the Mossop precedent. When speaking briefly about Haig, a senior department official allowed, "You don't continue beating away at something if you feel the judgment was probably right and probably would be sustained."60 As well, and as noted earlier, not appealing the provincial court ruling restricted the jurisprudential impact, which could have been overwhelming if upheld by the Supreme Court. Although the federal government had conducted an internaI review of its legislation for compliance with s.15' s enumerated grounds, it is quite unlikely that the government understood the full impact of adding sexual orientation, or was prepared to address constitutional challenges on this basis. In short, the Justice Department (and the government in general) may have been buying time.

Rosenberg involved an equality rights challenged to the Incorne Tax Act's heterosexual definition of "spouse." A trade union registered its pension plan with

Revenue Canada to quaI ify for significant tax deferral advantages under the ITA, but when the union sought to change the plan's definition of "spouse" to include same-sex couples, Revenue Canada refused, citing a conflict with the ITA' s definition. On behalf of two lesbian employees (Rosenberg among them), the union filed a section 15 challenge against the ITA. At trial, the government conceded the violation, and persuaded the judge that the law was reasonable under section 1. The Ontario Court of Appeal reversed the trial judge, and "read in" a same-sex definition of spouse, rather th an nullify the law and

60 Graham Garton, Department of Justice Canada, personal interview (July 31, 2001, Ottawa) [emphasis added].

145 deprive heterosexual couples of their tax benefit. The government refused to appeal to the Supreme Court, going so far as to fight off an extraordinary attempt by the Official

6 Opposition to force a vote on the decision. \

While the use of judicial extension in Rosenberg was a pressing reason to appeal, other factors should have been even more compelling. Although the redetinition of

"spouse" was not a novel interpretation here-the OCA had accomplished this in 1996 in

63 M v. H 62 (which the Ontario government appealed, unsuccessfully, in 1998 )-the decision nonetheless carried profound implications for dozens of provisions in tax and social welfare legislation. Moreover, unlike Haig, the OCA's ruling in Rosenberg would have direct budgetary consequences, in the form of lost tax revenue. Further still, as

Morton and Knopff observe, "until this case the government had never lost a Charter appeal involving a discrimination claim against the ITA.,,64

An important case in point is the Supreme Court ofCanada's 1995 ruling in Egan v. Canada. 65 In Egan, a gay couple was unable to persuade the Court that the Otd Age

Security Act' s heterosexual definition of spouse violated the Charter' s equality rights.

While the government conceded that sexual orientation was an "analogous ground" under section 15, it submitted that the OSA was intended to support couples in their old age who have raised children, in recognition of the inherent tinancial costs associated with raising

61 Morton and Knopff (2000), 118. Notably, the govemment's ability to secure the support of caucus suggests the Liberal rank-and-file were either significantly more supportive of gay rights than their Tory counterparts in 1992, or under significantly greater control by the Liberal leadership. The fact that a faction of the Liberal caucus secured a provision in the legislation stating that the Canadian Human Rights Act would not affect the definition ofmarriage suggests the latter. 62 [1996] 142 D.L.R. (4th) 1 (O.C.A.). 63 [1999] 2 S.C.R. 3 (S.C.C.). 64 Morton and Knopff (2000), 118. 65 [1995]2S.C.R.513.

146 children (inc1uding the frequent situation where one parent, usually the mother, has foregone paid employment for child-rearing). Ottawa contended that because this situation only very rarely applies to homosexual couples, the law was not, therefore, discriminatory, or at leastjustifiably so under section 1. The Supreme Court agreed, finding no section 15 violation, and explicitly concluded that the legislation would survive under section 1 analysis regardless. The government thus had a favourable, recent precedent directly on the issue in Rosenberg-namely, spousal recognition of same-sex couples-and an overt case of noncompliance by the Ontario Court of Appeal with said ruling.66 The incentive to appeal appears obvious.

However, appearances can be deceiving. As with Haig, there were several factors that should have given the govemment pause. The first was an unfavourable precedent on sexual orientation that was even more recent in 1998 than Egan: Vriend v. Alberta.67

The Supreme Court' s high-profile decision in Vriend, issued only weeks before

Rosenberg, amended Alberta's human rights code (the Individual's Rights Protection Act) to inc1ude sexual orientation, just as the Ontario Court of Appeal had rewritten the

Canadian Human Rights Act in Haig. Although spousal recognition was not explicitly addressed by the Court, the decision represented a strong endorsement of sexual orientation c1aims under section 15, in contrast to its 1993 decision in Mossop.

Another, less well-known aspect of the Vriend decision also weakened the government' s case in Rosenberg. Vriend saw the Supreme Court alter the rules of

66 In the "Cases referred to" section at the beginning of Rosenberg, Egan is listed with the notation, "not followed." 67 [1998] 1 S.C.R. 493 (S.c.c.).

147 analysis under the section l "reasonable limits" tests. The initial formulation of the

section 1 test laid down by the Court in Oakes was as follows:

1. Was the violation prescribed by law? (this exc1udes from section 1 protection "policies embedded in administrative practice, rather than explicit legislation" or "vague enabling legislation" where "the actual standards used by the law' s administrators are their standards, not the law's.")68

2. Is the violation justified by a "pressing and substantial objective"?

3. Are the legislative means "proportional" to the objective (are they (i) "rationally connected"; (ii) is there a "minimal impairment" of the rights, that is, are the "least drastic means" used; and (iii) does the bene fit to general welfare outweigh the cost to the individual's rights?)

Importantly, the "pressing and substantial objective" in the second stage referred to the

legislative objective of the impugned law. Following Oakes, the Court demonstrated

considerable deference to legislative objectives, and focused its attention on the

proportionality component, leading Knopff and Morton to observe that "the courts will

rarely invalidate legislation because its objective is not important enough.,,69 In Vriend, the Court rewrote this step, so that the question was not whether the legislative objective

was pressing and substantial, but whether the violation was, a significantly more difficult

criterion for the government to satisfy. Notably, the Ontario Court of Appeal cited this

amended section 1 test in Rosenberg, finding that while supporting dependent widowed women was a valid legislative objective, violating the rights of gay and lesbian couples

68 Knopffand Morton (1992), 40 [emphasis in original]. 69 Knopff and Morton (1992), 47. The authors highlight the Court's decision in Rocket v. Royal Ca//ege of Dental Surgeons of Ontario, [1990] 2 S.C.R. 232 (S.C.C.), where the Justices "dropped any reference to the 'pressing and substantial' standard, arguing simply that the legislative objective 'must be ofs!lffÏcient importance to warrant overriding a constitutionally protected right."" (46)

148 was not. In light of Vriend, the govemment' s questionable decision to concede the rights

violation in Rosenberg-even though it had successfully argued in Egan that a

heterosexual definition of spouse was not discriminatory-meant that Ottawa had given

up the strategically most important issue, as a section 1 defense would be virtually

impossible.

A third factor discouraging an appeal concems the Supreme Court's composition,

which changed between Egan and the Rosenberg period. The Egan decision was sharply

divided, with Justices Lamer, LaForest, Sopinka, Gonthier and Major ruling in Ottawa's

favour, and with Justices L'Heureux-Dubé, Cory, McLachlin and lacobucci in dissent.

By April 1998, when Rosenberg was decided by the OCA, two members of the Egan majority were gone: Justice LaForest had retire d, and Justice Sopinka had passed away.

LaForest had been replaced by Justice Bastarache-who had ruled with the majority in

Vriend-while Sopinka's seat had been filled by Justice Binnie, who, while not a completely unknown quantity (he had been a former Associate Deputy Minister of Justice for Canada (1982-86) who subsequently represented the govemment on several occasions), had been appointed directly from private practice, with no previous bench experience. The government could not, therefore, count on the current Court to follow

Egan, especially iffewer than ni ne Justices heard the case. Finally, as a provincial court ruling, Rosenberg, like Haig, entailed greater jurisprudential risks on appeal than a federal court case.

Notably, the foregoing raises a problem for Morton and Knopffs argument: if, as they contend, the Justice Department's primary concem in Haig and Rosenberg was to promote the adoption of sexual orientation as an ana[ogous ground l,lnder s.15, why did it

149 not appeal to the Supreme Court? The Department strongly suspected the Court would rule in favour of protection (although perhaps more so in Rosenberg than Haig), and, in any case, Ottawa could have conceded the violation and offered a weak section 1 defense, or contested the violation and made no section 1 argument (the latter ofwhich the government had done at the lower appeal court in Haig). That the Justice Minister did not opt for this strategy suggests the need for a more nuanced interpretation, and the evidence presented supports the conclusion that government lawyers-while they did agree with the lower court decisions-acted strategically to minimize losses.

IV. Summary

A closer look at the cases discussed herein reinforces the conclusion of the previous chapter, that the federal government' s appeal decisions appear to be motivated by rational considerations. Chief among these are concerns about winning, minimizing losses, and, relatedly, protecting the discretion of legislative and administrative actors.

This is true not only in cases of obvious importance-those with novel interpretations and judicial amendment-but also in cases where the reason for appealing was not immediatelyevident. The following chapter extends the contextual analysis to aH cases where the lower court's remedy affected federallegislation (whether by judicial amendment or invalidation), in order to determine whether the government responded to the 10ss in the legislative rather than the judicial arena. Instances of true inter institutional "dialogue" between the government and the lower courts, for example, might help explain why less-intrusive remedies were not a statisticaHy significant factor in the government' s appeal decision-making.

150 CHAPTERSIX

Government Appeals as "Dialogue": Expanding a Contemporary Debate

Considerable scholarly attention has been directed recently to the relationship

between governments and the courts under the Charter of Rights and Freedoms. The

dominant metaphor used to de scribe this relationship is that of a "dialogue" over

constitutional interpretation. 1 However, the form of governmental expression within this

dialogue examined to date is legislative "sequels," or legislation passed in response to judicial decisions. This approach conceives of dialogue in purely inter-institutional terms, with the judiciary "speaking" through rulings, and government through

legislation.2 This overlooks the important, on-going dialogue that occurs through

government litigation. In contrast to inter-institutional dialogue, dialogue via litigation brings the two parties together within the court setting, with government lawyers arguing legal interpretation before judges. While dialogue via litigation can occur in several scenarios, the topic of this dissertation addresses one important form of dialogue with

Hiebert (2002); Hogg and Bushell (1997); Hogg and Thornton (1999); Manfredi and Kelly (1999, 2001); Morton (1999); Roach (2001); see also the Supreme Court ofCanada's opinions in Vriend, Corbiere, MiIls, and M v. H., particularly those penned by Justice Iacobucci. 2 While still focusing on legislation, Kelly (1 999b) and Hiebert (2002) examine the degree to which the federal bureaucracy and Parliament address Charter rights infernal/y, in their policy-making processes. Vallinder (1995) somewhat confusingly refers to this as a form of "judicialization," or "the spread of judicial decision-making methods outside the judicial province proper ("When the Courts Go Marching ln," The Global Expansion ofJudicial Power, eds. C. Neal Tate and Torbjorn Vallinder (New York: New York University Press), 13).

151 lower courts: appeals to higher courtS. 3 Manfredi and Kelly (1999) are exceptional for their acknowledgment ofthis form of dialogue, observing that "[a]n appeal explicitly communicates a democratic actor' s judgement that a judicial decision is wrong in sorne sense.,,4

The chapter proceeds in two parts. The first reviews the existing literature on the dialogue metaphor, with special attention to the various attempts to define "dialogue" and the normative implications ofthese definitions. Foremost among my findings is that pro- dialogue writers rely on the potential use of the Charter's legislative override (s.33), which remains an unrealistic option for high-profile, politically salient issues. On the other hand, critics of the metaphor, who require sorne form oflegislative noncompliance as evidence of dialogue, define away the potential for genuine agreement resulting from inter-institutional exchanges. In the second part, 1 measure the degree of dialogue between the federal government and the lower appeal courts, based on the various criteria put forth by the authors discussed in Part 1. 1 find a fairly robust level of dialogue at this level, but primarily in the form oflitigation (appeals), especially in response to judicial amendments of legislation. Moreover, legislative responses are typically delayed, and require extensive changes. Finally, legislative sequels provide little explanation for those instances when the federal government decides not to appeal judicial invalidations or amendments, as litigation do es not preclude legislation, or vice versa.

3 There are four such scenarios, the first three ofwhich are instituted by the government: (1) when government appeals from its loss in a lower court; (2) when government "intervenes" in a case as a third party; (3) reference cases. where the Government of Canada may ask the Court directly for an "advisory opinion" (there is no lower court "case" in this scenario); and (4) wh en the government appears before the Court as a respondent. that is, involuntarily, as the target of litigation. -+ Manfredi and Kelly (1999), 524.

152 J. The Dialogue Debate so Far

Hogg and Bushell contend that, "where a judicial decision is open to legislative

reversaI, modification, or avoidance, then it is meaningful to regard the relationship

between the Court and the competent legislative body as a dialogue.,,5 With this

assertion, the authors were the first to characterize the judicial-Iegislative relationship in

Canada as a dialogue, and the metaphor was quickly adopted by the Supreme Court of

Canada in Vriend. 6 The crux of Hogg and Bushell's argument is that concems about the

unelected judiciary frustrating the actions of democratic governments overestimate the

courts' real power, by ignoring the capacity of elected officiaIs to respond. Such

responses are facilitated, they contend, by four factors: (1) most overtly, the section 33

legislative override; (2) the provision in section 1 allowing "reasonable limits" of Charter rights; (3) the limited nature of sorne Charter rights; and (4) the court's ability to craft remedies other than invalidation, such as reading in. 7

Based on their study of "legislative sequels" to 65 invalidations by the Supreme

Court and selected lower courts, the authors conclude that "Charter cases nearly always can be, and often are, followed by new legislation that still accomplishes the same objectives as the legislation that was struck down."g As such, the court cannot usually

"block" legislative objectives, only "influence" them by articulating the Charter values at stake. They continue,

5 Hogg and Bushell (1997), 79. 6 [1998] 1 S.C.R. 493. Notably, A.C. Hutchinson addressed-and criticized-the depiction ofjudicial review as a "dialogue" in his 1995 work, Waitingfor CORAF: A Critique of Law and Rights (Toronto: University of Toronto Press), but his treatment referred to a dialogue with the public, not in the inter institutional sense it is used here. 7 Hogg and Bushell (1997), 82-91.

153 In that case, the judicial decision causes a public debate in which Charter values play a more prominent role than they would if there had been no judicial decision. The legislative body is in a position to devise a response that is properly respectful of the Charter values that have been identified by the Court, but which accomplishes the social or economic objectives that the judicial decision has impeded. ,,9

Manfredi and Kelly identify two major lines of criticism that may be directed at

Hogg and Bushell's argument. IO The first focuses on the many methodological problems in Hogg and Bushell's article, which, given the empirical basis oftheir core argument, are not lightly dismissed. These include, briefly: Hogg and Bushell's exclusive focus on nullifications, thereby ignoring judicial amendments such as reading in and severance; case selection bias associated with only considering lower court cases deemed

"important" by Hogg himself; the use of lower court cases at all, when such decisions may be reversed on appeal; and artificially inflating the number of "sequels" by counting multiple responses when there was only one legislative action following several judicial rulings on the exact same law. 11 While these criticisms are largely justified and are not convincingly rebutted by Hogg and Thorton,12 the inclusion of lower court cases cannot be ruled out so hastily. It is an empirical question whether lower court rulings producing a legislative sequel have been reversed by a higher court, and Manfredi and Kelly do not provide such data. As I demonstrate in Part II ofthis chapter, there are several occasions where the lower appeal court rendered the final judicial decision, or where there was a

8 Hogg and Bushell (1997), 75. 9 Hogg and Bushell (1997), 79-80. 10 Manfredi and Kelly (1999). 1 1 Manfredi and Kelly (1999), 515-521. 12 Hogg and Thornton (1999).

