SUPREME COURT OF CANADA

Appeal heard: December 2, 2015 Judgment rendered: July 21, 2016 Docket: 36200

Between:

K.R.J.

Appellant

and

Her Majesty The Queen

Respondent

• and -

Attorney General of Canada, Attorney General of Ontario,

Association des avocats de la défense de Montréal,

David Asper Centre for Constitutional Rights,

Criminal Lawyers’ Association (Ontario) and

British Columbia Civil Liberties Association

Interveners

Coram: McLachlin C.J. and Abella, Cromwell, Moldaver, Karakatsanis, Wagner, Gascon, Côté and Brown JJ.

Reasons for Judgment: (paras. 1 to 116) Reasons Dissenting in Part: (paras. 117 to 130) Reasons Dissenting in Part: (paras. 131 to 162)

Karakatsanis J. (McLachlin C.J. and Cromwell, Moldaver, Wagner, Gascon and Côté JJ. concurring) Abella J. Brown J.

R. v. K.R.J., 2016 SCC 31, [2016] 1 S.C.R. 906

K.R.J. Appellant

v.

Her Majesty The Queen Respondent

and

Attorney General of Canada,

Attorney General of Ontario,

Association des avocats de la défense de Montréal,

David Asper Centre for Constitutional Rights,

Criminal Lawyers’ Association (Ontario) and

British Columbia Civil Liberties Association Interveners

Indexed as: R. v. K.R.J.

2016 SCC 31

File No.: 36200.

2015: December 2; 2016: July 21.

Present: McLachlin C.J. and Abella, Cromwell, Moldaver, Karakatsanis, Wagner, Gascon, Côté and Brown JJ.

on appeal from the court of appeal for british columbia

Constitutional law — Charter of Rights — Benefit of lesser punishment — Sentencing — Accused pleaded guilty to incest and making child pornography — Retrospective application of amendments to Criminal Code expanding scope of community supervision measures sentencing judge can impose on sexual offenders — Offences committed prior to amendments but accused sentenced after — Whether new prohibition measures contained in Criminal Code constitute punishment such that their retrospective operation limits right protected by s. 11(i) of Charter — If so, whether limit is justified — Reformulation of s. 11(i) test for punishment — Canadian Charter of Rights and Freedoms, ss. 1 , 11(i) — Criminal Code, R.S.C. 1985, c. C‑46, s. 161(1) (c), (d).

Section 11( i ) of the Charter provides that, if the punishment for an offence is varied after a person commits the offence, but before sentencing, the person is entitled to “the benefit of the lesser punishment”. When offenders are convicted of certain sexual offences against a person under the age of 16 years, s. 161(1) of the Criminal Code gives sentencing judges the discretion to prohibit them from engaging in a variety of everyday conduct upon their release into the community, subject to any conditions or exemptions the judge considers appropriate. In 2012, Parliament expanded the scope of s. 161(1) , empowering sentencing judges to prohibit sexual offenders from having any contact with a person under 16 years of age ( s. 161(1) (c)) or from using the Internet or other digital network ( s. 161(1) (d)). In doing so, Parliament intended to give sentencing judges the discretion to impose the expanded prohibition measures on all offenders, even those who offended before the amendments came into force. In March 2013, the accused pleaded guilty to incest and the creation of child pornography. The offences were committed between 2008 and 2011. By virtue of the convictions and the age of the victim, the sentencing judge was required to consider whether to impose a prohibition under s. 161(1) . The question arose as to whether the 2012 amendments could operate retrospectively such that they could be imposed on the accused.

The sentencing judge concluded that an order under the new s. 161(1)(c) and (d) constitutes punishment within the meaning of s. 11( i ) of the Charter , such that the provisions cannot be applied retrospectively. He therefore imposed a prohibition order under s. 161, but limited the prohibited activities to those described in the version of s. 161(1) that existed when the accused committed the offences. On the Crown appeal, the majority of the Court of Appeal concluded that the 2012 amendments were enacted to protect the public, rather than to punish offenders, and therefore, they do not qualify as punishment within the meaning of s. 11 ( i ). The majority allowed the appeal and imposed the conditions in s. 161(1) (c) and (d) retrospectively on the accused.

Held (Abella and Brown JJ. dissenting in part): The appeal should be allowed in part. The amendments to s. 161(1) (c) and (d) of the Criminal Code qualify as punishment such that their retrospective operation limits the right protected by s. 11( i ) of the Charter . Under s. 1 of the Charter , while the retrospective operation of the no contact provision in s. 161(1)(c) is not a reasonable limit on the s. 11( i ) right, the retrospective operation of the Internet prohibition in s. 161(1) (d) is a reasonable limit. Accordingly, the appeal should be allowed with respect to s. 161(1) (c), but dismissed with respect to s. 161(1) (d).

Per McLachlin C.J. and Cromwell, Moldaver, Karakatsanis, Wagner, Gascon and Côté JJ.: Section 11( i ) of the Charter constitutionally enshrines the fundamental notion that criminal laws should generally not operate retrospectively. This constitutional aversion for retrospective criminal laws is primarily motivated by the desire to protect the fairness of criminal proceedings and safeguard the rule of law. Rules pertaining to criminal punishment should be clear and certain. To attract the protection of s. 11 ( i ), the new prohibition measures must qualify as “punishment”. In R. v. Rodgers , 2006 SCC 15 , [2006] 1 S.C.R. 554, this Court developed a two‑part test for determining whether a consequence amounts to punishment under s. 11 ( i ): (1) the measure must be a consequence of a conviction that forms part of the arsenal of sanctions to which an accused may be liable in respect of a particular offence; and (2) it must be imposed in furtherance of the purpose and principles of sentencing.

This test requires two clarifications. First, while not all measures imposed to protect the public constitute punishment, public protection is at the core of the purpose and principles of sentencing and is therefore an insufficient litmus test for defining punishment. Thus, sanctions intended to advance public safety do not constitute a broad exception to the protection s. 11 ( i ) affords and may qualify as punishment. Second, the s. 11 ( i ) test for punishment must embody a clearer, more meaningful consideration of the impact a sanction can have on an offender. Doing so enhances fairness and predictability in punishment and is consistent with this Court’s jurisprudence.

Accordingly, the s. 11 ( i ) test for punishment should be restated as follows: a measure constitutes punishment if (1) it is a consequence of conviction that forms part of the arsenal of sanctions to which an accused may be liable in respect of a particular offence, and either (2) it is imposed in furtherance of the purpose and principles of sentencing, or (3) it has a significant impact on an offender’s liberty or security interests. To satisfy the third branch of this test, a consequence of conviction must significantly constrain a person’s ability to engage in otherwise lawful conduct or impose significant burdens not imposed on other members of the public.

Applying this reformulated test, the 2012 amendments to s. 161(1) constitute punishment. The prohibitions found in these amendments are a consequence of conviction, imposed in furtherance of the purpose and principles of sentencing, and they can have a significant impact on the liberty and security of offenders. Clearly, the 2012 amendments constitute greater punishment than the previous prohibitions. Accordingly, the retrospective operation of these provisions limits the s. 11 ( i ) right as it deprives the accused of the benefit of the less restrictive community supervision measures captured in the previous version of s. 161 — that is, the lesser punishment.

