SUPREME COURT OF CANADA

2010 SCC 21, [2010] 1 S.C.R. 721

Date: 20100610

Docket: 33085, 32865

Between:

Toronto Star Newspapers Ltd., Canadian Broadcasting Corporation, Associated Press and CTV Television Inc.

Appellants/Respondents on cross‑appeal

• and -

Her Majesty The Queen in Right of Canada and A.A.

Respondents/Appellants on cross‑appeal

and

F.A., S.A., Qayyum Abdul Jamal, A.M.D., S.V.C. and Ahmad Mustafa Ghany

Respondents

and

Attorney General of Ontario, Attorney General of Alberta, N.S. (being a Young Person within the meaning of the Youth Criminal Justice Act), N.Y. (being a Young Person within the meaning of the Youth Criminal Justice Act), Canadian Civil Liberties Association, Canadian Newspaper Association, AD IDEM/Canadian Media Lawyers Association, RTNDA Canada/Association of Electronic Journalists and Canadian Association of Journalists

Interveners

And Between:

Canadian Broadcasting Corporation, Edmonton Journal, a Division of CanWest MediaWorks Publications Inc., CTV Television Inc. and Bell Globemedia Publishing Inc., carrying on business as The Globe and Mail

Appellants

and

Edmonton Sun, a Division of Sun Media Corporation

Appelant

• and -

Her Majesty The Queen and Michael James White

Respondents

and

Director of Public Prosecutions of Canada, Attorney General of Ontario and Canadian Civil Liberties Association

Interveners

Coram: McLachlin C.J. and Binnie, LeBel, Deschamps, Fish, Abella, Charron, Rothstein and Cromwell JJ.

Reasons for Judgment: (paras. 1 to 64) Deschamps J. (McLachlin C.J. and Binnie, LeBel, Fish, Charron, Rothstein and Cromwell JJ. concurring)

Dissenting Reasons: (paras. 65 to 77) Abella J.

Toronto Star Newspapers Ltd. v. Canada, 2010 SCC 21, [2010] 1 S.C.R. 721

Toronto Star Newspapers Ltd., Canadian Broadcasting Corporation, Associated Press and CTV Television Inc. Appellants/Respondents on cross‑appeal

v.

Her Majesty The Queen in Right of Canada and A.A. Respondents/Appellants on cross‑appeal

and

F.A., S.A., Qayyum Abdul Jamal, A.M.D., S.V.C. and Ahmad Mustafa Ghany Respondents

and

Attorney General of Ontario, Attorney General of Alberta, N.S. (being a Young Person within the meaning of the Youth Criminal Justice Act), N.Y. (being a Young Person within the meaning of the Youth Criminal Justice Act), Canadian Civil Liberties Association, Canadian Newspaper Association, Ad IDEM/Canadian Media Lawyers Association, RTNDA Canada/Association of Electronic Journalists and Canadian Association of Journalists Interveners

‑ and ‑

Canadian Broadcasting Corporation, Edmonton Journal, a Division of CanWest MediaWorks Publications Inc., CTV Television Inc. and Bell Globemedia Publishing Inc., carrying on business as The Globe and Mail Appellants

and

Edmonton Sun, a Division of Sun Media Corporation Appellant

v.

Her Majesty The Queen and Michael James White Respondents

and

Director of Public Prosecutions of Canada, Attorney General of Ontario and Canadian Civil Liberties Association Interveners

Indexed as: Toronto Star Newspapers Ltd. v. Canada

2010 SCC 21

File Nos.: 33085, 32865.

2009: November 16; 2010: June 10.

Present: McLachlin C.J. and Binnie, LeBel, Deschamps, Fish, Abella, Charron, Rothstein and Cromwell JJ.

on appeal from the courts of appeal for ontario and alberta

Constitutional law — Charter of Rights — Freedom of expression — Reasonable limits — Publication ban — Media organizations challenging constitutionality of statutory mandatory publication ban on bail hearing information — Whether mandatory ban justifiable infringement of freedom of expression — Canadian Charter of Rights and Freedoms, ss. 1, 2(b) — Criminal Code, R.S.C. 1985, c. C‑46, s. 517.

Under s. 517 of the Criminal Code, a justice of the peace is required, if an accused applies for one, to order a publication ban that applies to the evidence and information produced, to the representations made at a bail hearing and to any reasons given for the order. In the context of two high profile cases — a murder case in Alberta and an Ontario case involving terrorism‑related offences — a number of media organizations challenged the constitutionality of the mandatory aspect of the publication bans, contending the provision is an unjustifiable violation of freedom of expression guaranteed by the Canadian Charter of Rights and Freedoms. In the Alberta case, the media’s application was allowed, but the Court of Appeal set aside that decision and upheld the constitutional validity of s. 517. The court concluded that the mandatory ban, while it infringes freedom of expression, merely defers publication and that the values of protecting fair access to bail and the right to a fair trial were benefits which outweighed the deleterious effects of the restrictions on freedom of expression. In the Ontario case, the media’s application was dismissed. The Court of Appeal, in a majority decision, allowed the media’s appeal in part, finding that s. 517 was overbroad and read the provision down to exclude from the ban any cases in which the charges would not be tried by a jury. The dissenting judge would have declared the part of s. 517 relating to the mandatory ban to be invalid.

Held (Abella J. dissenting): The appeals should be dismissed and the cross-appeal in the Ontario case should be allowed. The constitutionality of s. 517 of the Criminal Code should be upheld.

Per McLachlin C.J. and Binnie, LeBel, Deschamps, Fish, Charron, Rothstein and Cromwell JJ.: Whether a discretion exists to issue a publication ban is not determinative of the validity of a limit on freedom of expression. The Dagenais/Mentuck test was not meant to apply to all limits on freedom of expression; rather, it was designed for and applies to discretionary orders. The validity of a statutory mandatory ban, such as the one at issue, must be determined by conducting an analysis based on the Oakes test. Bans are sometimes necessary, and whether they are justified depends on the context. The s. 517 mandatory publication ban is but one of numerous interrelated measures adopted as part of a sweeping reform of the rules on bail resulting from the 1969 Report on criminal justice and corrections. This Report recommended new rules to protect accused persons from the effects of pre‑trial incarceration and unsatisfactory conditions of detention, and to ensure that they were not punished at a time when they should be presumed innocent.

While the statutory mandatory publication ban limits freedom of expression, that limit can be justified in a free and democratic society. In adopting the various components of the bail reform and, more particularly, the mandatory ban, Parliament’s objectives were to ensure expeditious bail hearings and to safeguard the right to a fair trial. These objectives, which are undeniably pressing and substantial, were to be achieved by establishing a process that facilitated early release of an accused in order to mitigate the harshness of his or her interaction with the criminal justice system, limit the stigma as far as possible, and ensure that the trier of fact remains impartial.

When asking whether the mandatory publication ban is rationally connected to the objectives, the Court must consider other measures which might be linked to or even dependent on the ban. In this case, the mechanisms in place are closely linked and a rational connection can clearly be found in the interplay between the various components of the bail reform rules. They illustrate the expeditious nature of the bail hearing and the ultimate objective of safeguarding the right to a fair trial. The ban prevents dissemination of evidence which, for the sake of ensuring an expeditious hearing, is untested for relevance or admissibility.

The mandatory publication ban also meets the requirements of the minimal impairment stage of the Oakes test. If a publication ban hearing were to be held instead, an additional burden would be placed on the accused at a time when he or she may be overwhelmed by the criminal process, and may not have been able to consult his or her counsel of choice. Accused should be devoting their resources and energy to obtaining their release, not to deciding whether to compromise liberty in order to avoid having evidence aired outside the courtroom. In light of the delay and the resources which a publication ban hearing would entail, and of the prejudice which could result if untested evidence were made public, it would be difficult to imagine a measure capable of achieving Parliament’s objectives that would involve a more limited impairment of freedom of expression. Adding issues unrelated to the release of the accused to the bail hearing would require the consideration of matters extraneous to the bail process and could have a domino effect on other bail hearings in the same forum, thereby delaying the administration of justice. Moreover, the mandatory publication ban provided for in s. 517 is not an absolute ban either on access to the courts or on publication. The provision only prohibits the publication of evidence adduced, information given, representations made, and reasons given by the justice at a bail hearing. The media can publish the identity of the accused, comment on the facts and the offence with which the accused has been charged and for which the bail application has been made, and report on the outcome of the application. Journalists are also not prevented from informing the public of the legal conditions attached to the accused’s release. The temporary nature of the ban is another important factor. The ban ends when the accused is discharged after a preliminary inquiry or at the end of the trial. In essence, it applies only with respect to the bail process, and the information it covers can eventually be made public once more complete information produced in accordance with the standards applicable to criminal trials is available. Although information revealed at the bail hearing may no longer be newsworthy by the time the media can release it, the ban cannot be said to impair freedom of expression more than is necessary. The ban may make journalists’ work more difficult, but it does not prevent them from conveying and commenting on basic, relevant information.