154 legislative sequel to a lower court ruling despite a reversaI of that ruling by the Supreme

Court of Canada. To the extent one accepts Hogg and Bushell's dialogue metaphor, such

scenarios must quai ify as dialogue, and should be included.

Manfredi and Kelly's second line of criticism is a normative one, that "It is

unclear ... that even uncritical acceptance of the dialogue metaphor as operationalized by

Hogg and Bushell would answer the democratic critique of Charter-based judicial

review.,,13 Their criticism is explicitly rooted in Mark Tushnet's concept of "policy

distortion," which occurs when "legislators choose policies that are less effective but

more easily defensible than other constitutionally acceptable alternatives.,,'4 That is, law-

makers avoid policy choices for fear of judicial invalidation, regardless of whether the

law is actually invalidated. ludicial rulings therefore constrain legislative choices even

when "sequels" occur.

Tushnet' s argument is animated by the Madisonian theory of "coordinate construction," according to which the judiciary and the elected branches may each

legitimately give meaning to the constitution, through their respective interpretations; this explains the apparent paradox of how a "constitutionally acceptable" policy may not survive judicial review. Tushnet concludes that "more than minimal" judicial review is necessarily problematic for democracy, because the ensuing "policy distortion" implicitly constrains the legislature' s ability to articulate its understanding of the constitution, and to make laws in accordance with that understanding. Similarly, Manfredi and Kelly

13 Manfredi and Kelly (1999), 522. 14 Mark Tushnet (1995) "Policy Distortion and Democratie Debilitation: Comparative Illumination of the Countermajoritarian Difficulty," Michigan Law Review 94: 250, cited in Manfredi and Kelly (1999), 522.

1::) -- identify Hogg and BusheIl's "assumption ofajudicial monopoly on correct interpretation"-or what Knopff and Morton term "oracular legalism" 15_as the "most crucial flaw in [their] normative argument.,,16 They conclude, "Genuine dialogue [and by implication, democracy] only exists when legislatures are recognized as legitimate interpreters of the constitution and have an effective means to assert that interpretation.,,17

Manfredi and KeIly's normative argument thus engages the very definition of dialogue, a topic to which 1 retum shortly.

To these two lines of criticism, Morton adds two more. 18 First, he observes that while legislative responses to adverse judicial rulings may be possible, Hogg and

BusheIl's argument ignores Flanagan's finding that ajudicial ruling creates a new "policy status quo" that actually limits the ability of governments to respond. 19 On its face, this is an odd criticism, given the empirical basis for Hogg and BusheIl's position; ifFlanagan's theory is right, after aIl, legislative sequels should not be common, yet Hogg and Bushell conclude otherwise. Morton's argument therefore rests on an a priori rejection of Hogg and Bushell's empirical findings, and the "staying power" theory is more appropriately understood as an explanation for the limited degree of dialogue. Notably, Morton does,

15 Knopff and Morton (1992), 179; as an example this view, they point to the idea that legislatures may only suspendjudicial rulings under s.33 or justify violations under s.l, both ofwhich assume the initial judicial interpretation is the "correct" one. 16 Manfredi and Kelly (1999), 523. Notably, Manfredi and Kelly mistakenly charge that Hogg and Bushell gloss over the distinction between respecting the Charter's meaning and respecting the court's interpretation of the Charter (523). Hogg and Bushell do imply repeated1y thatjudicia1 interpretations are authoritative and should be respected by legislatures, but they are not legal positivists; that is, they acknowledge thatjudicial rulings are interpretations, and not simple applications of the Charter's "inherent" meaning. 17 Manfredi and Kelly (1999), 524 [emphasis in original]. 18 Morton (1999). 19 Thomas Flanagan (1997) "The Staying Power of the Status Quo: Collective Choice in Canada's Par1iament after Morgentaler," Canadian Journal of Political Science 30: 31-54.

156 in a separate argument, dismiss Hogg and Bushell's findings as based on a flawed definition of dialogue; this is discussed further below.

Morton also rejects the distinction Hogg and Bushell draw between legislative objectives (or ends) and means. According to Hogg and BusheU, "The effect of the

Charter is rarely ta black a legislative objective, but rather to influence the design of implementing legislation."20 Their reasoning is that most invalidations occur in section 1 analysis because the courts conclude the policy is not the "least restrictive means," rather than finding the objective is not "pressing and substantial." This distinction parallels that drawn by John Hart Ely between procedural and substantive review,21 and Morton's criticism echoes Ely's critics; namely, that that there is considerable substance in the procedure chosen. As Morton states, "politics is as much about me ans as ends ....No respectable person is willing ta defend child pomography, but many will argue that restrictions on it must be balanced with our respect for freedom of expression and privacy."22 Furthermore, "apparent disagreement about means sometimes tums out to be disagreement about ends. Everything depends on the purpose(s) ajudge attributes to the statute .... [A]ny half-clever judge can use procedural objections as a pretense to strike down legislation that he opposes for more substantive reasons."23

The most fundamental criticism raised by both Manfredi and Kelly and Morton regards the broad de±inition of "dialogue" implicit in Hogg and Bushell's treatment of any legislative response as a meaningful "sequel." Indeed, whether dialogue exists is

20 Hogg and Bushell (1997), 75. 21 John Hart Ely (1980) Democracy and Distrust: A Theory of}udicial Review (Cambridge, Mass.: Harvard University Press). 22 Morton (1999), 24.

157 secondary to the debate over the proper definition of the term, since the metaphor' s

accuracy follows directly from the definition used. On the surface, Manfredi, Kelly and

Morton agree with Hogg and Bushell's initial claim that dialogue occurs when "ajudicial decision is open to legislative reversaI, modification, or avoidance." Disagreement arises because Hogg and Bushell do not retain this strict definition in their empirical analysis, shifting their operational definition to "sorne action by the competent legislative body."24

Employing this far looser standard, Hogg and Bushell count as meaningful responses legislation which pre-dates the courts' rulings, legislative repeals, and arnendments which simply parrot the court.

Manfredi and Kelly distinguish six "degrees" of dialogue, only one ofwhich they consider a "positive" legislative response satisfying Hogg and Bushell's initial criteria. 25

A "positive" response occurs when "elected officiaIs reflect on the implications of legislative decisions, and revise statutes to advance legislative objectives in a manner that complies with the Charter. They cite as an exarnple the federal governrnent's rejection of the Supreme Court's decision in Daviault (extending the drunkenness defense to assault offenses), "since Parliarnent arnended the Criminal Code to include a prearnble that discussed the constitutionality of the new section. ,,26

In contrast, "Charter dialogue does not characterize the process whereby elected officiaIs simply repeal offending sections or replace entire Acts. Such responses ... border on Charter ventriloquism because elected officiaIs are simple expunging sections or

23 Morton (1999), 24. 24 Hogg and Bushell (1997), 82. This point was originally made by Manfredi and Kelly (2001), 325. 25 Manfredi and Kelly (1999), 520-521. 26 Manfredi and Kelly (1999), 520.

158 whole laws found to be offensive to judicial actors, and thus are simply complying with judicial decisions."27 Similarly, Morton bluntly observes, "[0 ]beying orders is not exactly

what most of us consider a dialogue. If 1 go to a restaurant, order a sandwich, and the

waiter brings me the sandwich 1 ordered, 1 would not count this as a 'dialogue. ",28 ln

addition to simply repealing Acts and replacing Acts to satisfy judicial requirements,

Manfredi and Kelly characterize legislative "prequels," judicial amendment of laws, and

the absence of a legislative sequel as the remaining "degrees" of "negative" dialogue.

Based on these redefined parameters, Manfredi and Kelly conc1ude that "true" dialogue

occurs in only a third of cases, and, contrary to Hogg and Bushell, most involve major

legislative changes rather than minor amendments.29

Before proceeding, Manfredi and Kelly' s ')udicial amendment" category requires

sorne explanation. Recall that Hogg and Bushell c1aim to inc1ude only instances of invalidation in their study; Manfredi and Kelly contend that one of these "invalidations" is in fact a case of reading in (Miron v. Trudel), or judicial amendment. However, the mere fact of judicial amendment does not constitute a form of dialogue. Rather, the case should simply have been dropped from analysis, or ail cases ofjudicial amendment should have been inc1uded, to test the hypothesis that invalidations are easier to respond to than judicial amendments (in either case leaving only five degrees of dialogue). 1 offer such comparative analysis of sequels to lower court rulings in Part II.

Other authors have recently waded into the dialogue debate to attempt to clarify

27 Manfredi and Kelly (1999), 520-521. 28 Morton (I999), 23. 29 Manfredi and Kelly (1999),521.

159 the metaphor's meaning. Kent Roach identifies three modes of dialogue that are

conceivable in the Canadian context: Cl) dialogue based on courts and legislatures each

interpreting the constitution for themselves, or "coordinate construction"; (2) dialogue

based on legislatures holding the judiciary accountable, usually in conjunction with an

appeal to majoritarian democratic values; and (3) dialogue based on courts and

legislatures playing distinct but complementary roles, specifically, independent courts

focusing on constitutional princip les and legislatures "reflect[ing] the will of the

majority."30 Rejecting the first two "strong" forms of dialogue as derived from the

American constitutional context-where there is no legislative override or "reasonable

limits" provision-Roach prefers the third option, where legislative noneomplianee

should at least be govemed by use of the override, and legislation should be drafted with

s.l analysis in mind, to educate judges about the factors and alternative means the

govemment eonsidered.

Roach's argument begins with an acknowledgment of the relativism, or

definitional flexibility, implied by dialogic arguments. That is, dialogue theorists eontend

that eonstitutional texts rarely provide clear guidelines for public poliey, and therefore

interpretation is necessary.31 Given this, the crucial issue is whose interpretation to

follow: unelected judges, or elected and duly accountable legislators? Roach eontends

that because "right answers" are not possible in the real world, we should not expect judges to get it "right"; rather, this is an argument for not leaving the final decision to judges-hence his support for the legislative override, which at least is consistent with

30 Kent Roach (2001), 239-251; 247. 31 Roach (2001), Chapter 12.

160 democratic principles. To restate: in the absence of moral absolutes, we should decide

issues democratically. Anyone familiar with political theory will recognize this as the

central tenet of liberal democracy.

A major flaw in Roach's preferred mode!, however, is the erroneous contention,

also articulated by Hogg and Bushell and the majority in Vriend, that Canadian

legislatures typically ignore minoritarian interests, and that the courts are uniquely placed

to prote ct such interests. Take, for example, the following passage from Roach:

The Court promotes democracy ... because it requires the elected government to take responsibility for and justify to the people its decisions to limit or override rights that are liable to be neglected in the legislative and administrative process. The Court forces Parliament to consider the rights of minorities and the unpopular, while not dictating the final word about matters involving these rights. 32

This view is a caricature of the legislative process, for at least two reasons. The first is the mistaken assumption of a c1ear majority will: on most serious policy and moral issues, there are multiple positions rather than a c1ear bifurcation between a majority and minority. That is, a plurality, not a majority, is the norm. This is not to say that majorities, either of the "moral" or numerical variety, do not exist on sorne issues; historically, this was the case for sexual orientation, or African-Americans in the U.S.

South. But on many issues--even highly contentious ones such as abortion-the

"majority" is a rhetorical construction rather than reality.

The second fallacy in this caricature is that the legislative process itself is majoritarian. There are several problems with this view. To begin with, the legislative

32 Roach (2001), 250.

161 process in Canada is hardly "legislative" at aU, as the core policy functions of agenda- setting, policy formulation and implementation are overwhelmingly controlled by the political executive, not Parliament. 33 In light of the embedded Cabinet executive, the institutional tendency toward majority government, strict party discipline, and the common invoking of closure on debate, Parliament' s role is typically that of rubber stamp. The "dialogue between courts and legislatures" is thus more accurately labelled the "judicial-executivellegislative" dialogue. However, for ease and consistency with the existing debate, the simpler "legislative" term is used here.

Recognizing the central role of the executive is crucial, for with it cornes an appreciation of the extensive clientele networks that include the very "minorities" (or more accurately, pluralities) Roach and others assume are ignored by law makers. As

Cairns captured with his concept of the "embedded state," and Pallater affirmed, many traditionaUy disadvantaged groups, including women, gays and lesbians, and cultural and linguistic minorities frequently enjoy state sponsorship and engage in what Hodgetts terms "direct exchanges" with the public service in the policy process. 34 Furthermore,

Kelly's recent study found that the federal Justice Department actively screens prospective legislation, including that from other departments, for potential Charter

33 See, for exarnple, Barbara Wake Carroll (1991) "The Structure of the Canadian Bureaucratie Elite: Sorne Evidence of Change," Canadian Public Administration 34: 359-372; Donald Savoie (1999) Ooverningfrom the Centre: The Concentration of Power in Canadian Politics (Toronto: University of Toronto Press); and Richard Schultz (1977) "Prime Ministerial Government, Central Agencies and Operating Departrnents: Towards a More Realistic Analysis," Apex of Power, ed. Thomas A. Hockin (Scarborough: Prentice-Hall). 34 Alan C. Cairns (1986) "The Embedded State: State-Society Relations in Canada," State and Society: Canada in Comparative Perspective, ed. Keith Banting (Toronto: University of Toronto Press); Leslie A. Pal (1993) Interests ofStme: The Politics of Language, Multiculturalism, and Feminism in Canada (Montreal: McGill-Queen's University Press); lE. Hodgetts (1973) The Cunudian Public Service: A Physiology ofOovernment. 1867-! 970 (Toronto: University of Toronto Press): 346.

162 violations; similar "Charter-proofing" (or more accurately, "Charter-vetting") processes also exist in sorne provinces, such as Ontario.35

In a related vein, Janet Hiebert's most recent work argues that Parliament can, sometimes has, and should proactively address minoritarian rights issues; that is, MPs

(and bureaucrats) should take it upon themselves to make government accountable regarding the consequences of proposed and existing legislation for Charter rights. 36 She contends that this would help rehabilitate the public's opinion ofParliament, by enhancing the legislature's role and relevance, thereby reversing the anti-democratic trend

Knopfp7 observes of people looking to the unelected judiciary for principled leadership.

Rather than premising her position on a mistrust of judges, or a belief that judges should never trump democratically-accountable actors, Hiebert echoes Tushnet's concept of policy distortion: "To assume that only judges can resolve Charter conflicts is not healthy for the polity, because it diminishes political responsibility to pursue important policy goals and may lead to the unnecessary use of non-ambitious or ineffective means to pursue these objectives.,,38

Hiebert attempts to locate her argument outside the dialogue metaphor, but her

"relational" approach is in most respects a refinement of Roach's conception of dialogue as "complementary roles." Consider, for example, the definition offered in the following passage:

35 Kelly (1999a). 36 Hiebert (2002). 37 Rainer Knopff (1998) "Populism and the Politics of Rights: The Dual Attack on Representative Democracy," Canadian Journai of Po/iticai Science 31: 683-705. 38 Hiebert (2002), xiii.