To be justified under s. 1 of the Charter , a law that limits a constitutional right must do so in pursuit of a sufficiently important objective that is consistent with the values of a free and democratic society. The legislative history, judicial interpretation, and design of s. 161 all confirm that the overarching goal of the section is to protect children from sexual violence perpetrated by recidivists. It follows naturally that the objective of the retrospective operation of the 2012 amendments — the infringing measure — is to better protect children from the risks posed by offenders like the accused who committed their offences before, but were sentenced after, the amendments came into force. This latter objective anchors the s. 1 analysis and is of sufficient importance to warrant further scrutiny.

There is clearly a rational connection between this objective and retrospectively giving sentencing judges the discretionary power to limit those offenders who pose a continuing risk to children in contacting children in person or online, and in engaging with online child pornography (the means chosen). Reason and logic suffice to establish that Parliament proceeded rationally in opting to give s. 161(1) (c) and (d) retrospective effect. Further, given the discretionary and tailored nature of s. 161 and the fact that a purely prospective application of the amendments would have compromised Parliament’s full objective, the retrospective operation of s. 161(1) (c) and (d) impairs the s. 11 ( i ) rights as little as reasonably possible.

Finally, the deleterious and salutary effects of the law must be assessed. This final stage of the proportionality inquiry is important because it allows courts to transcend the law’s purpose and engage in a robust examination of the law’s impact on Canada’s free and democratic society in direct and explicit terms. Although this examination entails difficult value judgments, it is preferable to make these judgments explicit, as doing so enhances the transparency and intelligibility of the ultimate decision. While the minimal impairment test has come to dominate much of the s. 1 discourse in Canada, this final step permits courts to address the essence of the proportionality enquiry at the heart of s. 1.

The deleterious effects flowing from the retrospective operation of s. 161(1)(c) are substantial. The new s. 161(1)(c) goes much further and prohibits any contact — including communicating by any means — with a person who is under the age of 16 years in a public or private space. By impacting people like the accused with a punishment of which they had no notice, the retrospective operation of s. 161(1) (c) undermines fairness in criminal proceedings and compromises the rule of law. Unfortunately, sexual offences against children have persisted for centuries. The Crown has failed to lead much, if any, evidence to establish the degree of enhanced protection s. 161(1) (c) provides in comparison to the previous version of the prohibition. The benefits society stands to gain are marginal and speculative. The Crown has provided no temporal justification for the retrospective limitation, yet, at its root, s. 11 ( i ) is about the timing of changes to penal laws. The retrospective operation of s. 161(1) (c) therefore cannot be justified under s. 1. As a result, s. 161(1) (c) should apply only prospectively — that is, only to offenders who committed their offences after the 2012 amendments came into force.

The deleterious effects resulting from the retrospective operation of s. 161(1)(d) are also significant. A complete ban on using the Internet or other digital network is more intrusive than the previous ban on using a computer system for the purpose of communicating with young people. As with the retrospective operation of s. 161(1) (c), the imposition of punishment without notice translates into broader societal harms, including compromising the fairness of criminal proceedings and challenging the rule of law. However, s. 161(1)(d) is directed at grave, emerging harms precipitated by a rapidly evolving social and technological context. This evolving context has changed both the degree and nature of the risk of sexual violence facing young persons. As a result, the previous iteration of s. 161 became insufficient to respond to the modern risks children face. By closing this legislative gap and mitigating these new risks, the benefits of the retrospective operation of s. 161(1)(d) are significant and fairly concrete. The previous prohibition was insufficient to address the evolving risks. On balance, Parliament was justified in giving s. 161(1)(d) retrospective effect in the unique context within which it was legislating. The harms at stake are particularly powerful. The statutory regime is highly tailored and discretionary. An Internet prohibition, while invasive, is not among the most onerous punishments, such as increased incarceration. The benefits of the law outweigh its deleterious effects.

In summary, the 2012 amendments to s. 161(1)(c) and (d) qualify as punishment based on both the objective and impact of the prohibitions. The retrospective imposition of these prohibitions therefore limits the right protected by s. 11( i ) of the Charter . While the retrospective operation of the no contact provision in s. 161(1)(c) is not a reasonable limit on the s. 11 ( i ) right, the retrospective operation of the Internet prohibition in s. 161(1)(d) is a reasonable limit.

Per Abella J. (dissenting in part): The Charter breach of s. 161(1)(d) cannot be justified. The wording of s. 11 ( i ) is unequivocal. The absolutist language used by the drafters of the Charter in s. 11 must colour the s. 1 analysis by demanding the most stringent of justifications.

The Crown has the highest possible evidentiary burden, namely, to demonstrate through compelling evidence that the previous provisions so significantly undermined the government’s objectives, that the retrospective application of the greater punishment was justified. The Crown’s evidentiary record here was insufficient to justify the retrospective application of the impugned provisions. Far from offering compelling evidence, the Crown offered no evidence in the context of s. 161(1)(d), to show that the former provisions so significantly undermined its objectives, that the retroactive application of greater restrictions was justified. If all that is needed to justify a breach of s. 11 ( i ) is the suggestion of a possible reduction in recidivism rates, whether based on changes in technology or otherwise, the state could, in theory, justify the retrospective application of more stringent punishments so routinely that s. 11( i ) is written out of the Charter . In this case, there was no evidence about how the retrospective application of s. 161(1)(d) was expected to, or would, reduce recidivism rates any more than those under the former restrictions. As a result, while there is agreement with the majority that both s. 161(1) (c) and (d) of the Criminal Code violate s. 11( i ) of the Charter and that s. 161(1)(c) cannot be justified under s. 1, neither can s. 161(1)(d) be justified.

Per Brown J. (dissenting in part): There is agreement with the majority that the conditions which a sentencing judge may impose under s. 161(1) (c) and (d) of the Criminal Code constitute punishment within the meaning of s. 11( i ) of the Charter and that their retrospective application infringes s. 11( i ) . There is also agreement that the Crown has met its burden of justifying the infringement of s. 11 ( i ) in respect of the conditions relating to Internet use contained in s. 161(1) (d). However, the Crown has also done so in respect of the conditions imposable under s. 161(1) (c) relating to contact with children. The retrospective application of both conditions should therefore be upheld under s. 1 of the Charter .

The harm addressed by s. 11( i ) is not the punishment itself, but rather the means by which it is imposed. This means‑based quality of the s. 11 ( i ) protection affects the analysis to be applied under s. 1, since the Oakes analysis considers the proportionality between a legislative objective and the Charter ‑infringing effects resulting from its pursuit, not the choice of means that, by itself, constitutes a Charter infringement. The Oakes test is not, and should not be treated as, a technical inquiry. The majority’s rigid and acontextual application of Oakes causes it to lose sight of the broader context and overall goals sought by Parliament. It holds Parliament to an exacting standard of proof, thereby denying Parliament the room necessary to perform its legislative policy‑development role when addressing a chronic social problem. And it also insists on direct evidence of anticipated benefits which, given that chronic nature of the harm, is likely impossible to obtain.