Finally, the mandatory ban has several salutary effects. The ban limits the deprivation of the accused’s liberty by confining the issues at the hearing to those specifically related to bail, thereby avoiding undue delay and permitting accused persons to focus their energy and resources on their liberty interests rather than on their privacy interests. The ban also ensures that the public will not be influenced by untested, one‑sided and stigmatizing information bearing on issues that are often irrelevant to guilt. The deleterious effects of the publication ban, however, should not be downplayed. The ban prevents full public access to, and full scrutiny of, the criminal justice process. Moreover, the bail hearing may attract considerable media attention and its outcome may not be fully understood by the public. In such cases, the media would be better equipped to explain the judicial process to the public if the information they could convey were not restricted. Nonetheless, on balance, the deleterious effects of the limits on the publication of information are outweighed by the need to ensure certainty and timeliness, to conserve resources, and to avert the disclosure of untested prejudicial information — in other words, to guarantee as much as possible trial fairness and fair access to bail. While not a perfect situation, the mandatory ban represents a reasonable compromise.

Per Abella J. (dissenting): The mandatory ban in s. 517 of the Criminal Code is not a justified infringement of freedom of expression because it does not meet the proportionality requirement between the measure’s deleterious and salutary effects. The appropriate remedy is to sever the mandatory aspect of s. 517 and leave in place the discretion to order a publication ban.

Preventing disclosure of a judge’s reasons and of any information at a bail hearing until the trial is complete, a chronology which can take years to unfold, has the effect, for all but the handful of people who are present in the courtroom, of denying access to information surrounding a key aspect of the criminal justice system — the decision whether or not to release an accused back into the community pending his or her trial. This denial is a profound interference with the open court principle. The harm of that interference is not outweighed by the benefits of a mandatory ban — the reduction in pre‑trial publicity and delay. Each of these concerns can largely be attenuated, and neither is sufficiently significant to represent a serious infringement of fair trial rights. Remedies such as a partial ban, challenges for cause, or a change of venue if there is a sufficient risk of prejudice can address speculative concerns over pre‑trial publicity, and the ability of a properly instructed jury in a criminal trial to disregard irrelevant evidence should also be taken into account. Moreover, in the absence of automatic notice to the media, to which they are not entitled, there will be, in the overwhelming number of cases, no undue delay caused by a discretionary ban. Those few cases where the media is most likely to contest a ban are those which have a higher profile. However, the desirability of a universal mandatory ban should not be judged on its effectiveness for a small percentage of cases.

Public confidence in the justice system requires relevant information delivered in a timely way. A mandatory ban on the evidence heard and the reasons given in a bail application is a ban on the information when it is of most concern and interest to the public. Restrictions on the release of such information are only justified if their benefits outweigh their detrimental impact. Given that the salutary effects of the ban under s. 517 are not proportional to the harmful effects flowing from the infringement of the open court principle, the mandatory aspect of s. 517 should be struck out.

Cases Cited

By Deschamps J.

Referred to: Thomson Newspapers Co. v. Canada (Attorney General), 1998 CanLII 829 (SCC), [1998] 1 S.C.R. 877; R. v. Oakes, 1986 CanLII 46 (SCC), [1986] 1 S.C.R. 103; Re Global Communications Ltd. and Attorney-General for Canada (1984), 1984 CanLII 2153 (ON CA), 44 O.R. (2d) 609; Canadian Broadcasting Corp. v. New Brunswick (Attorney General), 1996 CanLII 184 (SCC), [1996] 3 S.C.R. 480; Dagenais v. Canadian Broadcasting Corp., 1994 CanLII 39 (SCC), [1994] 3 S.C.R. 835; R. v. Mentuck, 2001 SCC 76, [2001] 3 S.C.R. 442; Attorney General of Nova Scotia v. MacIntyre, 1982 CanLII 14 (SCC), [1982] 1 S.C.R. 175; Named Person v. Vancouver Sun, 2007 SCC 43, [2007] 3 S.C.R. 253; Vancouver Sun (Re), 2004 SCC 43, [2004] 2 S.C.R. 332; RJR‑MacDonald Inc. v. Canada (Attorney General), 1995 CanLII 64 (SCC), [1995] 3 S.C.R. 199; Alberta v. Hutterian Brethren of Wilson Colony, 2009 SCC 37, [2009] 2 S.C.R. 567; R. v. Burlingham, 1995 CanLII 88 (SCC), [1995] 2 S.C.R. 206; R. v. Stillman, 1997 CanLII 384 (SCC), [1997] 1 S.C.R. 607; R. v. Big M Drug Mart Ltd., 1985 CanLII 69 (SCC), [1985] 1 S.C.R. 295; R. v. Butler, 1992 CanLII 124 (SCC), [1992] 1 S.C.R. 452; R. v. Keegstra, 1990 CanLII 24 (SCC), [1990] 3 S.C.R. 697; Canadian Newspapers Co. v. Canada (Attorney General), 1988 CanLII 52 (SCC), [1988] 2 S.C.R. 122; R. v. Hall, 2002 SCC 64, [2002] 3 S.C.R. 309; R. v. Jevons, 2008 ONCJ 559, [2008] O.J. No. 4397 (QL); Edmonton Journal v. Alberta (Attorney General), 1989 CanLII 20 (SCC), [1989] 2 S.C.R. 1326.

By Abella J. (dissenting)

Edmonton Journal v. Alberta (Attorney General), 1989 CanLII 20 (SCC), [1989] 2 S.C.R. 1326; Attorney General of Nova Scotia v. MacIntyre, 1982 CanLII 14 (SCC), [1982] 1 S.C.R. 175; Vickery v. Nova Scotia Supreme Court (Prothonotary), 1991 CanLII 90 (SCC), [1991] 1 S.C.R. 671; Canadian Broadcasting Corp. v. New Brunswick (Attorney General), 1996 CanLII 184 (SCC), [1996] 3 S.C.R. 480; R. v. Mentuck, 2001 SCC 76, [2001] 3 S.C.R. 442; Sierra Club of Canada v. Canada (Minister of Finance), 2002 SCC 41, [2002] 2 S.C.R. 522; Ruby v. Canada (Solicitor General), 2002 SCC 75, [2002] 4 S.C.R. 3; Vancouver Sun (Re), 2004 SCC 43, [2004] 2 S.C.R. 332; Toronto Star Newspapers Ltd. v. Ontario, 2005 SCC 41, [2005] 2 S.C.R. 188; Named Person v. Vancouver Sun, 2007 SCC 43, [2007] 3 S.C.R. 253; Dagenais v. Canadian Broadcasting Corp., 1994 CanLII 39 (SCC), [1994] 3 S.C.R. 835; R. v. Hall, 2002 SCC 64, [2002] 3 S.C.R. 309; R. v. Corbett, 1988 CanLII 80 (SCC), [1988] 1 S.C.R. 670; R. v. Vermette, 1988 CanLII 87 (SCC), [1988] 1 S.C.R. 985; R. v. White, 2005 ABCA 435, 56 Alta. L.R. (4th) 255.

Statutes and Regulations Cited

Bail Reform Act, S.C. 1970‑71‑72, c. 37.

Canadian Charter of Rights and Freedoms, ss. 1, 2(b), 11(e).

Criminal Code, R.S.C. 1985, c. C‑46, ss. 503(1)(a), 515, 516(1), 517, 518, 520, 539.

Authors Cited

Barak, Aharon. “Proportional Effect: The Israeli Experience” (2007), 57 U.T.L.J. 369.

Canada. Committee on Corrections. Report of the Canadian Committee on Corrections — Toward Unity: Criminal Justice and Corrections. Ottawa: Queen’s Printer, 1969.

Canada. House of Commons. House of Commons Debates, vol. III, 3rd Sess., 28th Parl., February 5, 1971, pp. 3113‑14.

Canada. House of Commons. Standing Committee on Justice and Legal Affairs. Minutes of Proceedings and Evidence, No. 11, 1st Sess., 28th Parl., March 18, 1969, pp. 501‑2.

Friedland, Martin L. Detention before Trial: A Study of Criminal Cases Tried in the Toronto Magistrates’ Courts. Toronto: University of Toronto Press, 1965.

Mirfield, Peter. “The Early Jurisprudence of Judicial Disrepute” (1987‑88), 30 Crim. L.Q. 434.

Ontario. Royal Commission. Inquiry into Civil Rights. Toronto: Queen’s Printer, 1968.

Pink, Joel E., and David C. Perrier, eds. From Crime to Punishment: An Introduction to the Criminal Law System, 6th ed. Toronto: Thomson/Carswell, 2007.

Trotter, Gary T. The Law of Bail in Canada, 2nd ed. Scarborough, Ont.: Thomson/Carswell, 1999.

APPEAL and CROSS‑APPEAL from a judgment of the Ontario Court of Appeal (Laskin, Rosenberg, Feldman, Simmons and Juriansz JJ.A.), 2009 ONCA 59, 94 O.R. (3d) 82, 239 C.C.C. (3d) 437, 302 D.L.R. (4th) 385, 245 O.A.C. 291, [2009] O.J. No. 288 (QL), 2009 CarswellOnt 301, setting aside in part a decision of Durno J. (2007), 84 O.R. (3d) 766, 2007 CarswellOnt 1224, 2007 CanLII 6249, upholding the constitutionality of s. 517 of the Criminal Code. Appeal dismissed and cross-appeal allowed, Abella J. dissenting.