163 A relational approach assumes that both Parliament and courts have valid insights into how legislative objectives should reflect and respect the Charter's normative values. Yet their judgments may be different. The benetits of conceiving Charter judgment in relational terms arise from the responsibility each body incurs to respect Charter values, from the exposure to judgments made by those differently situated, and from the opportunity to reflect upon the merits of contrary opinion .... Thus, a relational approach is informed by the assumption that parliamentary and judicial judgments be guided by a degree of mode sty about the superiority of their conclusions and by respect for the other' s contrary interpretation. Respect is measured by a sincere effort to understand the reasons and motivations that led to a contrary assessment, even if each ultimately disagrees with the other. 39

With respect, institutional actors involved in a relationship characterized by an exchange of opinions and, hopefully, mutual understanding, education and respect is the very definition of dialogue, albeit a thicker one than simply "talking."

Hiebert improves upon Roach's model by incorporating two existing arguments about judicial review. The tirst is a criticism of the court' s institutional capacity to make sound public policy, especially in comparison to the combined executive and legislative branches. As Horowitz,40 Manfredi,41 and Bogart42 have argued previously (but which

Hiebert notably fails to acknowledge), the bureaucracy and elected officiaIs are simply better at making public policy, thanks to their stronger connections to civil society, better appreciation of the complexities of balancing rights and societal needs, and access to policy and social science expertise. These factors underlie her suggestion that the

39 Hiebert (2002), 52. 40 Donald L. Horowitz (1977a) The Courts and Social Policy (Washington, DC: The Brookings Institution). 41 Christopher P. Manfredi (1994a) "'Appropriate and Just in the Circumstances': Public Policy and the Enforcement of Rights Under the Canadian Charter of Rights and Freedoms," Canadian Journal of Po/itical Science 24: 289-307. 42 W.A. Bogart (1994) Courts and Country: The Limits ofLitigatiol1 and the Social and Po/itical Life of Canada (Toronto: Oxford University_ Press).

164 judiciary should proceed with caution when evaluating the policy instruments chosen by

the government.43

The second argument, normally associated with conservative critics of judicial

activism, is that courts should exercise greater deference to legislation that does not

violate the Charter' s "core values" of political free speech, freedom of religion, freedom

of association, the rule of law, and basic procedural equality.44 Hiebert writes, "Section 1

analysis does not, and was not intended to, replace a court's criticaljudgment ofwhether

fundamental rights have been violated .... [T]he court should consider distinguishing core

from more marginal rights daims when evaluating the justification of impugned

legislation."45 Otherwise, "ruling that a right has been infringed becomes a hollow

gesture if legislation is easily sustained under the reasonableness evaluation," and "[the]

public and political officiaIs may not be duly attentive to the implications of policy

choices for fundamental rights.,,46 Conversely, whenjudges rule that core values have been violated, Parliament should exercise greater caution "before pursuing a course of action that is inconsistent with the spirit of the judicial ruling. ,,47 By the same token, governments should be more willing to use the s.33 override, and to openly defend such use, when the basis of judicial invalidation strays from these core values. In a refreshing turn, Hiebert neither categorically defends nor praises the override, but admits it can be

43 Hiebert (2002),60,65-67. 44 Hiebert (2002),57. Knopff and Morton identify this style of judicial review as the "intemallimits" approach, and the Court's CUITent practice ofreading rights broadly and dealing with limits only in a separate, second stage of analysis (under s.l) the "extemallimits" approach (1992, 42). 45 Hiebert (2002),69. 46 Hiebert (2002), 69. 47 Hiebert (2002),56.

165 used both appropriately and inappropriately; the more open and appropriate its use, she

contends, the more likely it is to gain public legitimacy.

Notably, Hiebert does not explicitly address the implication ofthis reliance on

"core" values. As Knopff and Morton observe, "the wording of the Charter is a tolerably

clear expression of the regime's core principles," but the kinds of questions most likely to

arise for judicial review under the Charter do not concern such basic challenges to the

'central core' ofliberal democratic principles. The bulk of Charter jurisprudence turns on

questions encountered 'as we move out from the central core ofthese values' toward the periphery," where the Charter's meaning is "arnbiguous and indeterrninate."48 In light of this, the courts should find violations ofrights only rarely, and invalidations should be accordingly uncommon.

Notwithstanding these refinements, Hiebert's approach suffers from two major shortcomings. First, like Roach and Hogg and Bushell, Hiebert ultimately relies on s.33 as the means of countering judicial hegemony, which, while defensible in terrns of democratic theory, is an unlikely saviour in practice (notably, section 1 cannot be considered a vehicle for the executive/legislative "last word," since the criteria and application of s.l are completely controlled by the judiciary). As Manfredi observes, s.33 is not a viable recourse, because the override, rightly or wrongly-and there is a very strong case for wrongly-is seen as illegitimate.49 Although Tsvi Kahana's recent work reveals that s.33 has actually been used 16 times, his analysis reinforces Manfredi's

48 Knopffand Morton (1992),147, citing Peter H. Russell (1983) "The Political Purposes of the Canadian Charter of Rights and Freedoms," Canadian Bar Review 61: 44-45. 49 M.~nfredi (2001).

166 conclusion. Kahana finds that aU but one use of the override have been pre-emptive, and

"inaccessible" or "invisible"-that is, dealing with a technicaUy-complex policy issue, or with an issue that was not on the public agenda prior to s.33 's enactment.50 The one exception-the Quebec government's 1988 shielding of the French-only sign law following the Supreme Court's decision in Ford-sparked a backlash throughout English

Canada that helped kill the Meech Lake Accord and delegitimized the use of s.33. In short, since 1988, no government has been willing to use the override in high-profile, politicaUy salient cases. This was graphicaUy demonstrated by the Alberta government's hasty retreat from the override on the forced sterilization compensation issue when faced with criticism from the national media and angry constituents. 51

Despite these impediments to the use of s.33, few of its advocates cited here offer a detailed programme for its rehabilitation in the court of public opinion (Manfredi 1993;

2001 is a notable exception). Such rehabilitation would require public recognition that rights are neither absolute nor clearly defined, and greater public appreciation of the complexities of constitutional democratic theory-a taU order indeed, but not impossible.

At any rate, the point remains, that hinging a normative theory which supports constitutional judicial review on s.33 requires engaging the existing factors limiting the override's use. While Hiebert appears to recognize the need to re-legitimize s.33,

50 Tsvi Kahana (2001) "The Notwithstanding Mechanism and Public Discussion: Lessons from the ignored practice of section 33 of the Charter," Canadian Public Administration 44(3): 257. 51 Kahana (2001), 271-272. In 1998, after one victim. Leilani Muir, sued the govemment of Alberta and was awarded $740,000 in compensation by the courts, the govemment tabled Bill 26, containing a notwithstanding clause, which would have limited its Iiability to $150,000. As Kahana reports, "it took the Alberta govemment less th an twenty-four hours ofmainly provincial pressure to back off from its plan to enact the steri.lization bill." (271)

167 Roach's and Hogg and BusheIl's failure to address this issue seriously weakens their advocacy of the override.

The second is her suggestion that elected officiaIs should defer to the judiciary on whether a right has been violated, a definitional task she describes as the "raison d'être for judicial involvement."52 If determining whether a violation has occurred was a simple matter of enforcing a right with a clear, self-evident meaning, her position would be quite convincing; unfortunately, this is not the case. So long as rights require interpretation- which, ironicaIly, Hiebert acknowledges53-it is not at aIl clear that elected officiaIs should surrender the important act of defining rights to the judiciary; indeed, we are back to square one of the debate over the legitimacy ofjudicial review. Moreover, her argument in this regard appears to employ the same means/ends distinction invoked by

Hogg and Bushell, which Morton and others persuasively rebut.

In spite of the dialogue metaphor' s apparent appeal, it suffers from several serious problems, in particular its advocates' reliance on the virtually dormant override provision.

As such, the most fundamental issue ofwhich institution's interpretation should prevail when "each ultimately disagrees with the other" is still unclear. Critics of judicial power who adhere to the theory of coordinate construction have an answer to this question- namely, the elected branches-but their approach reduces j udicial review to little more than an advisory opinion that may be ignored. On those issues where majorities are discriminating against unpopular minorities, the coordinate construction model legitimizes government behaviour that is inconsistent with liberal democracy's core

52 Hiebert (2002), 51. 53 Hiebert (2002),55.

168 principles. As noted earlier, such situations may be relatively rare, but the reservation is an important one, since it arises in the "hard cases" for liberal democracy.

A final observation about critics of the dialogue metaphor: Manfredi, Kelly and

Morton are right to criticize treating any legislative response as "dialogue," but by requiring sorne form of legislative noncompliance they define away the potential for genuine agreement resulting from inter-institutional discourse. Such agreement is certainly conceivable, given the potential for value change within the executive and legislative branches, and the practice of internally screening proposed legislation for potential Charter conflicts. This is particularly the case when the judicially-remedied legislation is relatively old. The irony is that historical and conventional characterizations of dialogue in the political context-for example, J.S. Mill's view of discourse in On Liberty and Considerations on Representative Government, Jürgen

Habermas's the ory of communicative action,54 the idea of diplomatic "dialogue," and aboriginal sentencing circles-emphasize its educative function and ability to promote compromise. Unfortunately, it is unclear how to operationalize this ontological distinction empirically; simply put, genuine agreement and grudging compliance "look" identical. In the absence of a satisfactory methodological solution, the appropriate course of action for researchers is to count as evidence of dialogue only those legislative responses which overtly signal government participation in a meaningful discourse. This

54 Jürgen Habermas (1984) The Theory ofCommunicative Action, translated by Thomas McCarthy (Boston: Beacon Press). For a more practical application of Habermas's theory, see Seyla Benhabib (1989) "Liberal Dialogue Versus a Critical TheOl·Y of Discursive Legitimation," Liberalism and the Moral Lijè, ed. Nancy L. Rosenblum (Cambridge: Harvard University Press). See also Stephen Holmes (1988) "Gag Rules or the Politics of Omission," Constitutionalism and Democracy, eds. Jon Elster and Rune Slagstad (Cambridge University Press), who criticizes democratic discourse's potential to promote social dissensus.

169 requires sorne element of creativity in the legislative response, such as reversaI,

modification or avoidance. 55

II. Empirical Analysis: Charter Dialogue with Lower Appeal Courts

The review of the empirical dimension of the dialogue debate reveals three related

shortcomings in the literature: first, an exclusive focus on the Supreme Court of Canada,

or, in the one exception (Hogg and Bushell), the findings are rendered meaningless by

fatal methodological flaws; second, as Manfredi and Kelly note but do not pursue, a

failure to include government appeals from lower courts as evidence of dialogue; and

third, an exclusive focus on responses to invalidations, ignoring cases involving judiciaI

amendment. In this section, 1 begin to fill the se gaps, by examining the federal

government's sequels to penultimate appeal court rulings in Charter cases. Following

from the discussion in Part l, 1 identify six possible outcomes to lower appeal court

invalidations or judicial amendments: "positive" or creative legislative response; appeal to the Supreme Court of Canada, or positive litigative response; repeal; replacement or amendment which simply complies with the court's decision ("ventriloquism"); "pre- emption," where the legislative "response" predates the lower court ruling; and no response.

55 Parliament's response to the Federal Court of Appeal's invalidation of the Incarne Tax Act's "inclusion/deduction" scheme for child support payments in Thibaudeau is a good example. Despite securing a reversai ofthis decision in the Supreme Court of Canada, the federal govemment repealed the inclusion/deduction scheme-which disadvantaged over a third of custodial parents, including Thibaudeau-and replaced it with a system of child tax credits and other benefits. In other words, the govemment used alternative policy instruments to pursue the core objective of tinancially aiding families with children. Specifically, Ihis demonstrates a (complex) strategy of"moditication."

170 The first step in this analysis is to isolate legislative responses to the lower court

from those to the Supreme Court. Accordingly, 1 exclude responses which followed a

loss in the Supreme Court, but include responses to appeal court decisions when any of

the following conditions were met:

(1) there was no appeal to the Supreme Court; (2) the response preceded the Supreme Court ruling; (3) the government won in the Supreme Court.

In addition, 1 exclude legislative responses to lower court rulings which are subsequently

struck down by the Supreme Court. 56 Together, these criteria also address Manfredi and

KeIly's criticism that "lower court decisions are simply too unstable to provide

unequivocal evidence of dialogue."57 The lower court's decision is "stable" for analytical

purposes if there is no subsequent appeal to the SCC, or if the response is independent of

the Supreme Court's ruling.

The database of aIl federal government losses in the highest federal and provincial

courts of appeal (not including the seC) yielded 45 cases involving an invalidation or judicial amendment. Counting a single legislative response where multiple cases dealt with the identicallegal issue,58 these cases generated 29 outcomes meeting the criteria

listed above, and consisted of seven "positive" legislative responses, Il cases where the only response was an appeal to the SCC, two repeals, four instances of "ventriloquism,"

56 There was only such response in the dataset, following M.N.R. v. Kruger, [1984] 2 F.C. 535 (F.C.A.); Vespoliv. The Queen, [1984] 55 N.R. 269 (F.C.A.). The response was nullified by the SCC in Baron v. Canada, [1993] 1 S.C.R. 416. 57 Manfredi and Kelly (1999),517. 58 To prevent statistical bias, such cases were combined regardless ofwhether the response was "positive" or "negative."

171 two pre-emptions, and two cases with no response of any kind; the cases are summarized

by category in Table 6.1., and legislative responses detailed in the Appendix to this

chapter.

Two important conclusions are immediately evident. The first is the

predominance of positive responses in the form of appeals when the government pursued

no other response (11/18, or 61 %). The finding highlights the weakness of Hogg and

Bushell's model, which excludes this form of dialogue. Notably, this response was

highly effective-that is, the Supreme Court reversed the lower court-in seven of Il

cases.59 Second, the proportion of "positive" responses (legislative and litigative, at

18/29, or 62%) is significantly higher than Manfredi and Kelly found with the Supreme

Court. My data support their conclusion that "Charter dialogue is less representative of

the Supreme Court than it is oflower courtS.,,60

While the same can be said with respect to legislative responses alone (positive

responses, repeals, and ventriloquism only), the discrepancy between the two levels of

courts is smaller. Positive responses constituted half (7/14) of such outcomes, compared to Manfredi and Kelly's finding of 33% for the SCC. The figure for lower Courts of

Appeal is probably generous, however, as the positive responses to two cases, R. v. Rao and R. v. LaPlante, came 13 and 9 years later respectively, shortly after similar rulings on

59 [fwe applied the same strict standard that dialogue requires "modification, avoidance or reversai" for appeals, only those appeais resuiting in Supreme Court reversai of the iower court wou id technically quai ify as "positive" responses. However, uniike "negative" iegisiative responses, which simpiy comply with the court's decision, appeais atlempt to modify, avoid, or reverse the iower court's ruling. It is therefore incorrect to treat unsuccessfui appeais the same as negative iegisiative responses. 60 Manfredi and Kelly (1999), 521.