A broad examination of Parliament’s purpose is necessary in order to anchor a useful proportionality analysis because of the unique means‑based quality of s. 11( i )’s protection. The measure that gave rise to the Charter infringement, and which should anchor the proportionality analysis, comprises the amendments to s. 161 as a whole. And, as to that measure, the majority’s characterization of the objective should be accepted: the objective is to enhance the protection s. 161 affords to children against the risk of harm posed by sexual offenders. The retrospective application of these amendments is rationally connected to that protective purpose, since the risk an offender poses to reoffend sexually against children is not affected by whether the offence occurred before or after the measure’s enactment. And, given Parliament’s objective of enhancing the protections that s. 161 affords to children, there are no less‑impairing alternate measure that would allow for s. 161(1) ’s protections to be realized in respect of an offender who committed his or her offence before the amendments came into force and who poses a risk to reoffend.

The final stage of the proportionality analysis is tied to the practical impacts and benefits of the law, but what is ultimately being weighed is much more abstract and philosophical: the detriment to Charter ‑protected rights against the public benefit sought. Insisting upon too strict an evidentiary burden must be carefully avoided. However, the majority does precisely that by demanding empiricism where none can exist. Given the complex social context in which Parliament develops policy, it will sometimes be difficult, if not impossible, for the state to provide reliable and direct evidence of the benefits its measure will achieve.

The majority errs by overstating the deleterious effects of s. 161(1)(c)’s retrospective operation while understating its salutary effects. Section 161(1)(c) prohibits only unsupervised contact with children, and is subject to any other exemptions that the sentencing judge sees fit to impose. The majority’s interpretation of the restriction on liberty worked by s. 161(1)(c) is over‑expansive and is at odds with the well‑established principle that the criminal law’s prohibitions on conduct should be construed strictly. Further, the majority’s insistence on a compelling temporal justification for the retrospective operation of s. 161(1) (c) when assessing the deleterious impact of its retrospective operation on the rule of law is inappropriate. The majority is, in substance questioning whether Parliament’s objective in enacting a retrospective increase in punishment was truly pressing and substantial. Temporal considerations are not relevant when assessing the deleterious effect of a retrospective punishment on the rule of law because all retrospective changes to the law derogate from the rule of law, irrespective of Parliament’s reasons for enacting them.

As to the salutary effects, the risk posed to children by offenders like the accused simply cannot be mitigated by the original version of s. 161(1) . The evidence before Parliament showed that a majority of sexual offences against children were committed by family members or acquaintances. The previous version of s. 161(1) could not be used to restrict an offender’s ability to interact with children in private, even if that is where the offender poses the greatest risk to reoffend sexually against children. The salutary effects of s. 161(1) (c)’s retrospective operation seem manifest.

All the reasons identified by the majority in support of the conclusion that the limit imposed on the s. 11 ( i ) right by the retrospective application of s. 161(1) (d) is justified are equally applicable to the retrospective application of s. 161(1) (c). The condition in s. 161(1) (c) is also highly tailored and discretionary, since it is imposed only where the sentencing judge deems it necessary, and also since it is subject to such exemptions as the sentencing judge sees fit to allow. If the retrospective operation of s. 161(1) (d) is a proportional and justified limit on an offender’s s. 11 ( i ) right, the retrospective operation of s. 161(1) (c) must be as well.

Balancing the salutary and deleterious effects of a Charter ‑infringing law is not an objective calculation because it requires the court to weigh incommensurables — in this case, to weigh the deleterious impact on the sexual offender and the rule of law against the possible benefit of protecting children from sexual offenders. However, despite the impossibility of weighing incommensurables objectively, a reviewing court must nevertheless come to a reasoned conclusion. The salutary effects pursued in this case are worth the cost in rights limitation: the harms sought to be addressed are grave, persistent, and worthy of Parliament’s efforts in the criminal law realm. The provisions are sufficiently tailored so that no offender’s s. 11 ( i ) rights will be unduly limited. Neither of the impugned provisions works a drastic increase in the punishment imposed. On balance, the potential salutary effect of the retrospective operation of s. 161(1)(c) and (d) of better protecting children from all sexual offenders who pose a risk to reoffend sexually against them, regardless of when the offender committed a designated offence, outweighs the modest impact on fairness and the rule of law.

Cases Cited

By Karakatsanis J.

Discussed: R. v. Rodgers , 2006 SCC 15 , [2006] 1 S.C.R. 554; referred to: R. v. Dineley , 2012 SCC 58 , [2012] 3 S.C.R. 272; R. v. Wigglesworth , 1987 41 (SCC) , [1987] 2 S.C.R. 541; Black‑Clawson International Ltd. v. Papierwerke Waldhof‑Aschaffenburg A.G. , [1975] A.C. 591 ; R. v. Kelly , 1992 62 (SCC) , [1992] 2 S.C.R. 170; R. v. Murrins , 2002 NSCA 12 , 201 N.S.R. (2d) 288; R. v. Lacasse , 2015 SCC 64 , [2015] 3 S.C.R. 1089; R. v. Ipeelee , 2012 SCC 13 , [2012] 1 S.C.R. 433; R. v. Lyons , 1987 25 (SCC) , [1987] 2 S.C.R. 309; R. v. Hooyer , 2016 ONCA 44 , 129 O.R. (3d) 81; R. v. Cross , 2006 NSCA 30 , 138 C.R.R. (2d) 163; R. v. Big M Drug Mart Ltd. , 1985 69 (SCC) , [1985] 1 S.C.R. 295; Canada (Attorney General) v. Whaling , 2014 SCC 20 , [2014] 1 S.C.R. 392; Cunningham v. Canada , 1993 139 (SCC) , [1993] 2 S.C.R. 143; R. v. Heywood (1992), 1992 6008 (BC CA) , 20 B.C.A.C. 166, aff’d 1994 34 (SCC) , [1994] 3 S.C.R. 761; R. v. A. (R.K.) , 2006 ABCA 82 , 208 C.C.C. (3d) 74; R. v. Perron , 2009 ONCA 498 , 244 C.C.C. (3d) 369; R. v. R.R.B. , 2013 BCCA 224 , 338 B.C.A.C. 106; R. v. Levin , 2015 ONCJ 290 ; R. v. Schledermann , 2014 ONSC 674 ; R. v. Proulx , 2000 SCC 5 , [2000] 1 S.C.R. 61; Carter v. Canada (Attorney General) , 2015 SCC 5 , [2015] 1 S.C.R. 331; R. v. Oakes , 1986 46 (SCC) , [1986] 1 S.C.R. 103; R. v. Sharpe , 2001 SCC 2 , [2001] 1 S.C.R. 45; Toronto Star Newspapers Ltd. v. Canada , 2010 SCC 21 , [2010] 1 S.C.R. 721; R. v. Moriarity , 2015 SCC 55 , [2015] 3 S.C.R. 485; R. v. Budreo (2000), 2000 5628 (ON CA) , 46 O.R. (3d) 481; RJR‑MacDonald Inc. v. Canada (Attorney General) , 1995 64 (SCC) , [1995] 3 S.C.R. 199; Alberta v. Hutterian Brethren of Wilson Colony , 2009 SCC 37 , [2009] 2 S.C.R. 567; Dagenais v. Canadian Broadcasting Corp. , 1994 39 (SCC) , [1994] 3 S.C.R. 835; Thomson Newspapers Co. v. Canada (Attorney General) , 1998 829 (SCC) , [1998] 1 S.C.R. 877; McKinney v. University of Guelph , 1990 60 (SCC) , [1990] 3 S.C.R. 229.