APPEAL from a judgment of the Alberta Court of Appeal (Conrad, Ritter and Slatter JJ.A.), 2008 ABCA 294, 93 Alta. L.R. (4th) 239, [2008] 10 W.W.R. 588, 437 A.R. 130, 433 W.A.C. 130, 236 C.C.C. (3d) 204, 298 D.L.R. (4th) 659, 179 C.R.R. (2d) 227, [2008] A.J. No. 956 (QL), 2008 CarswellAlta 1158, setting aside a decision of Brooker J., 2007 ABQB 359, 77 Alta. L.R. (4th) 98, 221 C.C.C. (3d) 393, [2007] 10 W.W.R. 250, 48 C.R. (6th) 300, 158 C.R.R. (2d) 270, 420 A.R. 1, [2007] A.J. No. 608 (QL), 2007 CarswellAlta 774, declaring s. 517 of the Criminal Code unconstitutional. Appeal dismissed, Abella J. dissenting.

Paul B. Schabas and Ryder Gilliland, for the appellants/respondents on cross‑appeal Toronto Star Newspapers Ltd. et al.

Frederick S. Kozak, Q.C., and Matthew A. Woodley, for the appellants Canadian Broadcasting Corporation et al.

Barry Zalmanowitz, Q.C., and Peter D. Banks, for the appellant Edmonton Sun, a Division of Sun Media Corporation.

John North and Steve Coroza, for the respondent/appellant on cross‑appeal Her Majesty The Queen in Right of Canada and for the intervener the Director of Public Prosecutions of Canada.

John Norris and Breese Davies, for the respondent/appellant on cross‑appeal A.A.

Dennis Edney and Raymond Motee, for the respondent F.A.

Peter G. Martin, for the respondent S.A.

Anser Farooq, for the respondent Qayyum Abdul Jamal.

Rocco Galati, for the respondents A.M.D. and Ahmad Mustafa Ghany.

Delmar Doucette and Michael Moon, for the respondent S.V.C.

Jolaine Antonio, for the respondent Her Majesty The Queen and for the intervener the Attorney General of Alberta.

Lauren Garcia and Kirk Starkie, for the respondent Michael James White.

M. David Lepofsky, Peter Scrutton and Daniel Guttman, for the intervener the Attorney General of Ontario.

Christopher Hicks and Catriona Verner, for the interveners N.S. and N.Y.

Jonathan C. Lisus and Alexi N. Wood, for the intervener the Canadian Civil Liberties Association.

Daniel W. Burnett, for the interveners the Canadian Newspaper Association et al.

The judgment of McLachlin C.J. and Binnie, LeBel, Deschamps, Fish, Charron, Rothstein and Cromwell JJ. was delivered by

[1] Deschamps J. — Upholding the rights of Canadian citizens by fostering trial fairness and safeguarding liberty interests is central to the criminal justice process. At the same time, access to the courts is central to a democratic society; it is a means to protect against arbitrary state action. Access to the courts is grounded in freedom of expression. Trial fairness and liberty interests must not clash with freedom of expression. They can be reconciled.

[2] A number of media organizations are urging this Court to find that s. 517 of the Criminal Code, R.S.C. 1985, c. C-46 (“Cr. C.”), unjustifiably violates the freedom of expression guaranteed by the Canadian Charter of Rights and Freedoms. Under this provision, a justice of the peace or provincial court judge (a “justice”) is required, if an accused applies for one, to order a publication ban that applies to the evidence and information produced, and representations made, at a bail hearing and to any reasons given for the order. There is no question that this order limits freedom of expression. Section 517 states:

517. (1) If the prosecutor or the accused intends to show cause under section 515, he or she shall so state to the justice and the justice may, and shall on application by the accused, before or at any time during the course of the proceedings under that section, make an order directing that the evidence taken, the information given or the representations made and the reasons, if any, given or to be given by the justice shall not be published in any document, or broadcast or transmitted in any way before such time as

(a) if a preliminary inquiry is held, the accused in respect of whom the proceedings are held is discharged; or

(b) if the accused in respect of whom the proceedings are held is tried or ordered to stand trial, the trial is ended.

The issue is whether that limit can be justified in a free and democratic society. For the reasons that follow, I conclude that the statutory mandatory publication ban is justified. These reasons apply to the two cases at bar, which were heard concurrently. I would dismiss the appeal and allow the cross-appeal in the Ontario case and dismiss the appeal in the Alberta case.

[3] Context is the key to understanding the scope and impact of a limit on a Charter right. As this Court said in Thomson Newspapers Co. v. Canada (Attorney General), 1998 CanLII 829 (SCC), [1998] 1 S.C.R. 877, “[t]he analysis under s. 1 of the Charter must be undertaken with a close attention to context. This is inevitable as the test devised in R. v. Oakes, 1986 CanLII 46 (SCC), [1986] 1 S.C.R. 103, requires a court to establish the objective of the impugned provision, which can only be accomplished by canvassing the nature of the social problem which it addresses” (para. 87). To properly assess the challenge raised by the appeals, it will be necessary to consider the historical and legislative context of the enactment of the interim release provisions found in the Criminal Code. I will then review the limits on freedom of expression in the criminal law context before discussing the issue of justification under the Oakes framework. The appellants argued these appeals on the basis of two different sets of facts that, in their view, demonstrate the need to promote openness in the context of the interim release provisions. I will thus begin by briefly setting out the facts in these two cases.

I. Facts and Judicial History

A. The Alberta Case

[4] Michael White was charged with the murder of his wife in Alberta. He was granted bail by Brooker J. of the Alberta Court of Queen’s Bench on October 7, 2005, and a publication ban was ordered pursuant to s. 517 Cr. C. According to the appellants, Mr. White’s release provoked public outrage. The constitutionality of the ban was then challenged successfully in the Court of Queen’s Bench (2007 ABQB 359, 77 Alta. L.R. (4th) 98). Brooker J. found that the legislative objective underlying the ban was to protect the right of the accused to a fair trial before an impartial jury, and that reason and logic alone were insufficient to establish a rational connection between the ban and this objective. He went on to find that the means also failed to meet the minimal impairment test.

[5] The Alberta Court of Appeal reversed Brooker J.’s decision (2008 ABCA 294, 93 Alta. L.R. (4th) 239). Slatter J.A., writing for a unanimous court, concluded that a s. 517 ban merely defers publication and that the values of protecting fair access to bail and the right to a fair trial were benefits that outweighed the deleterious effects of the restrictions.

B. The Ontario Case

[6] On June 2, 2006, twelve adults and five young persons were charged with various terrorism-related offences under the Criminal Code. The arrests attracted massive media attention between June 3 and June 12, 2006. One of the accused applied for a publication ban, while some of the others opposed it.

[7] On June 12, 2006, Justice of the Peace Currie ordered a ban. The appellants moved to quash the order. Durno J. of the Ontario Superior Court of Justice dismissed their application, holding that if one accused seeks a ban under s. 517, the order applies to all his or her co-accused ((2006), 2006 CanLII 25418 (ON SC), 211 C.C.C. (3d) 234, at para. 101). Some of the accused were released pending their trial, while others remained in custody. The appellants and two of the accused challenged the constitutionality of s. 517. Durno J., finding that he was bound by the decision in Re Global Communications Ltd. and Attorney-General for Canada (1984), 1984 CanLII 2153 (ON CA), 44 O.R. (2d) 609 (C.A.), held that s. 517 does not infringe the Charter ((2007) 2007 CanLII 6249 (ON SC), 84 O.R. (3d) 766, at para. 48).

[8] On appeal, Feldman J.A., writing for the majority of the Ontario Court of Appeal (Laskin and Simmons JJ.A. concurring), found that s. 517 was overbroad (2009 ONCA 59, 94 O.R. (3d) 82, at para. 159). She held that the objective of the provision — to safeguard the right to a fair trial by averting jury bias by means of a ban on the publication of prejudicial information — was pressing and substantial and that the ban was rationally connected to the objective. However, she concluded that, as drafted, the provision did not meet the minimal impairment test, because it applied to bail hearings in respect of all charges regardless of the mode of trial. As a remedy, Feldman J.A. read s. 517 down to exclude from the ban any cases in which the charges would not be tried by a jury. Rosenberg J.A., dissenting (Juriansz J.A. concurring), would have declared the part of s. 517 relating to the mandatory ban to be invalid on the basis that it did not meet the requirement of proportionality between the deleterious and the salutary effects of the measure. The appellants ask this Court to adopt the view of the dissent, while the respondents cross-appeal, arguing that the provision is valid.

II. Historical and Legislative Context

[9] Bail was developed in early English law not as a means to further the liberty interest of the accused but as a response to the deplorable conditions of jails: inmates in pre-trial detention died awaiting their trials. To avoid this, accused persons who were not at risk of failing to appear at trial were released on bail. Factors such as the seriousness of the offence, the likelihood that the accused was guilty and the status of the accused were seen as indicators of the likelihood that the accused would appear at trial. In Canada, too, the likelihood of attendance at trial was initially the predominant consideration. However, additional factors subsequently came to be developed, such as the need to protect the public from repeat offenders. The process for granting bail remained quite informal and discretionary until the middle of the 20th century (G. T. Trotter, The Law of Bail in Canada (2nd ed. 1999), at pp. 3-9).