172 Table 6.1: Charter Dialogue with Courts of Appeal, by Category

POSI rIVE RESPONSES NEGA TIVE RESPONSES

1. Section(s) creatively arnended or replaced (n=7) 3. Section(s) repealed (n=2)

*R. v. Ingebrigtson, [1990] 76 D.L.R. (4th) 481 (C.M.A.C.) International Fzmdfor Animal Welfare v. Canada, [1989] 1 F.C. 335 . Luseher v. Deputy Minister, Revenue Canada, Customs & Excise, [1985] 1 (F.C.A.) F.C. 85 (F.C.A.) Tétreault-Gadoury v. Canada, [1989] 2 F.C. 245 (F.C.A.)t * R. v. LaPiante, [1988] 48 D.L.R. (4th) 615 (Sk.C.A.) *R. v. Rao, [1984] 9 D.L.R. (4th) 542 (O.C.A.) 4. Section(s) amended or replaced to corn ply with Court of Appeal *Schachter v. Canada, [1990] 2 F.C. 129 (F.C.A.)t ("ventriloquisrn") (n=5) th S'omerville v. Canada (Attorney General), [1996] 136 D.L.R. (4 ) 205 : (A.C.A.) Canada (AG) v. Goguen (sub nom Goguen v. Shannon), [1989] 50 C.C.C. , Thibaudeau v. Minis/er of National Revenue, [1994] 2 F.C. 189 (F.C.A.)t (3d) 45 (N.B.C.A.) *Haig v. Canada, 94 D.L.R. (4th) 1 (O.C.A.) 2. Appealed to Suprerne Court (Exclusive1y) (n=ll) Pacifie Press v. Canada, [1991] 2 F.C. 327 (F.C.A.); Armadale Communications Ltdv. Acijudicator; [1991] 3 F.C. 242 (F.C.A.) th Canadav. Hydro-Quebee, [1995] 67 Q.A.C. 161 (Q.C.A.) R. v. Noble, [1985] 14 D.L.R. (4 ) 216 (O.C.A.) ;j *CanadianNewspapersv. A.G. Canada, [1985] 16 D.L.R. (4th) 642 *Rosenberg v. Canada, [1998] 158 D.L.R. (4th) 664 (O.C.A.). (O.C.A.) *Corbiere v. Canada (Minister of Indian and Northern Affairs) [sub nom NON-RESPONSES Batehel1'ana lndian Band (non-residents) v. Batchewana lndian Band], r1997] 1 F.C. 689 (F.C.A.). 5. Section(s) amended before Court of Appeal decision ("pre Del ZOllo v. Canada, [1997] 3 F.C. 40 (F.C.A.) emption") (n=2) *El Zein v. The Queen [sub nom Cotroni], [1987] 29 C.C.C. (3d) 560 (Q.C.A.) F.K. Clayton Group Ltd v. Minister of National Revenue, [1988] 2 F.C. 467 Lavallee, Rackel and Heintz v. Canada, [2000] 184 DLR (4th) 25 (F.C.A.) (A.C.A.) R. v. Lauzon, [1999] 129 C.C.C. (3d) 399 (C.M.A.C.); R. v. Bergeron, Osborne v. Treasury Board of Canada [sub nom MilIar], [1988] 3 F.C. [1999J 136 c.e.e. (3d) 327 (C.M.A.C.) 219 (F.C.A.) R. v. Higgins; R. v. Beare, [1987] 34 C.C.C. (3d) 193 (Sk.C.A.) 6. No response (Iegislative or Iitigative) (n=2) *R. v. Monne)', [1998] 153 D.L.R. (4th) 617 (O.C.A.) th * R. V. RI/zic, [1998] 128 C.C.C. (3d) 97 (O.C.A.) Gernhartv. Canada, [2000] 181 D.L.R. (4 ) 506 (F.C.A.) th * White, Ollenheimer and Baker v. Canada, [1998] 128 C.C.C. (3d) 97 R. v. Parker, [2000] 188 D.L.R. (4 ) 385 (O.C.A.) (O.C.A.) * case involved jlldicial amendment (n= 12) t denotes case also appea1ed to SlIpreme Court (n=3) the issue by the Supreme Court in three other cases. 61 As an aside, Rao demonstrates one

of the reasons the government and the courts are not equal partners in dialogue via

litigation: despite addressing a novel constitutional issue with important implications for

police procedure, the SCC refused the Crown's request for leave to appeal in Rao. The

Court's almost complete control over its docket must be kept in mind when assessing the

capacity for this form of dialogue.

My findings contradict Hogg and BusheU's assertions that legislative responses typically require little time and only minor changes. Regarding the "timeliness" of legislative sequels (appeals to the Supreme Court must be filed within 60 days of the lower court' s ruling), there was a significant difference between "positive" and

"negative" responses. The findings are summarized in Table 6.2, which employs the time increments used by Hogg and Bushell. AU complying responses (repeals and ventriloquism) took less than five years, and most less than three. In contrast, most positive responses came after at least three years, with three taking more than 5 years.

Similarly, while negative responses (other than simple repeals) typicaIly involved minor amendments, the positive responses, with one exception (Luscher), aIl entailed major policy revisions. The responses to Thibaudeau, Rao and LaPlante have already

61 R. v. Grant, [1993] 3 S.C.R. 223; R. v. Wiley, [1993] 3 S.C.R. 263; R. v. Plant, [1993] 3 S.C.R. 281. Notably, Manfredi and Kelly miscoded these cases as ventriloquism. The decisions, as in Rao, nullified the provision of the Narcotic Control Act authorizing warrantless searches ofnon-dwelling hou ses (for example, businesses and cars). The NCA (and Food and Drugs Act, s.42 in LaPlante) was repealed and replaced in 1996 with the Controlled Drugs and Substances Act, and the distinction between dwelling houses and non-dwelling houses was dropped with respect ta warrant requirements for "peace officers" (s.e. 1996, c.19, s.II). However, the 1996 Act created a new classification of enforcement official, "Inspector," who is empowered to enter and search "non-dwelling houses" without a warrant (s.e. 1996, c.19, ss.30-31). Although dealing with a viltually identical legal issue, 1 counted Rao and LaPlante separately because they dealt with two separate pieces of legislation.

174 Table 6.2: Timing of Legislative Responses

Response Type <3 Years 3-5 Years >5 Years

Positive Responses 2 2 3

Negative Responses* 4 2 0

Totals 6 4 3

*Information was unavailable for one case. been mentioned. In addition, fo11owing the court's finding in Ingebrigtson that Standing

Courts Martial violatedjudicial independence, the Department of National Defense re organized the entire military judges section, and created an independent commission to oversee staffing and reappointment. Parliament significantly altered third-party election advertising regulations following Somerville, and expanded parentalleave benefits while reducing individual claim amounts after Schachter.

On a related note, the evidence contirms the earlier hypothesis that it is markedly more difficult to respond legislatively to judicial amendments than invalidations (Table

6.3). All but one positive legislative response to judicial amendment took over five years

(8, 9 and 13), and two ofthese (Rao and LaPlante) could be coded "no response," as explained above. As we11, a11 four responses required major legislative revisions. In contrast, none of the three positive responses to invalidation took over five years, and one required only minor revisions. A similar pattern emerged for negative responses, as compliance with invalidations took less than three years, and judicial amendments between three and five years. The same conclusion is suggested by the types of judicial

175 Table 6.3: Timing of Legislative Responses, by Type of Judicial Remedy

Remedy Type Response Type <3 Years 3-5 Years >5 Years

Positive Responses 1 2 0 Invalidations Negative Responses* 4 0 0

Positive Responses 1 0 3 Amendments Negative Responses 0 2 0

Totals 6 4 3

*Information was unavailable for one case. remedies the government appealed rather than addressed through legislation. Of 12 judicial amendments, six were appealed (50%), versus five of 17 invalidations (29%); that is, judicial amendments were appealed at almost twice the rate as invalidations.

The final issue to consider is the degree to which the decision not to appeal may be explained by the presence of a positive legislative response or positive pre-emption. It is possible that governments decide not to appeal when they are able to craft a legislative response that modifies, reverses or avoids the Charter violation found by the court, or if the decision is rendered moot by a creative "prequel" (although the latter situation still does not qualify as "dialogue").62 Because appeals to the Supreme Court must be filed

62 Legislative responses which comply with the lower court cannot be considered an "explanation" for the decision not to appeal, but simply another indicator of the government's decision to concede.

176 within 60 days of the lower court's decision,63 legislative responses must occur fairly quickly for this explanation to be credible. Based on these criteria, government legislation may explain only a handful of the decisions not to appeal. As detailed in

Appendix 6.1, Ottawa responded creatively to Luscher and Minister ofNational Revenue v. Kruger/Vespoli v. The Queen64 within two years, and was in the pro cess of amending the legislation at issue in R. v. Lauzon and R. v. Bergeron.65

However, the foregoing hypothesis is based on the mistaken assumption that legislative and litigative responses to lower court rulings are mutually exclusive strategies, when this is not the case. Both strategies were employed in three cases,

Tétreault-Gadoury, Thibaudeau, and Schachter. In Tétreault-Gadoury, the government complied with the lower court by repealing the age restriction on claiming unemployment insurance, but simultaneously appealed to the Supreme Court. Similarly, after Schachter,

Parliament immediately recognized in legislation yet appealed the Federal Court of

Appeal's reading in of "natural parents" to the Unemployment Insurance Act provision allowing adoptive parents to divide parentalleave benefits. In both Tétreault-Gadoury and Schachter, the Court allowed Ottawa's appeals on technical grounds while ruling against the government on the substantive Charter issues (which, nevertheless, reversed the lower courts' remedies). In contrast, Ottawa successfully appealed the Charter issue to the Supreme Court in Thibaudeau, but subsequently repealed the inclusionldeduction

63 Supreme Court Act, R.S. 1985, c.S-26, s.58(1)(a). 64 [1984] 2 F.C. 535 (F.C.A.); [1984] 55 N.R. 269 (F.C.A.) 65 In addition, the govemrnent rnay have chosen not to appeal the invalidation in Somerville of election advertising and third-party spending regulations in light of the Supreme Court's obiter dicta shortly thereafter in Libman v. Quebec (A. G.), [1997] 3 S.C.R. 569 that it would have upheld sorne regulation. Four years after Somervi/le, Ottawa did introduce new (albeit rnuch looser) regulations.

177 scheme for taxation of child support payments.

The explanation for pursuing both strategies turns heavily on the concept of micro-constitutional politics discussed in Chapter One. In aU of these cases, the government appealed to challenge interpretive rules developed by the lower courts. In

Tétreault-Gadoury, the AG Canada urged the Court to exercise a lower degree of scrutiny regarding age discrimination claims, on the grounds that unlike s.15' s other enumerated grounds, age is not an immutable personal characteristic.66 In Thibaudeau, Ottawa discouraged thick systemic equality analysis, and instead advocated judicial deference to complex socio-economic policy, particularly in the area of taxation. As part ofthis argument, the federal government criticized judicial review which focuses on only individual sections ofbroader policies. And in Schachter, the AG Canada argued, unsuccessfully, that the judicial remedy power did not include the authority to "read in," but was limited to nullification. Micro-constitutional politics also helps explain the

Court's decision to hear the two moot cases, as they provided an opportunity for the

Justices to render judgment on these interpretive issues. These cases therefore underline the calculated behaviour of both j udicial and governmental actors.

III. Conclusion

My goal in this chapter has been two-fold. The first has been to examine the use of the dialogue metaphor and its implications for the appropriate role of judges and

66 Carole Bureau and Gaspard Côté, Attorney General of Canada Factum in Tétreault-Gadoury v. EIC, 29-30.

178 elected governments. My review of the literature to date reveals that much of the

"dialogue" debate is really a disagreement over what constitutes a valid governmental response, with critic persuasively rejecting Hogg and Bushell's broad criterion of

"any action" in favour of requiring sorne creativity on the government' s part. A second key point of contention regards whether the elected branches may legitimately attempt to interpret constitutional rights (as opposed to just reacting to judicial pronouncements), with advocates of coordinate construction at odds with oracular legalists who assume majoritarian governments would ignore rights without judicial intervention.

Second, 1 have expanded the focus of the dialogue debate in three related directions. The first is to extend systematic analysis of legislative-executive responses to lower appeal courts. My findings suggest that there is, in fact, a fairly robust level of inter-institutional dialogue at this level, although stilliess than Hogg and Bushell initially claimed. Furthermore, legislative responses may be greatly delayed and require major changes. This is particularly true when the court has rewritten rather than invalidated the law; including the hitherto-neglected scenario ofjudicial amendment is the dissertation's second novel contribution to the empirical dimension of the dialogue debate. The third involves a conscious redefinition of "dialogue" to include government litigation. Herein,

1 have analysed one such form of litigative dialogue with the lower courts, namely, appeals to the Supreme Court of Canada. The evidence indicates that this is the predominant form of dialogue with the judiciary, especially following judicial amendment. In view of this finding, future research should be directed to enhancing our understanding of substantive govermnent litigation strategies, such as intervention, where governments engage judges directly on issues of constitutional interpretation.

179 Appendix 6.1: Legislative Outcornes for Judicially Rernedied Laws67

R. v. Oakes; R. v. Carroll; R. v. Cook; R. v. o 'Day; R. v. Stanger and 3 otlter appeals; R. v. Landry; R. v. Stock [1983] 145 D.L.R. (3d) 123 (o. C.A.); [1983] 147 D.L.R. (3d) 92 (PE.I.S.C. en banco); [1983] 147 D.L.R. (3d) 687 (N.S.C.A.); [1983] 148 D.L.R. (3d) 371 (N.B.C.A.); [1984] 2 D.L.R. (4th) 121 (A.C.A.); [1984] 2 D.L.R. (4th) 518 (Q.C.A.); [1984] 10 C.C.C. (3d) 319 (B.C.C.A.); Law Affected: Reverse onus clause for drug trafficking: Narcotic Control Act, R.S.C. 1970, c.N-l, s.8. Charter Section Breached: s.11 (d). Legislative Sequel: Act repealed and replaced by Controlled Drugs and Substances Act, which substitutes reverse onus provision with absolute liability definition of trafficking based on quantity seized: S.C. 1996, c.19, s.11. [response to SCC] Litigative Sequel: appealed to SCC and lost.

Hunter, et al. v. Soutltam Inc. [1983] 147 D.L.R. (3d) 420 Law Affected: warrantless search and seizure provisions: Combines Investigation Act, R.S.C. 1970, c.C- 23, s.IO. Charter Section Breached: s.8. Legislative Sequel: A new procedure introduced for obtaining search warrants, requiring approval of Federal Court judge: Competition Act, S.C. 1986, c.26, s.13. [response to SCC] Litigative Sequel: appealed to SCC and lost.

Minister ofNational Revenue v. Kruger; Vespoli v. tlte Queen [1984] 2 F.C. 535 (F.C.A.); [1984] 55 N.R. 269 (F.C.A.) Law Affected: broadly-worded search and seizure provisions: Income Ta:" Act, R.S.C. 1970-71-72, c.63, s.23 1(4). Charter Section Breached: s.8. Legislative Sequel: The provisions were slightly amended, but retained broad authority: S.C. 1986, c.6, s.121. The new section was invalidated by the SCC in Baron v. Canada, [1993] 1 S.C.R. 416. Litigative Sequel: none

R. v. Rao [1984] 9 D.L.R. (4th) 542 (O.C.A.) Law Affected: warrantless search provision for non-dwelling houses: Narcotic Control Act, R.S.C. 1970, c.N-l, s.IO(I)(a). Charter Section Breached: s.8. Legislative Sequel: Act repealed and replaced by Controlled Drugs and Substances Act, which dropped distinction between dwelling houses and non-dwelling hou ses with respect to warrant requirements for "peace officers": S.C. 1996, c.19, s.ll. However, the 1996 Act created a new classification of enforcement official, "Inspector," who is empowered to enter and search "non-dwelling houses" without a warrant: S.c. 1996, c.19, ss.30-31. Litigative Sequel: request for leave to appeal to SCC denied.