By Abella J. (dissenting in part)

Canada (Attorney General) v. Whaling , 2014 SCC 20 , [2014] 1 S.C.R. 392; Liang v. Canada (Attorney General) , 2014 BCCA 190 , 311 C.C.C. (3d) 159.

By Brown J. (dissenting in part)

R. v. Oakes , 1986 46 (SCC) , [1986] 1 S.C.R. 103; R. v. Keegstra , 1990 24 (SCC) , [1990] 3 S.C.R. 697; RJR‑MacDonald Inc. v. Canada (Attorney General) , 1995 64 (SCC) , [1995] 3 S.C.R. 199; Alberta v. Hutterian Brethren of Wilson Colony , 2009 SCC 37 , [2009] 2 S.C.R. 567; Public School Boards’ Assn. of Alberta v. Alberta (Attorney General) , 2000 SCC 2 , [2000] 1 S.C.R. 44; M. v. H. , 1999 686 (SCC) , [1999] 2 S.C.R. 3; Mounted Police Association of Ontario v. Canada (Attorney General) , 2015 SCC 1 , [2015] 1 S.C.R. 3; Harper v. Canada (Attorney General) , 2004 SCC 33 , [2004] 1 S.C.R. 827; R. v. Butler , 1992 124 (SCC) , [1992] 1 S.C.R. 452; Thomson Newspapers Co. v. Canada (Attorney General) , 1998 829 (SCC) , [1998] 1 S.C.R. 877; R. v. Sharpe , 2001 SCC 2 , [2001] 1 S.C.R. 45; R. v. Bryan , 2007 SCC 12 , [2007] 1 S.C.R. 527; R. v. McIntosh , 1995 124 (SCC) , [1995] 1 S.C.R. 686; R. v. Orbanski , 2005 SCC 37 , [2005] 2 S.C.R. 3; R. v. St‑Onge Lamoureux , 2012 SCC 57 , [2012] 3 S.C.R. 187; New Jersey v. T.L.O. , 469 U.S. 325 (1985).

Statutes and Regulations Cited

Act to amend the Criminal Code and the Young Offenders Act , S.C. 1993, c. 45, s. 1.

Canadian Charter of Rights and Freedoms , ss. 1, 11 .

Constitution Act, 1982 , s. 52(1).

Criminal Code , R.S.C. 1985, c. C‑46, ss. 161, 258(1) (d.1), 718 et seq., 718.1, 718.2, 810, 810.1, 810.2.

Safe Streets and Communities Act , S.C. 2012, c. 1, s. 16(1).

Authors Cited

Areheart, Bradley Allan, and Michael Ashley Stein. “Integrating the Internet” (2015), 83 Geo. Wash. L. Rev. 449.

Babchishin, Kelly M., R. Karl Hanson and Heather VanZuylen. “Online Child Pornography Offenders are Different: A Meta‑analysis of the Characteristics of Online and Offline Sex Offenders Against Children” (2015), 44 Arch. Sex. Behav. 45.

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

Barak, Aharon. “Proportionality and Principled Balancing” (2010), 4 L. & Ethics Hum. Rts. 1.

Beatty, David M. The Ultimate Rule of Law . Oxford: Oxford University Press, 2004.

Cameron, Jamie. “The Past, Present, and Future of Expressive Freedom Under the Charter ” (1997), 35 Osgoode Hall L.J. 1.

Canada. House of Commons. House of Commons Debates , vol. 145, No. 110, 3rd Sess., 40th Parl., December 3, 2010, p. 6787.

Canada. House of Commons. House of Commons Debates , vol. 145, No. 144, 3rd Sess., 40th Parl., March 11, 2011, p. 8967.

Canada. House of Commons. Standing Committee on Justice and Human Rights. Evidence , No. 44, 3rd Sess., 40th Parl., January 31, 2011, pp. 5‑6.

Canada. House of Commons. Standing Committee on Justice and Human Rights. Evidence , No. 46, 3rd Sess., 40th Parl., February 7, 2011, pp. 5‑6.

Canada. House of Commons. Standing Committee on Justice and Human Rights. Evidence , No. 49, 3rd Sess., 40th Parl., February 16, 2011, p. 7.

Canada. House of Commons. Standing Committee on Justice and Human Rights. Evidence , No. 50, 3rd Sess., 40th Parl., February 28, 2011, p. 4.

Canada. Public Safety and Emergency Preparedness. “Predictors of Sexual Recidivism: An Updated Meta‑Analysis”, by R. Karl Hanson and Kelly Morton‑Bourgon, February 2004 (online: www.publicsafety.gc.ca/cnt/rsrcs/pblctns/2004‑02‑prdctrs‑sxl‑rcdvsm‑pdtd/index‑en.aspx).

Canada. Public Safety and Emergency Preparedness. “Sex Offender Recidivism: A Simple Question”, by Andrew J. R. Harris and R. Karl Hanson, March 2004 (online: www.publicsafety.gc.ca/cnt/rsrcs/pblctns/sx‑ffndr‑rcdvsm/index‑en.aspx).

Choudhry, Sujit. “So What Is the Real Legacy of Oakes ? Two Decades of Proportionality Analysis under the Canadian Charter ’s Section 1 ” (2006), 34 S.C.L.R. (2d) 501 .

Fuller, Lon L. The Morality of Law , rev. ed. New Haven, Conn.: Yale University Press, 1969.

Gardner, John. “Introduction”, in H. L. A. Hart, Punishment and Responsibility: Essays in the Philosophy of Law , 2nd ed. Oxford: Oxford University Press, 2008, xiii.

Grimm, Dieter. “Proportionality in Canadian and German Constitutional Jurisprudence” (2007), 57 U.T.L.J. 383.

Lyons, David. Ethics and the rule of law . Cambridge: Cambridge University Press, 1984.

Pound, Roscoe. “Mechanical Jurisprudence” (1908), 8 Colum. L. Rev. 605.

Sampford, Charles, with the assistance of Jennie Louise, Sophie Blencowe and Tom Round. Retrospectivity and the Rule of Law . Oxford: Oxford University Press, 2006.

Schauer, Frederick. “Proportionality and the Question of Weight”, in Grant Huscroft, Bradley W. Miller and Grégoire Webber, eds., Proportionality and the Rule of Law: Rights, Justification, Reasoning . New York: Cambridge University Press, 2014, 173.

Stewart, Hamish. “Punitive in Effect: Reflections on Canada v. Whaling ” (2015), 71 S.C.L.R. (2d) 263 .