[10] Disturbing data then emerged from an empirical study by Prof. Martin Friedland that was published in 1965 (Detention before Trial: A Study of Criminal Cases Tried in the Toronto Magistrates’ Courts). Friedland observed that the existing procedures resulted in the detention of many individuals whose attendance could have been secured by less restrictive means. In addition, he noted a relationship between pre‑trial detention, conviction and custodial sentences. This study prompted a re-examination of the rules on bail by the Ontario Royal Commission (Inquiry into Civil Rights (1968)) and by the Canadian Committee on Corrections (Toward Unity: Criminal Justice and Corrections (1969) (“Ouimet Report”)).

[11] The authors of the Ouimet Report found that considerations other than ensuring the attendance of the accused at trial were relevant to the decision whether to order interim release. They stated, as a guiding proposition, that “[t]he basic purposes of the criminal law should be carried out with no more interference with the freedom of individuals than is necessary” (p. 11). The Report’s authors noted that the initial period after an arrest is determinative. Pre-trial incarceration could lead to loss of employment and make it impossible for accused persons to fulfill their family obligations, thereby weakening their family and social ties. The authors observed that the conditions of detention were unsatisfactory. In addition, although they were careful in weighing the statistical evidence, they relied on it to conclude that pre-trial custody had a negative impact on the chances of acquittal. In their opinion, it was almost obvious that how an accused was treated between arraignment and trial had an impact on the corrective measures that would be called for if the accused were convicted. They accepted that incarceration could result in a permanent stigma, even if the accused were eventually acquitted. Moreover, they were of the view that the release of accused persons until trial was intended to ensure that individuals were not being punished at a time when they should be presumed innocent. From a human rights perspective, the authors considered it clear that an accused should not be incarcerated while awaiting trial unless the protection of society made it necessary to do so.

[12] In light of their findings, the authors of the Ouimet Report made a number of recommendations, including that police officers be allowed to release accused persons pending their appearance before a justice, that the onus be placed on the prosecution to justify detention, that flexible rules be adopted to ensure early bail hearings, that provision be made for a publication ban at the request of the accused and that, in a case in which an accused was not represented by counsel, the justice be required to inform the accused of the right to a ban. Finally, one of the most important features of the proposed reform was that limited grounds for refusing interim release were identified. Under reformed rules, the authors said, release should be the rule and custody the exception.

[13] The Ouimet Report stressed the right to a fair trial and the need to implement new rules to facilitate the early release of the accused whenever that was appropriate. As can be seen from the subsequent amendments to the Criminal Code (Bail Reform Act, S.C. 1970-71-72, c. 37), which were enacted almost immediately after the release of the Ouimet Report, the government embraced most of the Report’s recommendations. In brief, the amendments required that an arrested person be brought before a justice without unreasonable delay and that the person be released unless the Crown showed cause for ongoing detention. Only limited grounds were available for denying release, and the Crown was responsible for gathering any evidence it intended to produce at the bail hearing. Also, a mandatory publication ban would have to be ordered if requested by the accused. The new legislation was promoted as protecting individual rights. John Turner, the then Minister of Justice, declared in the House of Commons:

I said that as soon as we could, I intended to turn once again along the road of law reform and continuing enhancement and protection of civil liberties. . . . This bill is directed at making that first contact between citizens and the criminal judicial process less abrasive.

(House of Commons Debates, vol. III, 3rd Sess., 28th Parl., February 5, 1971, at pp. 3113-14)

[14] The reform of the rules governing interim release was implemented long before the Charter but foreshadowed the period of consciousness of individual rights. The purpose of the new rules was to avoid alienating accused persons without exposing society to undue risks. Today, the constitutional right to bail is not just an element of the protection against arbitrary detention, but is also explicitly recognized in s. 11(e) of the Charter. The protection is both procedural and substantive in that not only is reasonable bail guaranteed, but the cause for the denial of bail must be just. Hence, as I will discuss below, the bail process is inextricably linked to the right to bail itself.

III. Publication Bans in the Criminal Law Context

[15] As La Forest J. stated in Canadian Broadcasting Corp. v. New Brunswick (Attorney General), 1996 CanLII 184 (SCC), [1996] 3 S.C.R. 480, at para. 23, “[o]penness permits public access to information about the courts, which in turn permits the public to discuss and put forward opinions and criticisms of court practices and proceedings.” A publication ban therefore infringes freedom of expression. However, it should not be concluded that there is a conflict between freedom of expression and the rights of the accused. The “clash model” was rejected in Dagenais v. Canadian Broadcasting Corp., 1994 CanLII 39 (SCC), [1994] 3 S.C.R. 835, at pp. 882 et seq., in which Lamer C.J. stated that, rather than giving one right priority over the other, the Court must engage in a balancing exercise that takes into account the salutary and deleterious effects of the measure and any alternatives. In Dagenais, because the objection of the accused related to trial fairness in the context of adverse pre-trial publicity (p. 879), the Court focussed on that concern. However, it soon became clear that the analytical framework could not be limited to that context. In New Brunswick (at para. 69), La Forest J., writing for a unanimous Court, recognized that the interest of the proper administration of justice could justify banning the press from the courtroom for a limited time. The Court found that the discretion conferred on the judge to assess the specific circumstances of each individual case was crucial to the analysis of the validity of the challenged provision. Then, in R. v. Mentuck, 2001 SCC 76, [2001] 3 S.C.R. 442, the Court applied the balancing test to another aspect of the proper administration of justice. Iacobucci J. stressed that while Dagenais was the starting point of the analysis, a publication ban could involve a broad range of objectives:

However, the common law rule under which the trial judge considered the publication ban in this case is broader than its specific application in Dagenais. The rule can accommodate orders that must occasionally be made in the interests of the administration of justice, which encompass more than fair trial rights. [para. 31]

[16] The bans in Dagenais and Mentuck were based on the courts’ common law jurisdiction to order publication bans. The authority for the exclusion order in New Brunswick was statutory. In all three cases, the orders were discretionary. In the cases at bar, the appellants argue that the order provided for in s. 517 Cr. C. fails the New Brunswick test because it is mandatory: the judge does not have a discretion to consider the justification for the ban in light of the circumstances of the case. The appellants add that there is no rational connection between the ban and the objective of the legislation and that the ban fails to meet the requirements of the minimal impairment and proportionality stages of the Oakes test. As can be seen from the above cases, bans are sometimes necessary, and whether they are justified depends on the context.

IV. Discretion as a Constitutional Threshold

[17] The appellants accept that a limited publication ban may be valid in some narrow circumstances. However, they contend that the existence of a judicial discretion amounts to a constitutional threshold. In my view, their position does not reflect this Court’s approach to limits on freedom of expression.

[18] Whether a discretion exists is not determinative of the validity of a limit on freedom of expression. For example, the limit on access to the content of a search warrant prior to the execution of the warrant is not discretionary (Attorney General of Nova Scotia v. MacIntyre, 1982 CanLII 14 (SCC), [1982] 1 S.C.R. 175). To consider mandatory bans unconstitutional because the circumstances in which they apply cannot be scrutinized in a Dagenais analysis would be to turn the rule on its head. In Dagenais, Lamer C.J. explicitly stated that his analysis did not concern bans required by statute (pp. 856-57). Moreover, as Bastarache J. stated in Named Person v. Vancouver Sun, 2007 SCC 43, [2007] 3 S.C.R. 253, at para. 36, the Dagenais test was not meant to apply to all limits on freedom of expression; rather, it was designed for and applies to discretionary orders (see also Vancouver Sun (Re), 2004 SCC 43, [2004] 2 S.C.R. 332, at para. 31). Discretionary bans are constitutional because the test developed in Dagenais/Mentuck incorporates the essence of the balancing exercise mandated by the Oakes test. Indeed, as Lamer C.J. said in Dagenais, “[i]f legislation requires a judge to order a publication ban, then any objection to that ban should be framed as a Charter challenge to the legislation itself” (p. 874 (emphasis in original)). The validity of a statutory mandatory ban, such as the one at issue and the one provided for in s. 539 Cr. C. with respect to evidence led at a preliminary inquiry, will be determined by conducting an analysis based on the Oakes test.

V. The Oakes Test

[19] The various stages of the Oakes test are well known. When a protected right is infringed, the government must justify its action by identifying a pressing and substantial objective, by demonstrating that there is a rational connection between the objective and the infringement, and by showing that the means chosen interferes as little as possible with the right and that the benefits of the measure taken outweigh its deleterious effects.

A. Pressing and Substantial Objective

[20] The identification of Parliament’s objectives in adopting the mandatory ban is of great importance, as this will have a considerable impact on the analysis of the remaining stages of the test. As McLachlin J. (as she then was) noted in RJR-MacDonald Inc. v. Canada (Attorney General), 1995 CanLII 64 (SCC), [1995] 3 S.C.R. 199, at para. 144, “[c]are must be taken not to overstate the objective. The objective relevant to the s. 1 analysis is the objective of the infringing measure, since it is the infringing measure and nothing else which is sought to be justified. If the objective is stated too broadly, its importance may be exaggerated and the analysis compromised” (emphasis in original). Moreover, in Alberta v. Hutterian Brethren of Wilson Colony, 2009 SCC 37, [2009] 2 S.C.R. 567, at para. 76, McLachlin C.J. endorsed the following comment by President Barak: “Whereas the rational connection test and the least harmful measure test are essentially determined against the background of the proper objective, and are derived from the need to realize it, the test of proportionality (stricto sensu) examines whether the realization of this proper objective is commensurate with the deleterious effect upon the human right” (A. Barak, “Proportional Effect: The Israeli Experience” (2007), 57 U.T.L.J. 369, at p. 374). This suggests that all steps of the Oakes test are premised on a proper identification of the objective of the impugned measure.