Lusclter v. Deputy Minister, Revenue Canada, Customs and Excise [1985] 1 F.C. 85 (F.c.A.) Law Affected: prohibition on importing "immoral or indecent" books: Customs Tarif!, R.S.C. 1970, C.c- 41, S.14, Schedule C. tariff item 99201-1.

67 The list includes cases where the legislative response followed a ruling by the sec. As explained, such cases were excluded from analysis above.

180 Charter Section Breached: s.2(b). Legislative Sequel: when Customs TariffSchedules updated, "immoral and indecent" replaced by item 9956 which employs "obscene" or "hate propaganda" as defined by s.159(8) and s.281.3(8), respectively, of the Criminal Code, R.S.C. 1970, c-34 as am. by R.S.C. 1970 (pt Supp.), c.l1, s.l: S.C. 1987, c.49 to R.S.C. 1985 (3d Supp.), c.41, Schedule VII, item 9956. Litigative Sequel: none.

R. v. Noble th [1985] 14 D.L.R. (4 ) 216 (O.C.A.) Law Affected: search and seizure of dwelling-houses authorized by writs of assistance: Narcotic Control Act, R.S.C. 1970, c.N-l, s.10(1)(a); Food and Drugs Act, R.S.C. 1970, c.F-27, s.37. Charter Section Breached: s.8. Legislative Sequel: reference to writs of assistance removed: Criminal Law Amendment Act, 1985, S.C. 1985, c.19, ss. 195, 199, 200. This pre-empted the SCCs rulings in R. v. Sieben, [1987] 1 S.C.R. 295 and R. v. Hamill, [1987] 1 S.C.R.30l. Litigative Sequel: none.

Canadian Newspapers v. A.G. Canada [1985] 16 D.L.R. (4th) 642 (O.C.A.) Law Affected: publication ban on identity of sexual assault complainants: Criminal Code, s.442(3) as am. R.S.C. 1980-81-82-83, c.II0, s.74; idem, c.125, s.25. Charter Section Breached: ss.2(b), II(d). Legislative Sequel: none. Litigative Sequel: appealed to SCC and won.

El Zein v. The Queen [sub nom Cotroni v. U.S.A.] [1987] 29 C.C.C. (3d) 560 (Q.C.A.) Law Affected: extradition ofCanadians from Canada for international crimes (drug trafficking): Extradition Act, R.S.C. 1970, c.E-21. Charter Section Breached: s.6. Legislative Sequel: none. Litigative Sequel: appealed to SCC and won.

R. v. Higgins; R. v. Beare [1987] 34 C.C.C. (3d) 193 (Sk.C.A.) Law Affected: compulsory fingerprinting of people charged with indictable offense: Identification of Criminals Act, R.S.C. 1970, c.I-l, s.2 and related provisions of Criminal Code and Narcotic Control Act. Charter Section Breached: s.7. Legislative Sequel: none. Litigative Sequel: appealed to SCC and won.

F.K. Clay ton Group Ltd. v. Minister ofNational Revenue [1988] 2 F.C. 467 (F.C.A.) Law Affected: broad search and seizure power authorized only by Minister: Income Tax Act, R.S.C. 1970- 71-72, c.63, ss.231(1 )(d) and 231 (2). Charter Section Breached: s.8. Legislative Sequel: sections amended before Court of Appeal ruling (see MN.R. v. Kruger): S.C. 1986, c.6, s.121. The new section was invalidated by the SCC in Baron v. Canada, [1993] 1 S.C.R. 416. Litigative Sequel: none.

181 Osborne [Millar] v. Treasury Board of Canada [1988] 3 F.C. 219 (F.C.A.) Law Affected: limitations on federal public servants' participation in election campaigns and political parties: Public Service Employment Act, R.S.C. 1970, c.P-32, 5.32, as am. R.S.C. 1985, c.P-33, s.33. Charter Section Breached: s.2(b),(d) Legislative Sequel: none. Litigative Sequel: appealed to SCC and lost.

R. v. LaPlante [1988] 48 D.L.R. (4th) 615 (Sk.C.A.) Law Affected: warrantless searches of non-dwelling houses: Food and Drugs Act, R.S.C. 1970, c.F -27, s.37(1)(a), as am. R.S.C. 1985, c.F-27, sA2. Charter Section Breached: s.8. Legislative Sequel: repealed (S.c. 1996, c.19, 5.81) and replaced by Control/ed Drugs and Substances Act, which dropped distinction between dwelling houses and non-dwelling houses with respect to warrant requirements for "peace officers": S.C. 1996, c.19, s.ll. However, the 1996 Act created a new classification of en forcement official, "Inspector," who is empowered to enter and search "non-dwelling houses" without a warrant: S.c. 1996, c.19, ss.30-31. Litigative Sequel: none.

International Fundfor Animal Welfare v. Canada [1989] 1 F.C. 335 (F.C.A.) Law Affected: provision forbidding unlicensed persons from approaching within half a nautical mile of the seal hunt: Seal Protection Regulations, C.R.C., c.833, 5.11(6), as am. by SOR/78-167, 5.3. Charter Section Breached: s.2(b). Legislative Sequel: was repealed after decision, but specific information unavailable. Litigative Sequel: none.

Tétreault-Gadoury v. Canada (Employment and Immigration Commission) [1989] 2 F.C. 245 (F.C.A.) Law Affected: provision ending unemployment benefits at age 65: Unemployment Insurance Act, 1971, S.c. 1970-71-72, cA8, 5.31(1),(2),(4) as am. by S.c. 1974-75-76, c.80, 5.10. Charter Section Breached: s.15(1). Legislative Sequel: age limit removed: An Act to Amend the Unemployment Insurance Act, S.C. 1990, cAO, s.13. Litigative Sequel: appealed to SCC (despite legislative response) and lost on Charter issue (won on jurisdictional grounds).

A.G. Canada v. Goguen (Goguen v. Shannon) [1989] 50 C.C.C. (3d) 45 (N.B.C.A.) Law Affected: provision allowing judges to issue search warrant where they believe goods "may" be in premises: Customs Act, S.C. 1986, c.l, s.lll. Charter Section Breached: s.8 Legislative Sequel: amended section to read "are" instead of "may": S.C. 1992, c.l, s.143 (Schedule VI, item 9). Litigative Sequel: none

R. v. Wholesale Travel Group Ine. th [1990] 63 D.L.R. (4 ) 325 Law Affected: reverse on us to estab1ish "due diligence" defense to charge of false or misleading advertising, specifically, the requirement to publish a timely retraction: Competition Act, R.S.C. 1970, c.C- 23, $5.36(1)(a), 37.3(2), as am. R.S.C. 1985, c.C-34, 55.52, 60.

182 Charter Section Breached: s.7. Legislative Sequel: provision amended to add "recklessly" to offense, to rem ove problem of reverse onus associated with mens rea, and due diligence defense revoked: S.C. 1999, c.2, ss.12, 17.1. [response to SCC] Litigative Sequel: appealed to SCC and won on reverse onus, but lost on retraction requirement.

R. v. Ingebrigtson [1990] 76 D.L.R. (4th) 481 (C.M.A.C.) Law Affected: Standing Courts Martial violate judicial independence, and should only be used in times of emergency, as originally intended: Natianal Defense Act, R.S.C. 1985, c.N-5, s. 177. Charter Section Breached: s.ll(d). Legislative Sequel: the "emergency function" requirement was ignored, but the Department of National Defense re-organized the entire military judges section, and created an independent commission to oversee staffing and reappointment: R.S.C. 1998, c.35, s.42 [Standing Courts Martial now govemed by ss.173- 175]. Litigative Sequel: none.

Schachter v. Canada [1990] 2 F.C. 129 (F.C.A.) Law Affected: provision restricting matemity leave benefits to natural mothers, but allowing adoptive parents to div ide benefits: Unemplayment Insurance Act, 1971, s.c. 1970-71-72, c.48, s.32(1), as am. by s.c. 1980-81-82-83, c.150, s.5 and R.S.C. 1985, c.U-l, s.20. Charter Section Breached: s.15(1). Legislative Sequel: amended provision to allow both natural and adoptive parent to divide benefits, which were reduced: An Act to Amend the Unemployment Insurance Act, S.c. 1990, c.40, s.24. Litigative Sequel: appealed (despite legislative response) to SCC and won on reversing FCA's remedy of reading in, but lost on interpretive issue ofwhether courts cou Id ever read in.

Baron v. Canada [1991] 1 F.C. 688 (F.C.A.) Law Affected: provision appearing to give judges no discretion to deny search warrant request: Income Tax Act, R.S.C. 1970-71-72, c.63, s.231.3, as am. S.C. 1986, c.6, s.121 (see MN.R. v. Kruger above). Charter Section Breached: s.8. Legislative Sequel: provision amended to stipulate that judges "may" rather than "shall" grant a search warrant under conditions set out in the An Act ta Amend the Incame Tax Act: S.C. 1994, c.21, s.1 07. [response to SCC] Litigative Sequel: appealed to SCC and lost.

Pacific Press Ltd. v. Minister of Employment and Immigration Canada; Armadale Communications Ltd v. Adjudicator (Immigration Act) [1991] 2 F.C. 327 (F.C.A.); [1991] 3 F.C. 242 (F.C.A.) Law Affected: requirement that immigration hearings be held "in camera" (privately) unless proven that publicity wou Id not harm claimant or hislher family: Immigration Act, R.S.C. 1985, c. 1-2, s.29(3), as am. by R.S.C. 1985 (1st Supp.), c.31, s.99. Charter Section Breached: s.2(b). Legislative Sequel: section repealed and replaced to meet FCA's requirement that public hearings be the norm, and in camera only if judge feels it necessary: S.c. 1992, c. 49, 5.18. Litigative Sequel: none.

Belczowski v. Canada; Sauvé v. Canada [1992] 2 F.C. 440 (F.C.A.); [1992] 89 D.L.R. (4th) 644 (O.c.A.)

183 Law Affected: disqualification of prison inmates from voting in federal elections: Canada Elections Act, R.S.C. 1985, c.E-2, s.51(e). Charter Section Breached: s.3 Legislative Sequel: narrowed disqualification to inmates serving a sentence oftwo years or more (i.e., for an indictable offense): S.c. 1993, c.19, s.23(2). [response to SCC] Litigative Sequel: appealed to SCC and lost.

Haig v. Canada [1992] 94 D.L.R. (4th) 1 (O.C.A.) Law Affected: absence of explicit protections against discrimination on the basis of sexual orientation: Canada Human Rights Act, R.S.C. 1985, c.H-6, s.3(1); R.S.C. 1985, c.31 (1 st Supp.), s.3(1). Charter Section Breached: s.15( 1). Legislative Sequel: Act amended to add sexual orientation to protected grounds: S.C. 1996, c.14, s.2. Litigative Sequel: none.

Thibaudeau v. Minister ofNational Revenue [1994] 2 F.C. 189 (F.C.A.) Law Affected: provision requiring custodial parents to cIaim child support payments as taxable income (and allowing the payer to cIaim this amount as a deduction, although the latter provision was not challenged): Income Tax Act, R.S.C. 1970-71-72, c.63, s.56(1)(b) (and 60(b)). Charter Section Breached: s.11 (d). Legislative Sequel: despite securing a reversai in the SCC, the provisions were amended to exempt taxation of child support payments (see s.56.1(4)), effectively repealing the "incIusion/deduction" scheme: R.S.C. 1997, c.25, s.9(6). A new system ofchild tax credits and benefits were introduced to aid families with children. [response to FCA] Litigative Sequel: appealed to the SCC and won.

Canada v. Hydro-Quebec [1995] 67 Q.A.C. 161 (Q.C.A.) Law Affected: regulations regarding use of PCBs: Chlorobiphenyls Interim Order, P.c. 1989-296 and Canadian Environmental Protection Act, R.S.C. 1985, c.C-15.3, ss.34, 35. Charter Section Breached: s.7 (but case ultimately decided on federalism grounds). Legislative Sequel: none. Litigative Sequel: appealed to SCC and won.

Somerville v. Canada (A. G.) th [1996] 136 D.L.R. (4 ) 205 (A.C.A.) Law Affected: restrictions on advertising at the beginning and end of election campaigns ("black-out" periods) and on third-party spending for election advertising ($1000): Canada Elections Act, R.S.C. 1985, c.E-2, ss.213, 259.2(2). Charter Section Breached: s.2(b). Legislative Sequel: s.213 repealed and replaced with s.323(1) of the Canada Elections Act, S.C. 2000, c.9 which limits blackout to polling day only; third-party spending limit increased to $3000 per riding to a total of $150,000: S.C. 2000, c.9, s.350(1). Litigative Sequel: none.

Corbiere v. Canada (Minister ofIndian and Northern Affairs) [sub nom Batchewana Indian Band (non-residents) v. Batchewana Indian Band) [1997] 1 F.C. 689 (F.C.A.). Law Affected: reserve residency requirement for band council elections: Indian Act, R.S.C. 1985, c.I-5 (rep. and sub. R.S.C. 1985, c.32 (1 st Supp.), s.14). Charter Section Breached: s.15( J) in conjunction with s.25.

184 Legislative Sequel: none. Litigative Sequel: appealed to SCC and won.

Del Zotto v. Canada [1997] 3 F.C. 40 (F.C.A.) Law Affected: Ministerial authorization to conduct inquiry "into anything relating to the administration and enforcement of the Act," although the Act requires warrants: Incarne Tax Act, R.S.C. 1970-71-72, c.63, s.231.4 (as enacted by S.C. 1986, c.6, s.121). Charter Section Breached: s.8. Legislative Sequel: none. Litigative Sequel: appealed to SCC and won.

R. v. Ruzic [1998] 128 C.C.C. (3d) 97 (O.C.A.) Law Affected: restriction on defense of duress to wh en the accused faced a threat of immediate death or bodily harm, and the threatener was present when the offense was committed: Crirnina/ Code, R.S.C. 1985, c. C-46, s.17 (am. 1985, c.27 (1 st Supp.), s.40(2), Schedule l, item 1). Charter Section Breached: s.7. Legislative Sequel: none. Litigative Sequel: appealed to SCC and lost.

R. v. Monney [1998] 153 D.L.R. (4th) 617 (O.C.A.) Law Affected: provision empowering customs officers, based only a "reasonable suspicion," to detain travellers for "a reasonable period" aftertheir arrivai in Canada; in this case, court held provision did not authorize holding someone for several hours in a "drug 100" facility so as to obtain any drugs secreted internally: Custarns Act, R.S.C. 1985, c.l, s.98. Charter Section Breached: s.8. Legislative Sequel: none. Litigative Sequel: appealed to SCC and won.

Rosenberg v. Canada (A. G.) [1998] 158 D.L.R. (4th) 664 (O.C.A.) Law Affected: heterosexual definition of spouse: Incarne Tax Act, R.S.C. 1985, c.l (5 th Supp.), ss.l, 252(4). Charter Section Breached: s.15(1). Legislative Sequel: heterosexual definition removed: S.C. 2000, c.12, s.141(2). Litigative Sequel: none.