Stuart, Don. Charter Justice in Canadian Criminal Law , 6th ed. Toronto: Carswell, 2014.

Sullivan, Ruth. Sullivan on the Construction of Statutes , 6th ed. Markham, Ont.: LexisNexis, 2014.

Webber, Grégoire C. N. The Negotiable Constitution: On the Limitation of Rights . Cambridge: Cambridge University Press, 2009.

Zion, Mark. “Effecting Balance: Oakes Analysis Restaged” (2012‑2013), 43 Ottawa L. Rev. 431.

APPEAL from a judgment of the British Columbia Court of Appeal (Newbury, Kirkpatrick and Groberman JJ.A.), 2014 BCCA 382 , 316 C.C.C. (3d) 540, 14 C.R. (7th) 30, 321 C.R.R. (2d) 75, 362 B.C.A.C. 86, 622 W.A.C. 86, [2014] B.C.J. No. 2495 (QL), 2014 CarswellBC 2955 (WL Can.), setting aside in part a sentencing decision. Appeal allowed in part, Abella and Brown JJ. dissenting in part.

Eric Purtzki and Garth Barriere , for the appellant.

Lesley A. Ruzicka , for the respondent.

Richard Kramer and Marc Ribeiro , for the intervener the Attorney General of Canada.

Stacey D. Young and Jennifer A. Crawford , for the intervener the Attorney General of Ontario.

Nicholas St‑Jacques and Lida Sara Nouraie , for the intervener Association des avocats de la défense de Montréal.

John Norris and Cheryl Milne , for the intervener the David Asper Centre for Constitutional Rights.

Matthew R. Gourlay , for the intervener the Criminal Lawyers’ Association (Ontario).

Emily MacKinnon and Michael A. Feder , for the intervener the British Columbia Civil Liberties Association.

The judgment of McLachlin C.J. and Cromwell, Moldaver, Karakatsanis, Wagner, Gascon and Côté JJ. was delivered by

Karakatsanis J. —

I. Introduction

[ 1 ] People’s conduct and the legal consequences that flow from it should be judged on the basis of the law in force at the time. This is a basic tenet of our legal system.

[ 2 ] In recognition of this principle, s. 11( i ) of the Canadian Charter of Rights and Freedoms provides that, if the punishment for an offence is varied after a person commits the offence, but before sentencing, the person is entitled to “the benefit of the lesser punishment”. Like the other legal rights enshrined in s. 11 of the Charter , s. 11( i ) is fundamentally important to our justice system because it protects the fairness of criminal proceedings and safeguards the rule of law.

[ 3 ] When offenders are convicted of certain sexual offences against a person under the age of 16 years, s. 161(1) of the Criminal Code , R.S.C. 1985, c. C-46, gives sentencing judges the discretion to prohibit them from engaging in a variety of everyday conduct upon their release into the community, subject to any conditions or exemptions the judge considers appropriate. In 2012, Parliament expanded the scope of s. 161(1) , empowering sentencing judges to prohibit sexual offenders from having any contact with a person under 16 years of age ( s. 161(1) (c)) or from using the Internet or other digital network ( s. 161(1) (d)).

[ 4 ] In doing so, Parliament intended to give sentencing judges the discretion to impose the expanded prohibition measures on all offenders, even those who offended before the amendments came into force. In other words, Parliament intended the 2012 amendments to operate retrospectively.

[ 5 ] The issue in this appeal is whether the retrospective operation of the 2012 amendments to s. 161(1) (c) and (d) of the Criminal Code is constitutional. This issue engages two subsidiary questions. First, do the prohibition measures contained in s. 161(1) (c) and (d) constitute “punishment” such that their retrospective operation limits s. 11( i ) of the Charter ? Second, if so, is the limit a reasonable one as can be demonstrably justified under s. 1 of the Charter ? The application of these expanded prohibition measures to offenders who committed their offences after the amendments came into force is not at issue.

[ 6 ] I conclude that the 2012 amendments to s. 161(1)(c) and (d) qualify as punishment based on both the objective and impact of the prohibitions. The retrospective imposition of these prohibitions therefore limits s. 11( i ) of the Charter .

[ 7 ] Turning to s. 1 of the Charter , I reach opposite conclusions with respect to s. 161(1)(c) and (d): while the retrospective operation of the no contact provision in s. 161(1)(c) is not a reasonable limit on the s. 11( i ) right, the retrospective operation of the Internet prohibition in s. 161(1) (d) is a reasonable limit. My conclusion with respect to s. 161(1) (d) is chiefly due to the fact that Parliament enacted the provision within a rapidly evolving social and technological context, which changed both the degree and nature of the risk of sexual violence facing young persons. Accordingly, I would allow the appeal in part.

II. Facts and Legislative History

[ 8 ] On March 6, 2013, the appellant pleaded guilty to incest and the creation of child pornography. The offences were committed between 2008 and 2011, and involved the appellant’s preschool-aged daughter.

[ 9 ] When the appellant committed the offences, s. 161(1) of the Criminal Code read as follows:

161. (1) When an offender is convicted, or is discharged on the conditions prescribed in a probation order under section 730, of an offence referred to in subsection (1.1) in respect of a person who is under the age of 16 years, the court that sentences the offender or directs that the accused be discharged, as the case may be, in addition to any other punishment that may be imposed for that offence or any other condition prescribed in the order of discharge, shall consider making and may make, subject to the conditions or exemptions that the court directs, an order prohibiting the offender from

( a ) attending a public park or public swimming area where persons under the age of 16 years are present or can reasonably be expected to be present, or a daycare centre, schoolground, playground or community centre;

( b ) seeking, obtaining or continuing any employment, whether or not the employment is remunerated, or becoming or being a volunteer in a capacity, that involves being in a position of trust or authority towards persons under the age of 16 years; or

( c ) using a computer system within the meaning of subsection 342.1(2) for the purpose of communicating with a person under the age of 16 years.

[ 10 ] After the appellant committed the offences, but before he was sentenced, s. 161(1) was amended by the Safe Streets and Communities Act , S.C. 2012, c. 1, s. 16(1), which came into force on August 9, 2012. Section 161(1) (a) and (b) remained unchanged. But the Act modified s. 161(1) (c) to include prohibiting all contact with young persons, no matter the means, and introduced a new Internet prohibition through s. 161(1) (d). These amendments had the effect of expanding the scope of the community supervision measures a sentencing judge can impose on sexual offenders. Section 161(1) (c) and (d) now provide that a sentencing judge can prohibit an offender from:

(c) having any contact — including communicating by any means — with a person who is under the age of 16 years, unless the offender does so under the supervision of a person whom the court considers appropriate; or

(d) using the Internet or other digital network, unless the offender does so in accordance with conditions set by the court.

[ 11 ] After the 2012 amendments came into force, the appellant was sentenced to nine years’ imprisonment. By virtue of the appellant’s convictions and the age of the victim, the sentencing judge was required to consider whether to impose a prohibition order under s. 161(1) . The question arose as to whether the 2012 amendments could operate retrospectively such that they could be imposed on the appellant.