[21] The mandatory publication ban is but one of numerous interrelated measures adopted as part of a sweeping reform of the rules on bail. Parliament’s objective in adopting the ban has to be identified by examining the provision in question in the context of the entire reform package. To assess the validity of the limit, it is necessary to understand the part the provision plays within the broader scheme of the reform.

[22] In the Ontario Court of Appeal, both the majority and the dissent expressed the view that the objective of the provision is to foster trial fairness. Trial fairness is a concept that can be interpreted in different ways. Although in Dagenais it was limited to averting jury bias by banning pre-trial publicity, this narrow view is not the only one recognized in the caselaw. Trial fairness has also been understood as encompassing all measures whose purpose it is to protect the fundamental rights of the accused (see R. v. Burlingham, 1995 CanLII 88 (SCC), [1995] 2 S.C.R. 206, at para. 29; R. v. Stillman, 1997 CanLII 384 (SCC), [1997] 1 S.C.R. 607; and P. Mirfield, “The Early Jurisprudence of Judicial Disrepute” (1987‑88), 30 Crim. L.Q. 434, at pp. 444 and 452). To define the interests at issue, therefore, the context must be taken into account. Rosenberg J.A., who wrote the dissenting reasons in the Ontario case, embraced a view of trial fairness which accurately identifies the objectives Parliament appears to have been pursuing in adopting the various components of the bail reform and, more particularly, in enacting the provision establishing the right to a mandatory ban:

The interest in a fair trial embraces not simply the narrow interest of preventing potential jurors from being influenced by prejudicial material that might be disclosed at a bail hearing, but other interests intended to safeguard the accused’s and society’s interest in a fair trial. Those interests include preventing diversion of the accused’s scarce resources to fight opposition to a publication ban and preventing delay of the bail hearing. As regards the latter, keeping accused in custody interferes with their ability to defend the case. The objectives of ensuring expeditious bail hearings, avoiding unnecessary detention of accused and allowing accused to retain scarce resources to defend their cases are all inextricably linked to the objective of ensuring a fair trial. [para. 38]

[23] As I mentioned above, the bail reform was implemented following the Ouimet Report, according to which there appeared to be a correlation between pre-trial custody, conviction, and custodial sentences. The Report’s authors stressed the fact that a first offender’s initial encounter with the justice system was crucial and could have dire consequences for the accused and his or her family. Measures were required to protect an accused person, who could bear a stigma even after being acquitted. In this context, Parliament’s primary objective can be defined on the basis of an understanding of trial fairness that is not limited to averting jury bias. Given the particular emphasis placed in the Ouimet Report on ensuring expeditious bail hearings, I would define Parliament’s objectives as (1) to safeguard the right to a fair trial; and (2) to ensure expeditious bail hearings. These objectives were to be achieved by establishing a process that facilitated early release of an accused in order to mitigate the harshness of his or her interaction with the criminal justice system, limit the stigma as far as possible, and ensure that the trier of fact remains impartial. In view of the facts noted in the Ouimet Report, these objectives were undeniably pressing. Since they are solidly grounded in Charter values, they are also substantial.

[24] The appellants argue that to accept ensuring expeditious bail hearings as an objective would be to apply the “shifting purpose” doctrine this Court explicitly rejected in R. v. Big M Drug Mart Ltd., 1985 CanLII 69 (SCC), [1985] 1 S.C.R. 295. In R. v. Butler, 1992 CanLII 124 (SCC), [1992] 1 S.C.R. 452, at p. 494, the Court held that the concepts of moral corruption and harm to society are not distinct, but are inextricably linked. In the same way, the objectives of ensuring expeditious bail hearings and fostering trial fairness are inextricably linked, as the latter embraces the former. As a result, this formulation of the objectives not only does not attract the prohibition against the shifting purpose doctrine, it is rooted in the Ouimet Report’s analysis.

B. Rational Connection

[25] At the next stage of the Oakes test, it must be asked whether there is a rational connection between the adopted means and Parliament’s objectives. “The government must show that it is reasonable to suppose that the limit may further the goal, not that it will do so” (Hutterian Brethren, at para. 48). Absent conclusive scientific or empirical evidence of a rational connection, one can be found by applying reason and logic: RJR‑MacDonald, at para. 158; Butler, at p. 503; R. v. Keegstra, 1990 CanLII 24 (SCC), [1990] 3 S.C.R. 697, at pp. 768 and 776; Thomson Newspapers, at paras. 104-7.

[26] In the cases at bar, since the mandatory publication ban is only one part of a whole, the enquiry cannot be limited to the ban itself. Therefore, when asking whether the mandatory publication ban is rationally connected to the objectives of protecting the right of the accused to a fair trial and ensuring expeditious bail hearings, the Court must consider other measures which might be linked to, or even dependent on, the ban.

[27] The mechanisms in place are closely linked. They illustrate the expeditious nature of the bail hearing and the ultimate objective of safeguarding the right to a fair trial. For instance, s. 503(1)(a) Cr. C. requires that a person who is arrested and detained be taken before a justice “without unreasonable delay” and in any event within 24 hours after the arrest. Section 515 Cr. C. provides that the justice shall release the person unless the prosecution shows cause why the detention should be continued. The grounds that can be relied on to deny the person’s release are limited. In the short time it has before it must show cause why the detention of the accused is justified, the prosecution has to gather the evidence it intends to use at the bail hearing, which means it may have insufficient time to meet with witnesses and further investigate the matters relevant to bail. Section 516(1) Cr. C. provides that the adjournment of a bail hearing cannot exceed three days except with the consent of the accused; and orders can be reviewed at the request of the accused provided that the accused has given the prosecution two days’ notice (s. 520(1) and (2) Cr. C.).

[28] To avoid any delay prejudicial to an accused who ought to be released, while at the same time ensuring that those who do not meet the criteria for release are kept in custody, compromises had to be made regarding the nature of the evidence to be adduced at the bail hearing. There are practically no prohibitions as regards the evidence the prosecution can lead to show cause why the detention of the accused in custody is justified. According to s. 518(1)(e) Cr. C., the prosecutor may lead any evidence that is “credible or trustworthy”, which might include evidence of a confession that has not been tested for voluntariness or consistency with the Charter, bad character, information obtained by wiretap, hearsay statements, ambiguous post-offence conduct, untested similar facts, prior convictions, untried charges, or personal information on living and social habits. The justice has a broad discretion to “make such inquiries, on oath or otherwise, of and concerning the accused as he considers desirable” (s. 518(1)(a)). The process is informal; the bail hearing can even take place over the phone (s. 515(2.2)).

[29] The appellants consider it significant that the mandatory publication ban was not part of the initial reform package and was not commented on during the debate in Parliament when the legislation was introduced a few years later. Two points have to be made with respect to this argument. The first is found in the Ouimet Report itself, and the second relates to the ban on publishing evidence taken at preliminary inquiries.

[30] As we have seen, Parliament chose to follow most of the recommendations of the Ouimet Report. It is therefore safe to say that Parliament found sufficient justification for reform in the comments of the Report’s authors. They had recommended that, if the accused sought a ban, it be mandatory. In addition, they had recommended that the justice advise an unrepresented accused of the right to a ban. The explanation they had given for these recommendations was that as soon as the accused was brought before the justice, the prosecutor needed to be able to provide arguments sufficiently convincing to warrant detention. It was understood that information so provided would not necessarily be relevant or admissible at trial. The mandatory ban was seen as a measure “taken to prevent prejudicing the accused at his trial by the dissemination of prejudicial matter which would not be relevant or admissible at his trial” (Ouimet Report, at p. 110). It can be inferred from this that the interests to be protected included not only the avoidance of stigmatization of the accused, but also trial fairness. In view of the strength of the recommendation and the fact that no explanation was given for omitting the ban from the initial bill, it is safe to assume that the purpose of including it was to rectify this omission. Therefore, in my view, no argument can be drawn from the fact that the ban was not part of the initial version of the new bail provisions of the Criminal Code.

[31] Furthermore, mandatory publication bans were not unknown at the time of the Ouimet Report. For instance, s. 539 Cr. C. requires that a ban on the publication of evidence adduced at a preliminary inquiry be ordered should the accused apply for one. That ban was discussed by the Standing Committee on Justice and Legal Affairs. Minister Turner justified it as follows:

We are not talking about a situation of legitimate publicity at an open trial once the jury is empanelled. If the evidence at the preliminary inquiry is then brought into the trial it becomes part of the evidence of the trial. What we are trying to prevent is a preliminary pre-trial by newspaper prior to the time that a magistrate may have bound a man over for trial. He may find that the charges are dismissed but the damage has been done.

(Minutes of Proceedings and Evidence, No. 11, 1st Sess., 28th Parl., March 18, 1969, at pp. 501-2)

These comments go beyond averting jury bias. They address the broader goal of protecting the right to a fair trial.