R. v. Lauzon; R. v. Bergeron [1999] 129 C.C.C. (3d) 399 (C.M.A.C.); [1999] 136 c.C.C. (3d) 327 (C.M.A.C.) Law Affected: limited renewable terros and hierarchical accountability for miIitary judges violate judicial independence of Standing Courts Martial: National Defense Act, R.S.C. 1985, c.N-5, s.I77. Charter Section Breached: s.11 (d). Legislative Sequel: the changes implemented after Ingebrigtson (see ab ove) predated the Court of Appeals' decisions in these cases: R.S.C. 1998, c.35, s.42 [Standing Courts Martial now governed by 55. 173-175]. Litigative Sequel: none.

Gernhart v. Canada th [2000] 181 D.L.R. (4 ) 506 (F.C.A.)

185 Law Affected: provisions requiring Minister to transfer copies oftax information to the Tax Court of Canada wh en taxpayer appeals, and making Tax Court documents available to the public: Income Tox Act, R.S.C. 1985, c.1 (5 th Supp.), s.176(1); Tax Court of Canada Rules (General Procedure), SOR/90-688, s.16. Charter Section Breached: s.8. Legislative Sequel: none (although the government conceded s.176(1) was "an historical aberration which no longer serves any useful function). Litigative Sequel: none.

Lavallee, Racket and Heintz v. Canada; White, Ottenheimer and Baker v. Canada [2000] 184 D.L.R. (4th) 25 (A.C.A.); [2000] 187 D.L.R. (4th) 581 (Nf.C.A.) Law Affected: provision allowing seizure from solicitors of legal records that might be protected by privilege (whole section nullified in Lavallee, et al.; only the requirement that solicitor "claim" privilege severed in White, et al.): Criminal Code, R.S.C. 1985, c.C-46, s.488.1 (enacted 1985, c.27 (1 st Supp.), s.71 ). Charter Section Breached: s.8. Legislative Sequel: none. Litigative Sequel: both cases appealed to SCC and lost.

R. v. Parker th [2000] 188 D.L.R. (4 ) 385 (O.c.A.) Law Affected: Ministerial discretion overmedical exemption from prohibition on possession and cultivation of marijuana: Controlled Drugs and Substances Act, R.S.C. 1996, c.19, ss.4(l), 7, 56. Charter Section Breached: 5.7. Legislative Sequel: none. Litigative Sequel: none.

186 CONCLUSION

The dissertation opened with a simple question: why does the Government of

Canada appeal its lower court losses to the Supreme Court of Canada in Charter of Rights cases? As a resuIt ofthis study, we can now formulate an (albeit incomplete) answer: federal government appeal decisions are largely calculated choices, based on governmental self-interest and strategic factors associated with the risks of losing on appeal. Generating this answer required ascertaining the institutionallocus of litigating authority within the federal government, and establishing the process for making appeal decisions. This concluding chapter reviews the dissertation's major findings on these issues, and closes by considering what remains to be discovered by future research in this area.

Officially since 1868, and in practice for the last three decades, the Department of

Justice Canada has exercised a near-monopoly over litigation involving the federal government. While this might suggest to the casual observer that government litigation is therefore centralised, the reality is considerably more complex. The Department is internally decentralised by region and issue specialisation, with the bulk of litigation handled by counsel in regional field offices. Surprisingly, and in marked contrast to the

U.S., it is not uncommon for these initial counsel to represent the government before the

Supreme Court of Canada. On the other hand, appeal decisions are made by senior

Department officiais in Ottawa (the National Litigation Committee), including, uItimately, the Attorney Generai of Canada. However, even here, regional officiaIs and

187 client departments participate, through representation on or consultation by the National

Litigation Committee. The resulting decision making process must therefore deal with a

variety of perspectives, interests, and demands.

Possibly for this reason, but, more importantly, because of the Supreme Court's

limited docket capacity, the federal government does not routinely appeal its Charter

related losses, nor has it at any point since 1982. That said, when years are grouped

according to party or leader, successive governments have been progressively more likely

to appeal to the Supreme Court. The differences, while not large, do raise the question of

why the federal government has become more willing to defend its policies, and whether

this trend is tied to the Charter-review process detailed by Kelly.1 That is, does the

review process enhance the government' s confidence that it can successfully defend

policy choices, perhaps at the section 1 "reasonable limits" stage? Or, is the pattern

merely a function of personality differences between successive Attorneys General? The

complete answer is, unfortunately, beyond the scope ofthis study.

Quantitative analyses, employing correlations, logistic regression, and probability

simulations, support the dissertation's conceptual framework and central assumption, that the decision to appeal is a rational one, made by government lawyers weighing the costs and benefits of appealing. AU of the hypothesized factors were supported by the data, albeit to varying degrees. "Reviewability," or the government's concern with securing leave to appeal, evidenced the weakest influence, as measured by the appeal court's citation ofjurisdictional conflict. This finding is consistent, however, with recent studies

Kelly (1998; 1999a).

188 of Supreme Court of Canada leave decisions,2 and the fact that the federal govemment is

typically granted leave. Even so, the categories of factors are not watertight, and the

measures related to other factors (for example, lower court dissent) may also help

convince the Supreme Court of Canada to hear a case. Short-term financial costs also

have a weak influence, as cri minaI cases are less likely to be appealed than civil cases,

where monetary costs are typically greater. Immediate policy costs exercise a far

stronger influence, but only where the government's legislation has been rewritten or

"amended" by the court, through the remedies of "reading in" or "severance." Among

the dissertation's most important findings is that less-intrusive judicial remedies

inc1uding nullification of laws-do not provoke government appeals. The simple fact of judicial activism is, therefore, less important to government decision-makers than the

form this activism takes, with the government challenging only the greatest incursions on

its policy authority.

The strongest factor is the case's salience, or importance, as evidenced by the presence of interveners and novel interpretations of legal rules which limit

legislative/executive authority or expand judicial power. Contrary to expectation, neither of the se measures is significant by itself, but only in conjunction with each other. More detailed analysis of these measures reveals that the intuitive explanation, that interveners persuade the court to render novel interpretations, is incorrect. Rather, the federal government is simply more likely to appeal when both factors are present. Although the precise reason why this is so remains unc1ear, there is sorne evidence that particularly

important cases-that is, those in which novel interpretations are likely-both attract

:2 Flerilfuing and Krutz (2002), 238, 241.

189 interveners and spur appeals. That is, intervention is an additional indicator of a case's

importance, but is not causally related to either novel interpretations or the decision to

appeal. Notably, the findings regarding these two factors reveal that the government's

de ci sion not to appeal after losing in Haig v. Canada was truly exceptional: of the ten

cases witnessing both intervention and novel interpretation, only Haig was not appealed.

The answer to the Haig puzzle most likely lies in the "winnability" factor, or the

government's assessment ofits odds ofwinning on appeal. One measure ofwinnability,

dissent on the lower appeal court, emerges as one of the statistically strongest factors.

Similarly, the prospect ofwinning-or the jurisprudential risk oflosing in the Supreme

Court-is a recurring factor in deeper qualitative analysis, even where the reason for

appealing or conceding is not immediately obvious. For example, every ruling involving judicial amendment that was not appealed was issued by a unanimous court (inc1uding

Haig), and where recentjurisprudence was not in the government's favour. Regarding

Haig, Justice Department officiaIs recalled believing the Supreme Court would uphold the Ontario Court of Appeal, and not appealing meant that the OCA' s ruling was technically restricted to the province of Ontario. Notably, these findings implicitly

contradict the argument put forth by sorne critics, that the Attorney General conceded in

Haig to promote the adoption of sexual orientation as an analogous ground under s.15. If this was the AG's primary motive, she shouldhave appealed, thereby allowing the

Supreme Court to establish such an interpretation with the broadest and strongest legal

application. The evidence suggests, instead, that government lawyers acted strategically

to minimize losses.

190 These findings are supported by simulations calculating the probability of a

government appeal given the presence of certain factors. An appeal is much more likely

in the presence of judicial dissent, judicial amendment, or novel interpretation in

conjunction with an intervener than in any ofthese factors' absence. Equally important is

the finding that these factors have a cumulative effect, as the presence of any combination

of these significantly increases the probability of an appeal; when aU three variables are

present, the probability increases by 81 %, to 98.5%.

These findings are consistent with those from the U.S., that governments are

'''procedurally rational' when they decide to interact with the Court.,,3 Zorn also finds, in

his study of the U.S. Solicitor General's appeal decisions, that multiple factors play a

role, and exercise a cumulative influence.4 With the caveat that Zorn's work inc1uded

non-constitutional and unpublished cases and our measures varied somewhat, interesting

and important differences exist between his findings and my own regarding specifie

factors. While simple invalidation ("less-intrusive remedy") and interveners (on their

own) did not achieve statistical significance in the Canadian context, these were two of

the three strongest influences on the U.S. Solicitor General; the third-intercourt jurisprudential conflict-was only marginally significant in Canada, and exhibited no

relationship when analysed in isolation with the dependent variable. In contrast, judicial

dissent, which was among the strongest factors in Canada, was not statistically significant

in Zorn's study. Thus, the only factors that had comparable effects in both contexts were the immediate financial costs (measured by the criminal-civil case distinction), which

3 Waltenburg and Swinford (1999), 255. 4 Zorn (1997; 2001).

191 exercised a weak influence, and the lower court reversaI indicator for winnability, which influenced governments in the direction opposite to that hypothesized. Future studies of

Canadian government appeal decisions should be extended to non-constitutional cases and, if possible, unpublished decisions, to permit a more accurate comparison with existing U.S. studies.

Existing studies of de ci sion making within institutions, including Zorn and

Horowitz's work on government lawyers,5 conclude that individuals are motivated by diflerent factors depending on their position within an organisation. The U.S. Solicitor

General, for example, occupies the top of the Justice Department's hierarchy, and is accordingly less motivated by immediate policy costs than lower-Ievel counsel who de al regularly with those agencies responsible for the policy in question. Conversely, the factors that most concern the SG-jurisprudential development, the government' s reputation with the Supreme Court, and the related issues of reviewability and winnability-are of less concern to lower-level counsel. Given the greater level of decentralisation within the Justice Department Canada's appeal decision process, it was unc1ear whether a similar pattern would emerge. However, the Canadian evidence suggests a more centralised process. Immediate policy costs are not significant, unless the court has usurped the legislative function by issuing highly intrusive remedies; the fact that the federal government' stop lawyer is also a Cabinet Minister and Member of

Parliament may help explain this finding. The lower appeal rate for criminal prosecutions compared to more-costly civil cases also indicates a centralist orientation, as does the

5 Horowitz (1977b); Zorn (1997; 2001).

192 strong influence of case importance and winnability. However, the lack of evidence in support of reviewability suggests a contrary interpretation. A more accurate assessment of the role of institutional position requires comparing the appeal recommendations of regional officiaIs with those of the National Litigation Committee; unfortunately, these internaI documents are not available to the researcher at this time.

Finally, the dissertation expands the contemporary debate over judicial-Iegislative dialogue under the Charter, in three ways. The first is to provide a systematic assessment of the degree of dialogue with lower appeal courts, in contrast to the deeply flawed treatment these court decisions received in the leading work by Hogg and Bushell.6 The second is to redefine "dialogue" to include government appeals to the Supreme Court of

Canada, which clearly signal the government's disagreement with the lower courts' rulings. The third s to examine responses to judicial amendment, and not just invalidation, contrary to existing studies. l find that there is a significant level of dialogue with the lower courts, primarily in the form of appeals. Legislative responses which meet

Hogg and Bushell's initial criteria ofreversing, modifying, or avoiding the court's ruling, while possible, tend to take a considerable amount oftime, and require major changes, particularly where the court has "amended" rather than invalidated legislation. Notably, legislative responses offer little explanation for the government's decisions not to appeal. lndeed, there are high-profile examples-Schachter, Thibaudeau, and Tétreault

Gadoury-ofthe federal government pursuing both a legislative and litigative response, usually to advance a micro-constitutional political strategy in the latter.

Before closing, it is important to consider what the dissertation does not tell us.

193 First, the explanation offered here for government appeal decisions is not exhaustive, as much variance in the decision remains unexplained. While calculated decision making is c1early evident, there is room in the results for other influencing factors, which may or may not be traced to rational motives. Second, although the dissertation reveals a great deal about the external factors which influence government appeal decisions, it tells us virtually nothing about the substantive reasons why the government appeals. To put this another way, deciding to appeal to the Supreme Court is one thing-what you argue when you get there is quite another. As Morton and Knopff observe, the Justice

Department may appeal only to concede the Charter violation, or to provide a weak (or no) section one defense. 7 An interesting avenue for future research, therefore, is to examine how often governments offer weak section one defenses, especially after the

Oakes decision. Such a strategy should be attractive to governments wishing to establish a politically unpopular or contentious legal rule-such as extending legal protection or benefits to gays and lesbians, or expanding French-language education facilities for minority francophones outside Quebec-at the highest judiciallevel possible.

On a related note, we still know very little about the government' s substantive micro-constitutional strategy with respect to the Charter. While the evidence regarding the AG Canada's tendency to de fend governmental authority (as opposed to specific government policies) is suggestive, what the government actually argues on appeal is an open question. Moreover, appeals are only one litigative scenario in which the government can advance micro-constitutional arguments, or for that matter, contribute to

6 Hogg and Bushell (1997). 7 Morton and Knoptf (2000), 117.

194 the judicial-governmental dialogue. Three other scenarios, the first two of which are

instituted by the government, occur when: the government "intervenes" in a case as a

third party; when the government initiates a reference, asking the Court for an "advisory

opinion"; and when the govemment appears before the Court as a respondent, that is,

involuntarily, as the target of litigation. AlI ofthese forms of govemment litigation have received scant scholarly attention to date, but a complete understanding of govemmental-both federal and provincial-behaviour under the Charter, and use of litigation to advance policy goals, demands their study.

195 BIBLIOGRAPHY

Atcheson, M. Elizabeth (1984) Women and Legal Action: precedents, resources and strategies for the future. Ottawa: Canadian Advisory Council on the Status of Women.

Baar, Carl (1986) "Using Process Theory to Explain Judicial Decision Making." Canadian Journal ofLaw and Society 1: 57-79.

Bakan, Joel (1997) Just Words: Constitutional Rights and Social Wrongs. Toronto: University of Toronto Press.

Bayefsky, Anne and Mary Eberts, eds. (1985) Equality Rights and the Canadian Charter ofRights and Freedoms. Toronto: Carswell.

Beatty, David (1987) Putting the Charter to Work: Designing a Constitutional Labour Code. Montreal: McGill-Queen's University Press.

Beatty, David (1990) Talking Heads and the Supremes: The Canadian Production of Constitutional Review. Toronto: Carswell.

Beatty, David (1995) Constitutional Law in Theory and Practice. Toronto: University of Toronto Press.

Beaulieu, Mélanie (2001) "Traitement Fiscal Des Pensions Alimentaires Pour Enfants: Les Modifications De 1997 Étaient-Elles Une Réponse Adéquate à L'arrêt Thibaudeau?" University of Toronto Faculty ofLaw Review 59: 319-347.

Benhabib, Seyla (1989) "Liberal Dialogue Versus a Critical Theory of Discursive Legitimation." Liberalism and the Moral Life. Ed. Nancy L. Rosenblum. Cambridge: Harvard University Press.

Bogart, W.A. (1994) Courts and Country: The Limits ofLitigation and the Social and Political Life ofCanada. Toronto: Oxford University Press.

Bork, Robert H. (1973) "The Problems and Pleasures ofBeing Solicitor General." A.B.A. Antitrust Law Journal 42: 701-706.

Brennan, William J., Jr. (1996) "How the Supreme Court Arrives at Decisions." American Government: Readings and Cases. 12th edition. Ed. Peter Woll, NY: HarperCollins.