III. Decisions Below

A. British Columbia Provincial Court — Klinger Prov. Ct. J.

[ 12 ] The sentencing judge found that an order under s. 161 would be appropriate because “there is a serious risk to the safety of children under the age of 16 after [the appellant] is released”. However, on the basis of the test for punishment set out by this Court in R. v. Rodgers , 2006 SCC 15 , [2006] 1 S.C.R. 554, at para. 63 , he concluded that an order under the new s. 161(1) (c) and (d) constitutes punishment within the meaning of s. 11( i ) of the Charter , such that the provisions cannot be applied retrospectively. Since no formal constitutional challenge was brought and the sentencing judge merely used s. 11 ( i ) as a tool of statutory interpretation, no consideration was given to s. 1 of the Charter .

[ 13 ] In the result, the sentencing judge imposed a prohibition order under s. 161 for a period of seven years, but limited the prohibited activities to those described in the version of s. 161(1) that existed when the appellant committed the offences.

B. British Columbia Court of Appeal — 2014 BCCA 382 , 316 C.C.C. (3d) 540

[ 14 ] On the Crown appeal, the appellant filed a formal constitutional challenge to the retrospective operation of the 2012 amendments. The Court of Appeal split over whether a violation of s. 11 ( i ) had been established. Writing for the majority, Newbury J.A. concluded that the 2012 amendments were enacted to protect the public, rather than to punish offenders; therefore, they do not qualify as punishment within the meaning of s. 11 ( i ). Newbury J.A. allowed the appeal and imposed the conditions in s. 161(1) (c) and (d) retrospectively on the appellant for a period of seven years.

[ 15 ] Groberman J.A., dissenting in part, concluded that the retrospective application of the 2012 amendments infringes s. 11 ( i ). Applying Rodgers , Groberman J.A. concluded that s. 161 orders are consequences of conviction, imposed in furtherance of the purpose and principles of sentencing, and thus qualify as “punishment”.

[ 16 ] Because the majority found that s. 11 ( i ) was not engaged, the parties and the Court of Appeal did not address s. 1 of the Charter .

IV. Issues

[ 17 ] This case raises two constitutional questions:

(1) Does the retrospective operation of s. 161 (c) and (d) of the Criminal Code limit s. 11( i ) of the Charter ?

(2) If so, is the limitation a reasonable one prescribed by law as can be demonstrably justified in a free and democratic society under s. 1 of the Charter ?

V. Analysis

[ 18 ] As a preliminary matter, I observe that although there is a presumption against the retrospective application of legislation that affects substantive rights ( R. v. Dineley , 2012 SCC 58 , [2012] 3 S.C.R. 272, at para. 10 ), the parties do not dispute the Court of Appeal’s finding that the presumption has been rebutted in this case because Parliament intended the 2012 amendments to operate retrospectively. I agree.

[ 19 ] This appeal thus turns on whether such retrospective application complies with constitutional standards.

A. Do the 2012 Amendments Constitute Punishment Such That Their Retrospective Operation Limits Section 11(i) of the Charter ?

(1) The Purpose of Section 11( i ) of the Charter and the Interests It Protects

[ 20 ] Section 11 of the Charter protects the legal rights of accused persons when they are charged with an offence. Section 11 encompasses “crucial fundamental rights” ( R. v. Wigglesworth , 1987 41 (SCC) , [1987] 2 S.C.R. 541, per Wilson J., at p. 558), including the right to be tried within a reasonable time ( s. 11 ( b )); the right to be presumed innocent ( s. 11 ( d )); and the right against double jeopardy or punishment ( s. 11 ( h )).

[ 21 ] Section 11( i ) is another such right:

11. Any person charged with an offence has the right

( i ) if found guilty of the offence and if the punishment for the offence has been varied between the time of commission and the time of sentencing, to the benefit of the lesser punishment.

[ 22 ] Along with s. 11 ( g ) — which protects an accused’s right “not to be found guilty on account of any act or omission unless, at the time of the act or omission, it constituted an offence” — s. 11 ( i ) constitutionally enshrines the fundamental notion that criminal laws should generally not operate retrospectively.

[ 23 ] This constitutional aversion to retrospective criminal laws is in part motivated by the desire to safeguard the rule of law. As Lord Diplock put it, “acceptance of the rule of law as a constitutional principle requires that a citizen, before committing himself to any course of action, should be able to know in advance what are the legal consequences that will flow from it” ( Black-Clawson International Ltd. v. Papierwerke Waldhof-Aschaffenburg A.G. , [1975] A.C. 591 (H.L.), at p. 638) . One author expressed the rule of law implications of retrospective laws in these terms:

According to the ideal of the rule of law, the law must be such that those subject to it can reliably be guided by it, either to avoid violating it or to build the legal consequences of having violated it into their thinking about what future actions may be open to them. People must be able to find out what the law is and to factor it into their practical deliberations. The law must avoid taking people by surprise, ambushing them, putting them into conflict with its requirements in such a way as to defeat their expectations and frustrate their plans.

(J. Gardner, “Introduction”, in H. L. A. Hart, Punishment and Responsibility: Essays in the Philosophy of Law (2nd ed. 2008), xiii, at p. xxxvi)

[ 24 ] Retrospective laws threaten the rule of law in another way, by undercutting the integrity of laws currently in effect, “since it puts them under the threat of retrospective change” (L. L. Fuller, The Morality of Law (rev. ed. 1969), at p. 39).

[ 25 ] Relatedly, retrospective laws implicate fairness. “It is unfair to establish rules, invite people to rely on them, then change them in mid-stream, especially if the change results in negative consequences” (R. Sullivan, Sullivan on the Construction of Statutes (6th ed. 2014), at p. 754). For example, an accused who declines to consider a plea and is prepared to take the risk of going to trial should not be subsequently ambushed by an increase in the minimum or maximum penalty for the offence. A retrospective law such as this could not only cause unfairness in specific cases, but could also undermine public confidence in the criminal justice system. Instead, fairness in criminal punishment requires rules that are clear and certain. As McLachlin J. wrote in R. v. Kelly , 1992 62 (SCC) , [1992] 2 S.C.R. 170:

It is a fundamental proposition of the criminal law that the law be certain and definitive. This is essential, given the fact that what is at stake is the potential deprivation of a person of his or her liberty and his or her subjection to the sanction and opprobrium of criminal conviction. This principle has been enshrined in the common law for centuries, encapsulated in the maxim nullum crimen sine lege, nulla poena sine lege — there must be no crime or punishment except in accordance with law which is fixed and certain. [p. 203]

[ 26 ] Clearly, the concerns with retrospective laws are particularly potent in proceedings that are criminal, quasi -criminal, or in which a “true penal consequence” is at stake — the context to which s. 11 applies ( Wigglesworth , at p. 559).

[ 27 ] In sum, s. 11 ( i ) is rooted in values fundamental to our legal system, including respect for the rule of law and ensuring fairness in criminal proceedings.

(2) The Framework for Defining Punishment in Section 11( i ) of the Charter

[ 28 ] In Rodgers , this Court developed a two-part test for determining whether a consequence amounts to “punishment” under s. 11 ( i ): (1) the measure must be a consequence of a conviction that “forms part of the arsenal of sanctions to which an accused may be liable in respect of a particular offence”; and (2) it must be “imposed in furtherance of the purpose and principles of sentencing” (para. 63).