[32] A direct, and helpful, analogy can be drawn between the ban at a preliminary inquiry and the one at a bail hearing. In both these types of proceedings, the evidentiary threshold is far lower than proof beyond a reasonable doubt. The Crown need tender only enough evidence to make out a prima facie case, and the defence may for strategic reasons choose not to call witnesses or otherwise challenge the Crown’s evidence. It follows that the publication of proceedings at the preliminary hearing may result in a one-sided view of the case that could have an impact on trial fairness.

[33] In summary, in my view, a rational connection can clearly be found in the interplay between the various components of the reform. I will discuss the benefits of the ban below; it will suffice at this point to mention that the ban prevents the dissemination of evidence which, for the sake of ensuring an expeditious hearing, is untested. This brings me to the third stage of the Oakes analysis, that of minimal impairment, which entails an assessment of the impact of the ban on free expression.

C. Minimal Impairment

[34] In assessing the impact of the ban, the court must consider the nature of the expression at issue (New Brunswick, at para. 63). At this point, I would stress that the bail hearing is conducted at the very beginning of the criminal proceeding. Bail hearings often attract considerable media attention, especially in highly visible cases like the ones presently before this Court.

[35] To determine whether the limit impairs a right as little as possible, it must be assessed in light of the other measures adopted to meet Parliament’s objectives. I touched on this issue above in the discussion on the rational connection. I would stress here that the fairness of the decision whether to grant bail, which is one of the components of the constitutional protection, will depend to a large extent on its timeliness. As pointed out by Rosenberg J.A. in the Ontario case (para. 38) and Slatter J.A. in the Alberta case (para. 36), if the justice were to hold a publication ban hearing, the accused would have to prepare for that hearing in addition to preparing a rebuttal to the grounds the prosecution might raise to justify detaining him or her. The hurdles the accused would face in such a hearing are real.

[36] The authors of the Ouimet Report observed that the initial stage of the process is crucial (pp. 101-2). If a publication ban hearing were to be held, an additional burden would be placed on the accused at a time when he or she is extremely vulnerable. The accused might be a first-time offender who is overwhelmed by the criminal process, and may not have been able to consult his or her counsel of choice. At this point, accused persons will not have had the opportunity to learn what evidence the prosecution intends to adduce. They should be devoting their resources and energy to obtaining their release, not to deciding whether to compromise on liberty in order to avoid having evidence aired outside the courtroom. It is interesting that the Alberta Court of Appeal found it significant that neither the Crown nor Mr. White had attempted to justify a common law restriction in that court, to which the statutory mandatory ban does not apply; the court attributed this to the resources that would have to be expended to support such a restriction (para. 36).

[37] In light of the delay and the resources a publication ban hearing would entail, and of the prejudice that could result if untested evidence were made public, it would be difficult to imagine a measure capable of achieving Parliament’s objectives that would involve a more limited impairment of freedom of expression. If issues unrelated to the release of the accused were added to the bail hearing, this would require the consideration of matters extraneous to the bail process and could have a domino effect on other bail hearings in the same forum, thereby delaying the administration of justice.

[38] It is worth noting that the mandatory publication ban provided for in s. 517 is not an absolute ban either on access to the courts or on publication. The provision only prohibits the publication of evidence adduced, information given, representations made, and reasons given by the justice at a bail hearing. But the media can publish the identity of the accused, comment on the facts and the offence that the accused has been charged with, and that an application for bail has been made, as well as report on the outcome of the application. Journalists are also not prevented from informing the public of the legal conditions attached to the release of the accused.

[39] The temporary nature of the ban is another important factor. The ban ends when the accused is discharged after a preliminary inquiry, or at the end of the trial. In essence, it applies only with respect to the bail process, and the information it covers can eventually be made public once more complete information produced in accordance with the standards applicable to criminal trials is available.

[40] In summary, although information revealed at the bail hearing may no longer be newsworthy by the time the media can release it, the ban cannot be said to impair freedom of expression more than is necessary. The ban is limited to a preliminary stage of the criminal justice process and is not absolute, and the information the media are prevented from publishing is untested, and is often one-sided and largely irrelevant to the search for truth. The ban may make journalists’ work more difficult, but it does not prevent them from conveying and commenting on basic, relevant information.

[41] The appellants suggest alternatives that would in their view be equally effective and trench less on freedom of expression. They submit that any risk to a fair trial can be addressed by means of procedures that arise later in the process, such as challenges for cause, changes of venue and sequestration. The alternatives put forward are unsatisfactory. All of them relate solely to the need to avert bias. They do not address the other considerations which favour a publication ban, namely the need to ensure an expeditious bail hearing and an early release of the accused.

[42] The appellants also suggest that a time-limited publication ban could be imposed at the outset of the bail hearing that would last only until the end of that hearing, at which time a hearing would be held on the merits of continuing the ban. This suggestion is no more acceptable than the first. The parties need to know at the bail hearing if the information given there will be published. Accused persons could alter their approaches to the bail hearing, or perhaps even forego requesting bail, if they know that everything they say could be in the newspaper the next morning.

[43] If they know that evidence could be published, accused persons might have to make decisions they would not otherwise have made at a time when they can only speculate on what the Crown intends to adduce at the bail hearing. Such decisions would take time, would require them to make strategic choices and could compromise their rights to silence and to liberty. In Canadian Newspapers Co. v. Canada (Attorney General), 1988 CanLII 52 (SCC), [1988] 2 S.C.R. 122, this Court upheld s. 442(3) Cr. C., which provided for a mandatory ban on publishing the identity of a complainant in a sexual assault case should the complainant apply for one. Lamer J. (as he then was) stressed the importance of the certainty component of the ban at issue in that case:

Obviously, since fear of publication is one of the factors that influences the reporting of sexual assault, certainty with respect to non-publication at the time of deciding whether to report plays a vital role in that decision. Therefore, a discretionary provision under which the judge retains the power to decide whether to grant or refuse the ban on publication would be counterproductive, since it would deprive the victim of that certainty. Assuming that there would be a lesser impairment of freedom of the press if the impugned provision were limited to a discretionary power, it is clear, in my view, that such a measure would not, however, achieve Parliament’s objective, but rather defeats it. [Emphasis in original; p. 132.]

In the Ontario case, the Ontario Court of Appeal accepted the certainty argument in the context of the bail hearing. I agree with Rosenberg J.A. that, “[w]ithout knowing whether or not the publication ban will be in place, the accused cannot know whether to take the risk of contesting bail to possibly obtain the immediate reward of release at the cost of the more serious risk of poisoning the minds of jurors at the subsequent trial” (para. 63). On this same basis, I would reject the suggestion that the publication of information be authorized at the time of the bail hearing and that a discretionary ban then be imposed closer to the trial date.

[44] It follows that a publication ban hearing, whether held before or after the bail hearing, is not a reasonable alternative to a mandatory ban. I would add that in view of the timing of such a hearing and of how little the accused would know about the information the prosecutor would be conveying, it would be difficult for the accused to discharge the burden of showing that the ban would be “necessary in order to prevent a real and substantial risk to the fairness of the trial” (Dagenais, at p. 878 (emphasis deleted)). Therefore, the justice would have difficulty performing his or her duties in a judicial manner.

[45] For similar reasons, I am also of the opinion that to hold that the ban is discretionary solely with respect to the justice’s reasons would not serve the objectives of the provision. Even if it could be argued that the justice could expeditiously give reasons that would not unduly risk jury bias or otherwise feed dangerous pre‑trial publicity, to impose such constraints on the justice at this time would be unacceptable. This would inevitably invite the parties to present arguments on the discretionary ban and on the content of the reasons likely to be published. Preparing arguments takes time and resources. Moreover, one should not underestimate the complexity of crafting meaningful reasons without disclosing potentially determinative untested evidence at a stage at which neither the justice nor the accused can foresee what information will be prejudicial to the right of the accused to a fair trial.

[46] The fact that Parliament chose to make the ban discretionary where the prosecutor is the applicant does not undermine the legislation’s objectives. The prosecutor is in a better position than the accused to meet the Dagenais test. Unlike the accused, the prosecutor knows exactly what allegations are to be made against the accused and also knows what evidence will likely be introduced at trial. Moreover, the prosecutor does not face the problem of finding resources such as, most importantly, for counsel for both the publication ban and the bail hearings.

[47] The majority of the Ontario Court of Appeal read the provision down in part, limiting the mandatory publication ban to cases in which a jury trial is possible. They concluded that “[f]air trial rights cannot be said to be at risk where a judge, sitting alone, is exposed to prejudicial information which should not be admitted at trial” (para. 185). The dissent would have read out the mandatory nature of the ban. As they pointed out, the practical impact of the majority’s conclusion is limited, since at the time of the bail hearing, the accused has usually not yet made an election and not yet ruled out the possibility of being tried by a jury. If the case involves a hybrid offence, the prosecution may not yet have decided on the mode of prosecution. The fact that only a small percentage of criminal charges are actually heard by a jury is therefore not determinative. Because the bail hearing is held at the beginning of the process, even if the provision is read down as the majority have done, the ban would still apply in the vast majority of cases. Thus, this alternative cannot be accepted. Not only does it fail to respond to the appellants’ concerns, but it fails to settle the timing and resource issues that arise in respect of the proposed publication ban hearing.