196 Brigman, William E. (1966) "The Office of the Solicitor General of the United States." Ph.D. dissertation. University of North Carolina (Chapel Hill).

Brodie, Ian (1995) "The Charter and the Policy Process." Presented at the annual meeting of the Canadian Political Science Association. Montreal.

Brodie, Ian (2001) "Interest Group Litigation and the Embedded State: Canada's Court Challenges Program." Canadian Journal ofPolitical Science 34: 357-376.

Brunet, Mélanie (2000) Out ofthe Shadows: The Civil Law Tradition in the Department ofJustice Canada, 1868-2000. Ottawa: Department of Justice Canada.

Cairns, Alan C. (1986) "The Embedded State: State-Society Relations in Canada." State and Society: Canada in Comparative Perspective. Ed. Keith Banting. Toronto: University of Toronto Press.

Cairns, Alan C. (1991) Disruptions: Constitutional Struggles, from the Charter to Meech Lake. Ed. Douglas E. Williams. Toronto: McClelland & Stewart Inc.

Cairns, Alan C. (1992a) The Charter Versus Federalism: The Dilemmas ofConstitutional Reform. Montreal: McGill-Queen's University Press.

Cairns, Alan C. (1992b) "Reflections on the Political Purposes of the Charter: The First Decade." The Charter: Ten Years Later. Ed. G.A. Beaudoin. Cowansville, QC: Les Editions Yvon Blais.

Caldiera, Gregory A. and John R. Wright (1988) "Organized Interest and Agenda-Setting in the U.S. Supreme Court." American Political Science Review 82 (December): 1109-1127.

CapIan, Lincoln (1987a) "The Annals of the Law-The Tenth Justice (I)." The New Yorker (10 August): 29-58.

CapIan, Lincoln (1987b) "The Annals of the Law-The Tenth Justice (II)." The New Yorker (17 August): 30-62.

Carroll, Barbara Wake (1991) "The Structure of the Canadian Bureaucratie Elite: Sorne Evidence of Change." Canadian Public Administration 34: 359-372.

Chamberlain, Ronald S. (1987) "Mixing Politics and Justice: The Office of the Solicitor General." The Journal ofLaw and Politics 4: 379-428.

Chapman, Susan and John McInnes (1996) "The Role of the Attorney-General in . Constitutional Litigation: Re-Defining the Contours of the Public Interest in a

197 Charter Era." The Charter's Impact on the Criminal Justice System. Ed. Jamie Cameron. Toronto: Carswell.

Clayton, Cornell, ed. (1995) Government Lawyers: The Federal Legal Bureaucracy and Presidential Politics. Lawrence: University Press of Kansas.

Clayton, Cornell (1992) The Politics ofJustice: The Attorney General and the Making of Legal Policy. New York: M.E. Sharpe Inc.

Cox, Archibald (1963) "The Government in the Supreme Court." Chicago Bar Record 44: 221-29.

Crane, Brian A. and Henry S. Brown (1999) Supreme Court ofCanada Practice 2000. Toronto: Carswell.

Crosbie, John C. (1997) No Holds Barred: My Life in Politics. Toronto: McClelland & Stewart.

Dawson, Mary (1992) "The Impact of the Charter on the Public Policy Process and the Department of Justice." Osgoode Hal! Law Journal 30: 595-603. de Montigny, Yves (1997) "The Impact (Real or Apprehended) of the Canadian Charter of Rights and Freedoms on the Legislative Authority of Quebec." Charting the Consequences: The Impact ofCharter Rights on Canadian Law and Politics. Eds. David Schneiderman and Kate Sutherland. Toronto: University of Toronto Press.

Department of Justice Canada (1986) Toward Equality: The Response to the Report ofthe Parliamentary Committee on Equality Rights. Minister of Supply and Services.

Department of Justice Canada (2000) Federal Prosecution Service Deskhook. Ottawa: Department of Justice Canada.

Department of Justice Canada (2001) Serving Canadians: Report ofthe Federal Prosecution Service Review. Ottawa: Department of Justice Canada.

Dickson, Brian (1994) "The Canadian Charter of Rights and Freedoms: Dawn of a New Era?" Review ofConstitutional Studies 2: 1-19.

Dobrowolsky, Alexandra (1997) "The Charter and Mainstream Political Science: Waves ofPractical Contestation and Changing Theoretical Currents." Charting the Consequences: The Impact of Charter Rights on Canadian Law and Politics. Eds. David Schneiderman and Kate Sutherland. Toronto: University of Toronto Press.

Dobrowolsky, Alexandra (2001) The Politics ofPragmatism: Women, Representation, and Constitutionalism in Canada. Toronto: Oxford University Press.

198 Dotan, Y oav and Menachem Hofnung (2000) "Legal Defeats-Political Wins: Why Elected MK's Go to Courts." Presented at the annual meeting of the American Political Science Association, Washington, DC.

Duclos, Nitya (1992) "A Remedy for the Nineties: Schachter v. R. and Haig & Birch v. Canada." Constitutional Forum 4: 22-28.

Eberts, Mary (1986) "A Strategy for Equality Litigation Under the Canadian Charter of Rights and Freedoms." Litigating the Values ofa Nation: The Canadian Charter ofRights and Freedoms. Eds. Joseph Weiler and Robin Elliot. Toronto: Carswell.

Edwards, J.Ll.J. (1964) The Law Officers ofthe Crown: a study ofthe offices ofAttorney General and Solicitor-General ofEngland with an account ofthe office ofthe Director ofPublic Prosecutions ofEngland. London: Maxwell & Sweet.

Edwards, J.Ll.J. (1980) Ministerial Responsibility for National Security as it Relates to the Offices ofthe Prime Minister, Attorney General and Solicitor General of Canada. Commission of Inquiry Concerning Certain Activities of the RCMP [MacDonald Commission]. Ottawa: Supplyand Services Canada.

Edwards, J.Ll.J. (1984) The Attorney General, Politics and the Public Interest. London: Maxwell & Sweet.

Edwards, J.Ll.J. (1987) "The Attorney General and the Charter of Rights." Charter Litigation. Ed. Robert J. Sharpe. Toronto: Butterworths.

Edwards, J.Ll.J. (1995) "The Office of Attorney General: New Levels of Public Expectations and Accountability." Accountability for Criminal Justice: Selected Essays. Ed. Philip C. Stenning. Toronto: University of Toronto Press.

Eisenstein, James (1978) Counsel for the United States: u.s. Attorneys in the Political and Legal Systems. Baltimore: Johns Hopkins University Press.

Ely, John Hart (1980) Democracy and Distrust: A TheoryofJudicial Review (Cambridge, Mass.: Harvard University Press).

Epp, Charles R. (1998) The Rights Revolution: Lawyers, Activists, and Supreme Courts in Comparative Perspective. Chicago: University of Chicago Press.

Epstein, Lee (1999) "The Comparative Advantage." Law and Courts: Newsletter ofthe Law and Courts Section ofthe American Political Science Association 9 (3): 1,3- 6.

199 Epstein. Lee and Jack Knight (1999) "Mapping Out the Strategie Terrain: The Informational Role of Amici Curiae." Supreme Court Decision-Making: New Institutional Approaches. Eds. Comell W. Clayton and Howard Gillman. Chicago: University of Chicago Press.

Ferguson, Mary, ed. (1999-2000) Federal Guidebook: A Guide to the Canadian Federal Government and ifs Decision-Makers. Ottawa: J-K Carruthers.

Flanagan, Thomas (1997) "The Staying Power of the Status Quo: Collective Choice in Canada's Parliament after Morgentaler." Canadian Journal ofPo/itical Science 30: 31-54.

Flemming, Roy B. and Glen S. Krutz (2002) "Selecting Appeals for Judicial Review in Canada: a Replication and Multivariate Test of American Hypotheses." Journal of Politics 64: 232-48.

Fried, Charles (1991) Drder and Law: Arguing the Reagan Revolution-A Firsthand Account. New York: Simon & Schuster.

Gallagher, Sheila A. (1993) "The Public Policy Process and Charter Litigation." The Impact ofthe Charter on the Public Policy Process. Eds. Patrick Monahan and Marie Finkelstein. North York: York University Centre for Public Law and Public Policy.

Galanter, Marc (1974) "Why the 'Haves' Come Out Ahead: Speculations on the Limits of Social Change." Law & Society Review 9: 95-160.

Government Electronic Directory (2001) http://direct.srv.gc.calcgi-bin/wgweng. Ottawa: Public Works and Govemment Services Canada.

Green, Donald and Ian Shapiro (1994) Pathologies ofRational Choice Theory: A Critique ofApplications in Political Science. New Haven: Yale University Press.

Griswold, Erwin N. (1969) "The Office of the Solicitor General-Representing the Interests of the United States Before the Supreme Court." Missouri Law Review 34: 527-537.

Habermas, Jürgen (1984) The Theory of Communicative Action. Translated by Thomas McCarthy. Boston: Beacon Press.

Hall, Peter and Rosemary Taylor (1996) "Political Science and the Three New Institutionalisms: Review Article." Political Studies 44: 936-957.

Hausegger, Lori and Rainer Knopff(1994) "The Effectiveness oflnterest Group Litigation: An Assessment ofLEAF's Participation in Supreme Court Cases."

200 Presented at the annual meeting of the Canadian Political Science Association, Calgary.

Haynie, Stacia and C. Neal Tate (2000) "Comparative Judicial Politics in the Year 2000 and the Role of Research Collaboration between Indigenous and American Scholars." Presented to the annual meeting of the American Political Science Association, Washington, DC.

Hein, Gregory (2001) "Interest Group Litigation and Canadian Democracy." Judicial Power and Canadian Democracy. Eds. Paul Howe and Peter H. Russell. Montreal: McGill-Queen's Press.

Hennigar, Matthew A. (1996) "Litigating Pan-Canadianism: The Constitutional Litigation Strategy of the Canadian Federal Government in Charter Cases, 1982- 1993." MA Thesis. University of Calgary.

Hennigar, Matthew A. (1998) "Attorney General - Executive Relations in Charter Litigation." Presented at the annual meeting of the Canadian Political Science Association, Ottawa.

Herman, Didi (1997) "The Good, the Bad and the Smugly: Sexual Orientation and Perspectives on the Charter." Charting the Consequences: The Impact ofCharter Rights on Canadian Law and Politics. Eds. David Schneiderman and Kate Sutherland. Toronto: University of Toronto Press.

Hiebert, Janet L. (1994) "The Charter and Federalism: Revisiting the Nation-Building Thesis." Canada: The State ofthe Federation 1994. Eds. Douglas M. Brown and Janet Hiebert. Institute ofIntergovernmental Relations. Kingston: McGill-Queen's University Press.

Hiebert, Janet L. (1996) Limiting Rights: The Dilemma ofJudicial Review. Montreal: McGill-Queen's University Press.

Hiebert, Janet L. (2002) Charter Conflicts: What is Parliament's Role? Montreal & Kingston: McGill-Queen's University Press.

Ho, Shawn (1994) "The Macro- and Micro-Constitutional Strategies of Provincial Governments in Charter Politics: A Study of Alberta, Saskatchewan and Ontario, 1982-1992." M.A. Thesis. University of Calgary.

Hodgetts, J. E. (1973) The Canadian Public Service: A Physiology ofGovernment, 1867- 1970. Toronto: University of Toronto Press.

201 Hogg, Peter W. and Allison Bushell (1997) "The Charter Dialogue Between Courts and Legislatures (Or Perhaps the Canadian Charter of Rights Isn't Such a Bad Thing After AIl." Osgoode Hal! Law Journal 35: 75-124.

Hogg, Peter W. and Allison Thornton (1999) "Reply to 'Six Degrees of Dialogue.'" Osgoode Hal! Law Journal 37: 529-536.

Holmes, Stephen (1988) "Gag Rules or the Politics of Omission." Constitutionalism and Democracy. Eds. Jon Elster and Rune Slagstad. Cambridge University Press.

Horowitz, Donald L. (1977a) The Courts and Social Poliey. Washington, DC: The Brookings Institution.

Horowitz, Donald L. (1977b) The Jurocracy: Government Lawyers, Agency Programs, and Judicial Decisions. Lexington, Mass: Lexington Books.

Hutchinson, Allan C. (1995) Waitingfor CORAF: A Critique ofLaw and Politics. Toronto: University of Toronto Press.

Huscroft, Grant (1996) "The Attorney General and Charter Challenges to Legislation: Advocate or Adjudicator?" National Journal ofConstitutional Law 5: 125-162.

Immergut, Ellen (1998) "The Theoretical Core of the New Institutionalism." Politics & Society 26: 5-34.

Irons, Peter (1982) The New Deal Lawyers. Princeton: Princeton University Press.

Irons, Peter (1983) Justice at War. New York: Oxford University Press.

Jaccard, James (2001) Interaction Effects in Logistic Regression. Quantitative Applications in the Social Sciences Series, No. 135. Thousand Oaks: Sage Publications.

Jai, Julie (1997-98) "Policy, Politics and Law: Changing Relationships in Light of the Charter." National Journal ofConstitutional Law 9: 1-25.

Jenkins, John A. (1983) "The Solicitor General's Winning Ways." American Bar Association Journal 69: 734-738.

Kahana, Tsvi (2001) "The Notwithstanding Mechanism and Public Discussion: Lessons from the ignored practice of section 33 of the Charter." Canadian Public Administration 44(3): 255-291.

Kelly, James B. (1998) "Charter Activism and Canadian Federalism: Rebalancing Liberal Constitutionalism in Canada, 1982-1997." Ph.D. dissertation. McGill University.

202 Kelly, James B. (1999a) "Bureaucratie Activism and the Charter of Rights and Freedoms: the Department of Justice and its entry into the centre of government." Canadian Public Administration 42: 476-511.

Kelly, James B. (1999b) "The Charter of Rights and Freedoms and the Rebalancing of Liberal Constitutionalism in Canada, 1982-1997." Osgoode Hal! Law Journal 37: 625-695.

Kelly, James B. (2001) "Reconciling Rights and Federalism during Review of the Charter of Rights and Freedoms: The Supreme Court of Canada and the Centralization Thesis, 1982-1999." CanadianJournal ofPolitical Science 34: 321-356.

Knopff, Rainer (1998) "Populism and the Politics of Rights: The Dual Attack on Representative Democracy." Canadian Journal ofPolitical Science 31: 683-705.

Knopff, Rainer and F.L. Morton. (1985) "Nation-Building and the Canadian Charter of Rights and Freedoms." Constitutionalism, Citizenship and Society in Canada. Eds. Alan Cairns and Cynthia Williams. Toronto: University of Toronto Press.

Knopff, Rainer and F.L. Morton (1992) Charter Politics. Scarborough: Nelson Canada.

Kraus, Brent (1996) "Disentangling the Tax Unit: the Subsidization of Child Support after Thibaudeau." Saskatchewan Law Review 60: 295-314.

Law Reform Commission of Canada (1990), Controlling Criminal Prosecutions: The Attorney General and the Crown Prosecutor, Law Reform Working Paper 62, Ottawa: The Commission.

Lee, Rex E. (1986) "Lawyering for the Government: Politics, Polemics & Principle." Ohio State Law Journal 47: 595-601.

Liao, Tim Futing (1997) Interpreting Probability Models: Logit, Probit, and Other Generalized Linear Models. Quantitative Applications in the Social Sciences Series, Number 101. Thousand Oaks, CA: Sage Publications.