[ 29 ] In the course of articulating this test, Charron J. observed that a “liberal and purposive approach” must be taken to defining punishment (para. 61), but also cautioned that “punishment” does not “encompas[s] every potential consequence of being convicted of a criminal offence” (para. 63). For example, if a consequence advances a legitimate non-punitive state interest, such as solving future crimes, it will likely not constitute punishment, even if it indirectly furthers a sentencing objective like deterrence ( Rodgers , at para. 64 ). Applying this test, Charron J. concluded that post-conviction DNA databank orders do not constitute punishment because they are imposed to assist in the investigation of future crimes, not in furtherance of the purpose and principles of sentencing. The fact that a DNA profile may deter offenders is merely a “residual benefit” (para. 64, quoting R. v. Murrins , 2002 NSCA 12 , 201 N.S.R. (2d) 288 (C.A.), at para. 102 ).

[ 30 ] While the first branch of the s. 11 ( i ) test for punishment (consequence of conviction) has proven to be relatively straightforward, the second branch (imposed in furtherance of the purpose and principles of sentencing) has given rise to two key ambiguities. First, do laws that are primarily aimed at protecting the public necessarily fail to satisfy the second branch of the Rodgers test? Second, what role does the impact a sanction can have on an offender play in the analysis? I address each question in turn.

(a) Do Laws Primarily Aimed at Public Protection Necessarily Fail to Satisfy the Second Branch of the Rodgers Test?

[ 31 ] In this case, the Court of Appeal interpreted Rodgers as indicating that sanctions principally aimed at public protection necessarily fall outside the ambit of punishment. The Crown echoes this position before this Court. As I will explain, this position overreaches: while not all measures imposed to protect the public constitute punishment, public protection is at the core of the purpose and principles of sentencing. Public protection is therefore an insufficient litmus test for defining punishment.

[ 32 ] The purpose and principles of sentencing have been the subject of extensive jurisprudence and are reflected, at least in part, in ss. 718 et seq. of the Criminal Code : see R. v. Lacasse , 2015 SCC 64 , [2015] 3 S.C.R. 1089, at para. 1 ; see also R. v. Ipeelee , 2012 SCC 13 , [2012] 1 S.C.R. 433, at para. 35 . Section 718 provides that the “fundamental purpose of sentencing is to protect society ” and to contribute “to respect for the law and the maintenance of a just, peaceful and safe society ”. This overarching purpose is accomplished by “imposing just sanctions” ( s. 718 ) that reflect one or more of the traditional sentencing objectives: denunciation, deterrence, separation of offenders from society, rehabilitation, reparation, and promoting a sense of responsibility in offenders. Sections 718.1 and 718.2 go on to list a number of sentencing principles, including the fundamental principle of proportionality, that guide sentencing judges in crafting a fit sentence.

[ 33 ] It is clear from the plain language of s. 718 that public protection is part of the very essence of the purpose and principles governing the sentencing process, a point emphasized by this Court in R. v. Lyons , 1987 25 (SCC) , [1987] 2 S.C.R. 309, per La Forest J., at p. 329: “. . . the fundamental purpose of the criminal law generally, and of sentencing in particular, [is] the protection of society”. It is therefore difficult to distinguish between sanctions intended to protect the public and sanctions intended to punish offenders. Doherty J.A. highlighted this difficulty in the recent case of R. v. Hooyer , 2016 ONCA 44 , 129 O.R. (3d) 81. Although his comments were made in the context of defining the common law presumption against retrospectivity, they are apposite here:

The distinction between sanctions intended to protect the public and those intended to punish offenders is difficult to make in the context of sentencing for criminal offences. Many criminal sanctions are designed to both protect the public and punish the accused. In fact, some sanctions protect the public by punishing the accused. The objectives of public protection and punishment often cannot realistically be separated and treated as individual and competing purposes in the sentencing context. [para. 42]

For these reasons, sanctions intended to advance public safety do not constitute a broad exception to the protection s. 11 ( i ) affords and may qualify as punishment.

[ 34 ] To be clear, while measures imposed at sentencing for the purpose of protecting the public may constitute punishment under s. 11 ( i ), a public-protection purpose is not, on its own, determinative. To satisfy the second branch of the Rodgers test, a consequence of conviction must be imposed in furtherance of the purpose and principles of sentencing. As discussed, the purpose of sentencing is to “protect society” or advance “respect for the law and the maintenance of a just, peaceful and safe society” ( s. 718 of the Criminal Code ) by fulfilling one or more of the traditional sentencing objectives ( s. 718 (a) through (f)) in accordance with the principles of sentencing reflected in ss. 718.1 and 718.2 .

(b) What Role Does the Impact of a Sanction Play in the Analysis?

[ 35 ] Citing R. v. Cross , 2006 NSCA 30 , 138 C.R.R. (2d) 163, at paras. 45-46 , the Crown submits that the impact of a sanction on an offender is only relevant if it is out of proportion to the sanction’s legislative purpose. That is, “if the impact of the sanction aligns with its legislative purpose and is not of such magnitude that it reveals, instead, a punitive intent, it is not ‘punishment’” ( Cross , at para. 45 ).

[ 36 ] As I shall explain, I conclude that the impact of a sanction has broader significance. While a sanction’s impact was to some extent implicit in the Rodgers analysis, in my view, the s. 11 ( i ) test for punishment must embody a clearer, more meaningful consideration of the impact a sanction can have on an offender. This is important for a variety of reasons.

[ 37 ] First, it accords with “the liberal and purposive approach” that must be taken in interpreting Charter rights, including s. 11( i ) ( Rodgers , at para. 61 ). The purposes of s. 11 ( i ), which are centred on the rule of law and fairness in criminal proceedings, are compromised if the right is incapable of protecting offenders from the retrospective imposition of sanctions that have a significant impact on their liberty or security — regardless of the sanction’s objective. As the interveners the David Asper Centre for Constitutional Rights, the Criminal Lawyers’ Association (Ontario), the British Columbia Civil Liberties Association, and the Association des avocats de la défense de Montréal all submit, fairness and predictability in punishment are enhanced when there is a pragmatic consideration of the impact of an impugned sanction.

[ 38 ] A “liberal and purposive approach” to punishment is appropriate because s. 11( i ) is engaged only within a narrow sphere. As mentioned, in Wigglesworth , this Court held that s. 11 of the Charter applies only to proceedings that are criminal or quasi -criminal, or, regardless of the nature of the proceeding, if a “true penal consequence” such as imprisonment is at stake (p. 559). The Court in Wigglesworth gave s. 11 a narrow ambit so that “[t]he content of [the s. 11 ] rights [does not] suffer from a lack of predictability or a lack of clarity because of a universal application of the section” (p. 558). Although the “true penal consequence” test sets an indisputably high bar, it was developed to determine whether a person is nonetheless “charged with an offence” even if he or she is the subject of proceedings outside the criminal context. Within the criminal law context, the concerns motivating a narrow construction of “penal consequences” or “punishment” largely fall away.