[48] One last alternative was suggested: imposing a discretionary ban limited to prejudicial evidence. In my view, this suggestion presents the same difficulties as those that any publication ban hearing would entail. As I mentioned above, at this early stage of the procedure allocation of time and resources should focus on the fairness of the trial of the accused.

[49] I conclude that the mandatory publication ban is integral to a series of measures designed to foster trial fairness and ensure an expeditious bail hearing. It meets the requirements of the minimal impairment stage of the Oakes test. But this does not complete the analysis. The ban must also be found to have benefits that outweigh its deleterious effects.

D. Deleterious Versus Salutary Effects

[50] As the Chief Justice stated in Hutterian Brethren (at para. 76), the final stage of the Oakes analysis is not a reiteration of the previous ones. Whereas the focus of the first three stages is on Parliament’s objective, the final stage concerns the consequences of the impugned measure.

[51] When the consequences of the mandatory ban are considered, several salutary effects can be identified. First of all, it limits the deprivation of the liberty of the accused by confining the issues at the bail hearing to those specifically related to bail. The first day in custody may be overwhelming for an accused; this is especially true if the conditions of detention are unsatisfactory. The potential hardship at this initial stage cannot be underestimated. In R. v. Hall, 2002 SCC 64, [2002] 3 S.C.R. 309, at para. 47, Iacobucci J. (dissenting, but not on this point) stated:

At the heart of a free and democratic society is the liberty of its subjects. Liberty lost is never regained and can never be fully compensated for; therefore, where the potential exists for the loss of freedom for even a day, we, as a free and democratic society, must place the highest emphasis on ensuring that our system of justice minimizes the chances of an unwarranted denial of liberty.

A day in the life of an accused person may have a lifelong impact. In addition to protecting this very important liberty interest, the ban means that accused persons can focus their energy and resources on their liberty interests rather than on their privacy interests. It ensures that they will not renounce their right to liberty in order to protect their reputations. It also ensures that the public will not be influenced by untested, one-sided and stigmatizing information bearing on issues that are often irrelevant to guilt. It ensures consistency with the objectives of other publication bans provided for in the Criminal Code, such as the one under s. 539 concerning evidence produced at a preliminary inquiry.

[52] Another relevant aspect of the bail hearing which is of particular importance in the assessment of the effect of the ban is the fact that the information relevant to interim release often relates to the character of the individual accused and not to the crime (see J. E. Pink and D. C. Perrier, eds., From Crime to Punishment: An Introduction to the Criminal Law System (6th ed. 2007), at p. 92). Evidence will bear not on whether the accused committed the crime, but on the accused as an individual: what kind of person he or she is, and whether he or she is likely to be a danger for society or to appear at trial. This aspect is important where several accused persons have conflicting interests. In such a case, one accused might choose at the bail hearing to denounce a co-accused as the “bad guy” and to adduce additional untested evidence for that purpose (see (2006), 2006 CanLII 25418 (ON SC), 211 C.C.C. (3d) 234, at para. 116, per Durno J.). In such circumstances, a mandatory ban limits the pre-trial disclosure of information that may be inadmissible at trial or highly unreliable.

[53] In addition, two experienced Crown counsel confirmed in affidavits that duty counsel conduct most bail hearings in Ontario; they often prepare for and conduct them under difficult circumstances. For example, they may have to work under serious time constraints and other constraints related to facilities, and the circumstances in which they interview and provide information to the accused are far from ideal. Also, newly arrested accused persons often have only a limited understanding of the court system and the charges facing them, and a limited ability to instruct counsel (see Ontario Court of Appeal reasons, at para. 24, per Rosenberg J.A.).

[54] It is in this context that an accused can apply for a publication ban, and when such an application is made, the justice has no discretion. The ban must be ordered. It applies to the evidence taken, information given and representations made at the hearing, and to any reasons given for the order made under s. 515. It remains in effect until the accused is discharged after a preliminary inquiry or until the end of the trial, as the case may be. A large part of the evidence taken at the bail hearing is presumptively inadmissible at trial. Thus, criminal records, prior consistent statements and post‑offence conduct, which may be mentioned at the bail hearing, might not be admitted in evidence at trial. While it is true that all information about the accused might arouse the public’s curiosity, such information is often irrelevant to the search for truth in relation to the offence, which is the actual purpose of the criminal trial.

[55] A discretionary ban would entail additional issues and adjournments, and would result in longer hearings. Feldman J.A., writing for the majority of the Ontario Court of Appeal in the Ontario case, mentioned a current problem in some overburdened Ontario bail courts (at paras. 198-99): in R. v. Jevons, 2008 ONCJ 559, [2008] O.J. No. 4397 (QL), De Filippis J. had described the backlog plaguing the bail courts and concluded that the accused in that case was the victim of systemic delay in the bail system and that his right to reasonable bail had been infringed. Deficiencies in the organization of certain courts cannot of course justify limits on freedom of expression. However, constraints are a fact of life, and courts must be alert to those constraints when assessing the consequences of any change to a process, such as the inclusion of a publication ban hearing.

[56] Be that as it may, the Ontario case illustrates how quickly an accused can be brought before a justice. The arrests were made on June 2, 2006. Between June 3 and June 12, each of the accused appeared in court at least once; for some of them, it was the third appearance. The bail hearings were held on June 12, 2006. The publication ban was then ordered, and it was to apply to all the accused. In light of notice requirements and the conflicting positions taken by counsel for the various accused, this could not have been done as quickly had a publication ban hearing been held.

[57] The appellants argue that bail hearings would almost never be delayed if the ban were discretionary because the Dagenais test would rarely be met, since jury bias is purely speculative. As a result, counsel would seldom bring motions for bans. This proposition is based on the assumption that accused would renounce their interest in trial fairness to ensure an expeditious hearing. This is exactly the kind of compromise the mandatory ban is intended to avoid. The appellants’ argument is in fact based on the incorrect view that the ban has nothing to do with the rights of the accused to a fair trial and to fair access to bail. It is simply wrong to assume that neither the bail hearing itself nor the disclosure of information, evidence or the reasons for the justice’s order would have any effect on the accused’s interests.

[58] However, the deleterious effects of a publication ban should not be downplayed. Section 517 Cr. C. bars the media from informing the population on matters of interest which could otherwise be subject more widely to public debate: Edmonton Journal v. Alberta (Attorney General), 1989 CanLII 20 (SCC), [1989] 2 S.C.R. 1326. To this extent, it impairs “[t]he freedom of individuals to discuss information about the institutions of government, [and] their policies and practices” (New Brunswick, at para. 18).

[59] The ban prevents full public access to, and full scrutiny of, the criminal justice process. Moreover, the bail hearing may attract considerable media attention and its outcome may not be fully understood by the public, as was apparently the case when Mr. White in the Alberta case and certain of the accused in the Ontario case were initially released. In such cases, the media would be better equipped to explain the judicial process to the public if the information they could convey were not restricted.

[60] Nonetheless, on balance, I must find that in the context of the bail process, the deleterious effects of the limits on the publication of information are outweighed by the need to ensure certainty and timeliness, to conserve resources, and to avert the disclosure of untested prejudicial information; in other words, to guarantee as much as possible trial fairness and fair access to bail. Although not a perfect outcome, the mandatory ban represents a reasonable compromise.

VI. Application to the Co-accused in the Ontario Case

[61] An additional issue is raised in the Ontario case: does the publication ban apply to all the co-accused? The Criminal Code is silent on this point, and a justice must resort to his or her common law jurisdiction. In my view, the Ontario courts correctly exercised their jurisdiction in answering the question in the affirmative. A publication ban will be effective only if it applies to all the accused. Clearly, the evidence against the parties will overlap to a great extent.

VII. Additional Issues in the Alberta Case

[62] In the Alberta case, the appellants raise a number of additional issues based on jurisdiction, mootness and functus officio. They also mention the Court of Appeal’s criticism of the choice to raise the constitutional challenge in the context of a criminal proceeding rather than by bringing a civil action.

[63] Although these issues may give rise to interesting comments, the appellants expressly acknowledge that this Court does not need to rule on them, because it has jurisdiction to entertain the substance of the appeal in any event. It is therefore not necessary to deal with them.

VIII. Conclusion

[64] For these reasons, I would conclude that s. 517 Cr. C. infringes freedom of expression but that the limit can be demonstrably justified in a free and democratic society. Since the appellants sought to strike down the mandatory wording of the provision, I would dismiss their appeals. I would uphold the constitutionality of the provision and allow the cross-appeal in the Ontario case.

The following are the reasons delivered by

[65] Abella J. (dissenting) — This Court has a long pedigree in protecting the public’s right to be aware of what takes place in the country’s courtrooms. It is based on the premise that to maintain public trust in the justice system, the public must be able to see the judicial process at work. The public’s ability to engage in meaningful discussion about what a judge decides, depends primarily on knowing why the particular decision is made. The jurisprudence has, as a result, consistently attempted to enhance both the visibility of the system and the confidence of the public. This emphasis on transparency is known as the open court principle. As Cory J. stated in Edmonton Journal v. Alberta (Attorney General), 1989 CanLII 20 (SCC), [1989] 2 S.C.R. 1326:

[M]embers of the public have a right to information pertaining to public institutions and particularly the courts. . . . Only then can they make an assessment of the institution. Discussion of court cases and constructive criticism of court proceedings is dependent upon the receipt by the public of information as to what transpired in court. Practically speaking, this information can only be obtained from the newspapers or other media. [Emphasis added; pp. 1339‑40.]