Lochner, Todd (1993) "The Relationship Between the Office of the Solicitor General and the Independent Agencies: A Reevaluation." Virginia Law Review 79: 549-582.

MacPherson, Jim (1986) "The Case for the Defendant Government in Charter Litigation." Litigating the Values ofa Nation: The Canadian Charter ofRights and Freedoms. Eds. Joseph Weiler and Robin Elliot. Toronto: Carswell.

203 Mandel, Michael (1994) The Charter ofRights and the Legalization ofPolitics in Canada (Revised, Updated and Expanded Edition). Toronto: Thompson Educational Publishing.

Manfredi, Christopher P. (1993a) Judicial Power and the Charter: Canada and the Paradox ofLiberal Constitutionalism. Ist Edition. Toronto: McClelland & Stewart.

Manfredi, Christopher P. (1993b) "Litigation and Institutional Design: Micro Constitutional Politics and the Canadian Charter of Rights and Freedoms." Presented at the annual meeting of the Canadian Political Science Association, Ottawa.

Manfredi, Christopher P. (1994a) '''Appropriate and Just in the Circumstances': Public Policy and the Enforcement of Rights Under the Canadian Charter of Rights and Freedoms." Canadian Journal ofPolitical Science 24: 289-307.

Manfredi, Christopher P. (1994b) "Constitutional Rights and Interest Advocacy: Litigating Educational Reform in Canada and the United States." Equity and Community: the Charter, Interest Advocacy and Representation. Ed. F. Leslie Seidle. Montreal: IRPP.

Manfredi, Christopher P. (1995) "Constitutional Modification, Institutional Design and the Politics of Gender Equality." Presented at the annual meeting of the Canadian Political Science Association, Montreal.

Manfredi, Christopher P. (1997) "Institutional Design and the Politics of Constitutional Modification: Understanding Amendment Failure in the United States and Canada." Law & Society Review 31(1): 111-136.

Manfredi, Christopher P. (2001) Judicial Power and the Charter: Canada and the Paradox ofLiberal Constitutionalism. 2nd Edition. Toronto: Oxford University Press.

Manfredi, Christopher P. and James B. Kelly (1999) "Six Degrees of Dialogue: A Response to Hogg and Bushell." Osgoode Hall Law Journal 37: 513-527.

Manfredi, Christopher P. and James B. Kelly (2001) "Dialogue, Deference and Restraint: Judicial Independence and Trial Procedures." Saskatchewan Law Review 64: 323- 346.

Manfredi, Christopher P. and Scott Lemieux (1999) "Judicial Discretion and Fundamental Justice: Sexual Assault in the Supreme Court of Canada." The American Journal of Comparative Law 47: 489-514.

204 Manfredi, Christopher P. and Michael Lusztig (1998) "Why Do FormaI Amendments Fail? An Institutional Design Analysis." World Politics 50: 377-400.

McCormick, Peter (1993) "Party Capability Theory and Appellate Success in the Supreme Court of Canada, 1949-1992." Canadian Journal ofPolitical Science 26: 523-40.

McCormick, Peter (1994) Canada's Courts: A Social Scientist's Ground-Breaking Account ofthe Canadian Judicial System. Toronto: James Lorimer & Company, Publishers.

McGuire, Kevin T. (1998) "Explaining Executive Success in the U.S. Supreme Court." Po/itical Research Quarterly 51: 505-26.

Menard, Scott (2002) Applied Logistic Regression Analysis. 2nd Edition. Quantitative Applications in the Social Sciences Series, Number 106. Thousand Oaks, CA: Sage Publications.

Mendes, Errol P.(1993) "Sinking Again into the Quagmire ofConflicting Visions, Groups, Underinclusion, and Death by Referendum." The Charlottetown Accord, the Referendum, and the Future ofCanada. Eds. Kenneth McRoberts and Patrick 1. Monahan. Toronto: University of Toronto Press.

Mitchell, Graeme G. (1993) "The Impact of the Charter on the Public Policy Process: The Attorney General." The Impact ofthe Charter on the Public Policy Process. Eds. Patrick Monahan and Marie Finkelstein. North Y ork: York University Centre for Public Law and Public Policy.

Monahan, Patrick (1987) Politics and the Constitution: The Charter, Federalism and the Supreme Court of Canada. Toronto: Carswell.

Monahan, Patrick (1991) Meech Lake: The Inside Story. Toronto: University of Toronto Press.

Monahan, Patrick and Marie Finkelstein, eds. (1993) The Impact ofthe Charter on the Public Policy Process. North York: York University Centre for Public Law and Public Policy.

Morton, F.L. (1987) "The Political Impact of the Canadian Charter of Rights and Freedoms." Canadian Journal ofPo/itical Science 20: 31-55.

Morton, F.L. (1993) "The Charter Revolution and the Court Party." The Impact ofthe Charter on the Public Policy Process. Eds. Patrick Monahan and Marie Finkelstein. North York: York University Centre for Public Law and Public Policy.

205 Morton, F.L. (1995) "The Effect of the Charter of Rights on Canadian Federalism." Publius: The Journal ofFederalism 25 (Summer): 173-88.

Morton, F.L. (1999) "Dialogue or Monologue?" Policy Options 20: 23-26.

Morton, F.L. and Avril Allen (2001) "Feminists and the Courts: Measuring Success in Interest Group Litigation in Canada." Canadian Journal ofPolitical Science 34: 55-84.

Morton, F.L., Matthew Hennigar and Shawn Ho (1996) "The Role of Governments in Charter Litigation: Charter Politics as Centre-Periphery Politics." Presented at the annual meeting of the Canadian Political Science Association, St. Catharines.

Morton, F.L. and Rainer Knopff (2000) The Charter Revolution and the Court Party. Toronto: Broadview Press.

Morton, F.L., Peter H. Russell and Troy Riddell (1994) "The Canadian Charter ofRights and Freedoms: A Descriptive Analysis of the First Decade, 1982-1992." National Journal ofConstitutional Law 5: 1-58.

Norman-Major, Kristen A. (1994) "The Solicitor General: Executive Policy Agendas and the Court." Albany Law Review 57: 1081-1109.

O'Connor, Karen (1983) "The Amicus Curiae Role ofthe U.S. Solicitor General in Supreme Court Litigation." Judicature 66: 256-264.

Pal, Leslie A. (1993) Interests ofState: The Politics ofLanguage, Multicufturalism, and Feminism in Canada. Montreal: McGill-Queen's University Press.

Pampe1, Fred C. (2000) Logistic Regression: A Primer. Quantitative Applications in the Social Sciences Series, Number 132. Thousand Oaks, CA: Sage Publications.

Perry, H.W., Jr. (1991) Deciding to Decide: Agenda Setting in the United States Supreme Court. Cambridge: Harvard University Press.

Perry, H.W., Jr. (1998) "United States Attomeys-Whom Shall They Serve?" Law and Contemporary Problems 61: 129-148.

Petter, Andrew (1986) "The Politics of the Charter." Supreme Court Law Review 8: 473- 505.

Rathjen, Gregory J. (1978) "Lawyers and Appellate Choice: An Analysis of Factors Affecting the Decision to Appeal." American Politics Quarterly 6: 387-405.

206 Riddell, Troy Q. and F.L. Morton (1998) "Reasonable Limitations, Distinct Society and the Canada Clause: Interpretive Clauses and the Competition for Constitutional Advantage." Canadian Journal ofPolitical Science 31: 467-493.

Roach, Kent (1994) "The Role of Litigation and the Charter in Interest Advocacy." Equity and Community: The Charter, Interest Advocacy and Representation. Ed. F. Leslie Seidle. Montreal: Institute for Research on Public Policy.

Roach, Kent (2000) "The Attorney General and the Charter Revisited." University of Toronto Law Review 50: 1-40.

Roach, Kent (2001) The Supreme Court on Trial: Judicial Activism or Democratic Dialogue. Toronto: Irwin Law.

Rosenberg, Gerald N. (1991) The Hollow Hope: Can Courts Bring About Social Change? Chicago: The University of Chicago Press.

Russell, Peter H. (1983) "The Political Purposes of the Canadian Charter of Rights and Freedoms." Canadian Bar Review 61: 30-54.

Russell, Peter H. (1991) "Standing Up For Notwithstanding." Alberta Law Review 29: 293-309.

Russell, Peter H. (1992) "The Anti-Inflation Case: The Anatomy of a Constitutional Decision." Law, Politics and the Judicial Process in Canada. 2nd Edition. Ed. F.L. Morton. Calgary: University of Calgary Press.

Russell, Peter H. (1993a) Constitutional Odyssey: Can Canadians Become a Sovereign People? 2nd Edition. Toronto: University of Toronto Press.

Russell, Peter H. (1993b) "The End of Mega-Constitutional Politics in Canada?" The Charlottetown Accord, the Referendum, and the Future of Canada. Eds. Kenneth McRoberts and Patrick 1. Monahan. Toronto: University of Toronto Press.

Russell, Peter H. (1994) "The Political Purposes of the Charter: Have They Been Fulfilled?" Protecting Rights and Freedoms: Essays on the Charter's Place in Canada's Political, Legal and Intellectual Life. Ed. Phillip Bryden et al. Toronto: University of Toronto Press.

Russell, Peter, Rainer Knopff and Ted Morton (1990) Federalism and the Charter: Leading Constitutional Decisions. Ottawa: Carleton University Press.

Salokar, Rebecca Mae (1992) The Solicitor General: The Politics ofLaw. Philadelphia: Temple University Press.

207 Savoie, Donald 1. (1999) Governingfrom the Centre: The Concentration ofPower in Canadian Polilics. Toronto: University of Toronto Press.

Schultz, Richard (1977) "Prime Ministerial Government, Central Agencies and Operating Departments: Towards a More Realistic Analysis." Apex ofPower. Ed. Thomas A. Hockin. Scarborough: Prentice-Hall.

Scott, Ian (1989) "Law, Policy and the Role of the Attorney General: Constancy and Change in the 1980s." University of Toronto Law Journal 39: 109-126.

Segal, Jeffrey A. (1988) "Amicus Curiae Briefs by the Solicitor General During the Warren and Burger Courts." Western Political Quarterly 41: 135-144.

Shilton, Elizabeth J. (1993) "Charter Litigation and the Policy Process of Government: A Public Interest Perspective." The Impact ofthe Charter on the Public Policy Process. Eds. Patrick Monahan and Marie Finkelstein. North York: York University Centre for Public Law and Public Policy.

Sigurdson, Richard. (1993) "Left- and Right-Wing Charterphobia in Canada: A Critique of the Critics." International Journal ofCanadian Studies 7-8: 96-115.

Simeon, Richard (1985) "Ontario in Confederation." The Government and Politics of Ontario. Ed. Donald C. MacDonald. Scarborough: Nelson Canada.

Smith, Miriam (1999) Lesbian and Gay Rights in Canada: Social Movements and Equality-Seeking, 1971-1995. Toronto: University of Toronto Press.

Smithey, Shannon (1996) "The Effects of the Canadian Supreme Court's Charter Interpretation on Regional and Intergovernmental Tensions in Canada." Publius: The Journal ofFederalism 26: 83-100.

Songer, Donald and Reginald Sheehan (1992) "Who Wins on Appeal? Upperdogs and Underdogs in the United States Courts of Appeals." American Journal ofPolitical Science 36: 235-58.

Spaeth, Harold (1994) United States Supreme Court Judicial Database: 1953-1992 Terms. Ann Arbor: ICPSR.

Strayer, Barry (1988) The Canadian Constitution and the Courts: The Function and Scope ofJudicial Review. 3rd Edition. Toronto: Butterworths.

Strauss, David A. (1998) "The Solicitor General and the Interests of the United States." Law and Contemporary Problems 61: 165-177.

208 Stuart, Don (1995) "Prosecutorial Accountability in Canada." Accountability for Criminal Justice: Selected Essays. Ed. Philip C. Stenning. Toronto: University of Toronto Press.

Swinton, Katherine E. (1990) The Supreme Court and Canadian Federalism: The Laskin Dickson Years. Toronto: Carswell.

Tait, John (1995) "Policy Development and the Charter." Perspectives on Public Policy: 1-24.

Tanenhaus, Joseph, Joseph Schick, Matthew Muraskin, and Daniel Rosen (1963) "The Supreme Court's Certiorari Jurisdiction: Cue Theory." Judicial Decision Making. Ed. Glendon Schubert. New York: The Free Press.

Taylor, Charles (1991) "Shared and Divergent Values." Optionsfor a New Canada. Eds. Ronald L. Watts and Douglas M. Brown. Toronto: University of Toronto Press.

Tsebelis, George (1990) Nested Games: Rational Choice in Comparative Politics. Berkeley: University of California Press.

Tushnet, Mark (1995) "Policy Distortion and Democratic Debilitation: Comparative Illumination of the Countermajoritarian Difficulty." Michigan Law Review 94: 245-301.

Ulmer, S. Sidney (1984) "The Supreme Court's Certiorari Decisions: Conflict as a Predictive Variable." American Political Science Review 78 (December): 901- 91l.

Vallinder, Torbjorn (1995) "When the Courts Go Marching In." The Global Expansion of Judicial Power. Eds. c. Neal Tate and Torbjorn Vallinder. New York: New York University Press.

Wald, Patricia M. (1998) '''For the United States': Govemment Lawyers in Court." Law and Contemporary Problems 61: 107-128.

Waltenburg, Eric N. and Bill Swinford (1999) "The Supreme Court as a Policy Arena: The Strategies and Tactics of State Attorneys General." Policy Studies Journal 27 : 242-259.

Webber, Jeremy (1994) Reimagining Canada: Language, Culture, Community and the Canadian Constitution. Montreal: McGill-Queen's University Press.

Weinrib, Lorraine Eisenstat (1999) "The Activist Constitution." Policy Options 20: 27- 30.

209 Wheeler, Stanton, Bliss Cartwright and Robert A. Kagan (1987) "Do the 'Haves' Come Out Ahead? Winning and Losing in State Supreme Courts, 1870-1970." Law and Society Review 21: 403-45.

Zorn, Christopher (1997) "U.S. Government Litigation Strategies in the Federal Appellate Courts." Ph.D. dissertation. Ohio State University.

Zorn, Christopher (2002) "U.S. Government Litigation Strategies in the Federal Appellate Courts." Political Research Quarterly 55(1): 145-166.

210 ~ McGill

McGilI University Research Ethics Board 1

Certificate of Ethical Acceptability of Research Involving Humans

'roject Title: Deciding to Appeal: Appeals by the Government of Canada to the Supreme Court of :anada, 1982-1999

Ipplicant's Name: Matthew HENNIGAR Department: Political Science

!cademic Status: Ph.D. student Supervisor: Dr. C. Manfredi

'his project was approved on 'J2&.- 1 8, ;). 0 00 by

::Xpedited Review: Full Review: / ------"------)epartmental review: (For research projects that are carried out by undergraduate and graduate students spart oftheir course work):

Department Chair) (Signature)

~he signatures below indicate that the project as described in this application is acceptable on thical grounds.

4. - (Y' Grantham (Econ~mics) P. R~ (Sociology)

5. ~ 6. F. Sabetti (~otd~af Scl~n~e) O. Toulan (Faculty of Management)

7. S. Lamb (Community Member)

'rofessor George Wenzel, Chair ,: n iversity Research Ethics Board 1 ~lIrnside Hall. room 728 '~I: 398-8882/4346 Fax: 398-7347