(See also Attorney General of Nova Scotia v. MacIntyre, 1982 CanLII 14 (SCC), [1982] 1 S.C.R. 175; Vickery v. Nova Scotia Supreme Court (Prothonotary), 1991 CanLII 90 (SCC), [1991] 1 S.C.R. 671; Canadian Broadcasting Corp. v. New Brunswick (Attorney General), 1996 CanLII 184 (SCC), [1996] 3 S.C.R. 480; R. v. Mentuck, 2001 SCC 76, [2001] 3 S.C.R. 442; Sierra Club of Canada v. Canada (Minister of Finance), 2002 SCC 41, [2002] 2 S.C.R. 522; Ruby v. Canada (Solicitor General), 2002 SCC 75, [2002] 4 S.C.R. 3; Vancouver Sun (Re), 2004 SCC 43, [2004] 2 S.C.R. 332; Toronto Star Newspapers Ltd. v. Ontario, 2005 SCC 41, [2005] 2 S.C.R. 188; Named Person v. Vancouver Sun, 2007 SCC 43, [2007] 3 S.C.R. 253.)

[66] The issue in this case is whether a ban on bail hearing information should be mandatory in every case where it is requested by the accused, or whether it should only be imposed when the accused is able to demonstrate that his or her fair trial interests demand it. I have had the benefit of reading the reasons of Deschamps J., but with great respect, I am of the view, like Rosenberg J.A. in the Ontario Court of Appeal (2009 ONCA 59, 94 O.R. (3d) 82), and largely for his reasons, that the mandatory ban in s. 517 of the Criminal Code, R.S.C. 1985, c. C-46, does not survive the final stage of the Oakes test. I also agree with him that the appropriate remedy is to sever the mandatory aspect of s. 517 and leave in place the discretion to order a ban, to be exercised in accordance with Dagenais v. Canadian Broadcasting Corp., 1994 CanLII 39 (SCC), [1994] 3 S.C.R. 835, and Mentuck.

[67] Section 517 of the Criminal Code automatically prevents disclosure of the judge’s reasons and of any information at a bail hearing whenever the accused requests a ban. Neither the public nor the press is prevented from actually being in court during the hearing, but what is mandatorily prohibited is the public dissemination of what is disclosed there until the trial is complete, a chronology that can take years to unfold. This has the effect, for all but the handful of people who are present in the courtroom, of denying access to information surrounding a key aspect of the criminal justice system — the decision whether or not to release an accused back into the community pending his or her trial. This denial is a profound interference with the open court principle.

[68] The seriousness of the infringement was compellingly summarized by Rosenberg J.A. as follows:

Section 517 cuts off meaningful and informed public debate about a fundamental aspect of the administration of criminal justice, the bail system, at the very time that the debate may be most important — when the decision is made to grant or deny bail. It also hinders debate in other circumstances of great public interest, as where an accused on bail commits another, perhaps serious crime. The public is left to speculate about why the accused was released and the justice system is unable to provide a timely and meaningful response because of the statutorily imposed silence. [para. 32]

[69] The importance of public awareness of the bail decision‑making process was stressed in R. v. Hall, 2002 SCC 64, [2002] 3 S.C.R. 309, where McLachlin C.J. expressed what is to me the underlying and driving principle in considering whether the mandatory nature of the ban in s. 517 is justified:

Where justice is not seen to be done by the public, confidence in the bail system and, more generally, the entire justice system may falter. [para. 26]

[70] This brings us to whether this harm is outweighed by the benefits of a mandatory ban. The major benefits attributed to the ban are the reduction in pre-trial publicity and delay. Each of these concerns, in my view, can largely be attenuated, and neither represents a sufficiently serious infringement of fair trial rights.

[71] Concerns over pre-trial publicity were addressed by this Court when it considered the question of discretionary bans in Dagenais and Mentuck. The new threshold articulated in those cases was a high one, and bans were only to be imposed where they are “necessary” to protect against “real and substantial” risks to an accused’s fair trial rights (Dagenais, at p. 878), or “serious” risks to the administration of justice (Mentuck, at para. 32). Section 517, in granting an automatic ban at the request of an accused regardless of whether he or she can demonstrate such a degree of risk, completely collapses the constitutional framework in Dagenais/Mentuck, leaving out of the balance entirely the public’s presumptive right to know what goes on in a courtroom.

[72] Even if one is disinclined to accept what is to me the cogent evidence in the reasons of Rosenberg J.A. demonstrating how speculative the concerns over pre-trial publicity are, there remains the possibility of remedies such as a partial ban, challenges for cause, or a change of venue if there is a sufficient risk of prejudice. We should also be able to rely on the ability of a properly instructed jury to disregard irrelevant evidence, a reliance that is at the foundation of our belief in juries in criminal trials (Dagenais, at pp. 884-85; see also R. v. Corbett, 1988 CanLII 80 (SCC), [1988] 1 S.C.R. 670, at pp. 692-93; R. v. Vermette, 1988 CanLII 87 (SCC), [1988] 1 S.C.R. 985, at pp. 992-93). As Berger J.A. observed in the course of related proceedings in the Alberta appeal, where the accused had been released on bail after being charged with the murder of his pregnant wife:

The Applicant has been charged with second‑degree murder. His preliminary hearing will not take place until the new year. If committed to stand trial, jury selection would begin months later. I very much doubt that prospective jurors would retain and recall the details of a 30 second news clip or a seven inch column summarizing submissions made by counsel, or reasons for decision pronounced by a bail judge. Even if some did, the usual admonitions to the array, challenges for cause, and jury instructions themselves, are, in my opinion, sufficient safeguards to ensure that an impartial jury, true to their oaths, will be empanelled.

(R. v. White, 2005 ABCA 435, 56 Alta. L.R. (4th) 255, at para. 17)

[73] In any event, s. 517 only protects an accused from disclosure of pre-trial information from a bail hearing. There is no legislative protection from potentially prejudicial pre-trial information that emanates from sources other than the bail hearing. In the absence of such a generalized ban, the benefit of a ban only on bail hearing information seems to me to be too porous to justify the seriousness of the infringement.

[74] The second benefit of a mandatory ban is said to be the reduction in delay. This is a concern only if one assumes that the request for a discretionary ban will substantially elongate the hearing. It is hard for me to see what evidence would routinely result in a protracted bail proceeding. A delay would also only result if one assumes that the media is entitled to notice in every case. I make no such assumption. While the decision to give notice is a matter of discretion for the judge hearing the bail application, it seems to me to be unrealistic to expect that submissions be invited from the media in every case where a publication ban in a bail hearing is requested.

[75] In the absence of such automatic notice, there will be, in the overwhelming number of cases, no undue delay. Those cases where the media is most likely to contest a ban are those few which have a higher profile. But I would not judge the desirability of a universal mandatory ban based on its effectiveness for a small percentage of cases.

[76] Public confidence in the justice system requires relevant information delivered in a timely way. A mandatory ban on the evidence heard and the reasons given in a bail application is a ban on the information when it is of most concern and interest to the public. Restrictions on the release of such information are only justified if their benefits outweigh their detrimental impact.

[77] Given that the salutary effects of the ban under s. 517 are not proportional to the harmful effects flowing from the infringement of the open court principle, I would allow the appeal in both the Ontario and Alberta cases and strike out the words in s. 517 that render the ban mandatory.

Appeals dismissed and cross-appeal allowed, Abella J. dissenting.

Solicitors for the appellants/respondents on cross‑appeal Toronto Star Newspapers Ltd. et al.: Blake, Cassels & Graydon, Toronto.

Solicitors for the appellants Canadian Broadcasting Corporation et al.: Reynolds, Mirth, Richards & Farmer, Edmonton.

Solicitors for the appellant Edmonton Sun, a Division of Sun Media Corporation: Fraser Milner Casgrain, Edmonton.

Solicitor for the respondent/appellant on cross‑appeal Her Majesty The Queen in Right of Canada and for the intervener the Director of Public Prosecutions of Canada: Public Prosecution Service of Canada, Brampton.

Solicitor for the respondent/appellant on cross‑appeal A.A.: John Norris, Toronto.

Solicitor for the respondent F.A.: Dennis Edney, Edmonton.

Solicitor for the respondent S.A.: Peter G. Martin, Eganville, Ontario.

Solicitor for the respondent Qayyum Abdul Jamal: Anser Farooq, Toronto.

Solicitors for the respondents A.M.D. and Ahmad Mustafa Ghany: Rocco Galati Law Firm Professional Corporation, Toronto.

Solicitors for the respondent S.V.C.: Marlys Edwardh Barristers Professional Corporation, Toronto.

Solicitor for the respondent Her Majesty The Queen and for the intervener the Attorney General of Alberta: Attorney General of Alberta, Calgary.

Solicitors for the respondent Michael James White: Dawson Stevens & Shaigec, Edmonton.

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

Solicitors for the interveners N.S. and N.Y.: Hicks, Block, Adams, Toronto.

Solicitors for the intervener the Canadian Civil Liberties Association: McCarthy Tétrault, Toronto.

Solicitors for the interveners the Canadian Newspaper Association et al.: Owen Bird Law Corporation, Vancouver.