[ 39 ] Second, a consideration of the impact of a sanction is consistent with this Court’s jurisprudence. Since the early days of the Charter , this Court has always looked to both purposes and effects when considering the constitutionality of laws: see R. v. Big M Drug Mart Ltd. , 1985 69 (SCC) , [1985] 1 S.C.R. 295, at p. 331. And in the recent decision of Canada (Attorney General) v. Whaling , 2014 SCC 20 , [2014] 1 S.C.R. 392, this Court adopted “a functional rather than a formalistic perspective” (para. 52), observing that, “[i]t is the retrospective frustration of an expectation of liberty that constitutes punishment” (para. 60). The Court went on to conclude that the elimination of accelerated parole review violated s. 11 ( h ) as it had a sufficiently significant impact on “an offender’s settled expectation of liberty” (para. 60). In doing so, the Court focused on the impact the retrospective law had on the offender, rather than the purpose animating the law: see H. Stewart, “Punitive in Effect: Reflections on Canada v. Whaling ” (2015), 71 S.C.L.R. (2d) 263, at p. 269 . Although Whaling was concerned with the definition of punishment in the context of s. 11( h ) of the Charter , harmony between s. 11( i ) and ( h ) is desirable as fairness in punishment underlies both provisions.

[ 40 ] Third, an approach that accounts for a sanction’s impact will assist in identifying the “lesser punishment” to which an accused is entitled. The punishment with the less severe impact on the liberty or security of an offender will be deemed to be the “lesser punishment” for the purposes of s. 11 ( i ). A definition of punishment that focuses heavily on the objective of the sanction obscures this inquiry.

[ 41 ] Thus, I would restate the test for punishment as follows in order to carve out a clearer and more meaningful role for the consideration of the impact of a sanction: a measure constitutes punishment if (1) it is a consequence of conviction that forms part of the arsenal of sanctions to which an accused may be liable in respect of a particular offence, and either (2) it is imposed in furtherance of the purpose and principles of sentencing, or (3) it has a significant impact on an offender’s liberty or security interests. [1]

[ 42 ] As this Court wrote in Cunningham v. Canada , 1993 139 (SCC) , [1993] 2 S.C.R. 143: “The Charter does not protect against insignificant or ‘trivial’ limitations of rights . . . . The [state action] must be significant enough to warrant constitutional protection” (p. 151). That is why, if a consequence of conviction is not imposed in furtherance of the purpose and principles of sentencing, it must have a significant impact on an offender’s constitutionally protected liberty or security interests before it will qualify as punishment for the purposes of s. 11 ( i ). To satisfy this requirement, a consequence of conviction must significantly constrain a person’s ability to engage in otherwise lawful conduct or impose significant burdens not imposed on other members of the public. Again, Doherty J.A.’s comments in Hooyer are helpful: “. . . a prohibition that significantly limits the lawful activities in which an accused can engage, where an accused can go, or with whom an accused can communicate or associate, would sufficiently impair the liberty and security of the accused to warrant characterizing the prohibition as punishment” (para. 45).

[ 43 ] Having reformulated the s. 11 ( i ) test for punishment, I now turn to the sanctions at issue in this appeal. I first discuss s. 161 of the Criminal Code in more detail before applying the test for punishment to the 2012 amendments.

(3) History and Operation of Section 161 of the Criminal Code

[ 44 ] The legislative history, judicial interpretation, and design of s. 161 all confirm that the section has an overarching protective function: to shield children from sexual violence.

[ 45 ] Section 161 was enacted in 1993 in response to the decision in R. v. Heywood (1992), 1992 6008 (BC CA) , 20 B.C.A.C. 166, in which the British Columbia Court of Appeal struck down under s. 7 of the Charter the offence of loitering: see An Act to amend the Criminal Code and the Young Offenders Act , S.C. 1993, c. 45, s. 1. After 1993, s. 161 continued to evolve and, in 2012, the impugned amendments were introduced through the Safe Streets and Communities Act . The protective function of s. 161 generally, and the 2012 amendments specifically, was repeatedly emphasized throughout the legislative debates. For example, at the Bill’s third reading, the Minister of Justice stated that the proposed amendments are “an important step forward in the protection of children in this country” ( House of Commons Debates , vol. 145, No. 144, 3rd Sess., 40th Parl., March 11, 2011, at p. 8967).

[ 46 ] The jurisprudence interpreting and applying s. 161 confirms the provision’s protective purpose: see, e.g., R. v. Heywood , 1994 34 (SCC) , [1994] 3 S.C.R. 761, at p. 803; R. v. A. (R.K.) , 2006 ABCA 82 , 208 C.C.C. (3d) 74, at para. 20 ; R. v. Perron , 2009 ONCA 498 , 244 C.C.C. (3d) 369, at para. 13 .

[ 47 ] As well, the design of s. 161 is consistent with its purpose of protecting children from sexual violence. Section 161 orders are discretionary and “subject to the conditions or exemptions that the court directs” (s. 161(1)). They can therefore be carefully tailored to the circumstances of a particular offender. The discretionary and flexible nature of s. 161 demonstrates that it was designed to empower courts to craft tailored orders to address the nature and degree of risk that a sexual offender poses to children once released into the community. Failure to comply with the order can lead to a term of imprisonment of up to four years (s. 161(4)).

[ 48 ] Further, I agree with the line of cases holding that s. 161 orders can be imposed only when there is an evidentiary basis upon which to conclude that the particular offender poses a risk to children and the judge is satisfied that the specific terms of the order are a reasonable attempt to minimize the risk: see A. (R.K.) , at para. 32 ; see also R. v. R.R.B. , 2013 BCCA 224 , 338 B.C.A.C. 106, at paras. 32-34 . These orders are not available as a matter of course. In addition, the content of the order must carefully respond to an offender’s specific circumstances. [2]

(4) Application of the Test for Punishment to the 2012 Amendments to Section 161 of the Criminal Code

[ 49 ] Applying the reformulated test, I conclude that the 2012 amendments constitute punishment.

[ 50 ] First, the 2012 amendments form part of the arsenal of sanctions to which an accused may be liable in respect of a particular offence. Section 161(1) directs sentencing judges to consider whether to exercise their discretion to impose the community supervision measures once an offender is convicted of an enumerated sexual offence involving a person under the age of 16. Section 161 orders are therefore a consequence of conviction, a fact that the Crown does not dispute.

[ 51 ] Second, the sanctions contained in the 2012 amendments are imposed in furtherance of the purpose and principles of sentencing and can have a significant impact on an offender’s Charter -protected interests — although, to be clear, both are not required to satisfy the test.

[ 52 ] As to the objective, the 2012 amendments are intended to protect children by separating offenders from society, assisting in rehabilitation, and deterring sexual violence, sentencing goals that all find expression in s. 718 of the Criminal Code . In addition, the discretionary and flexible process through which s. 161 orders are imposed aligns with the principles of sentencing articulated in ss. 718.1 and 718.2. As noted above, the fact that such orders are imposed to protect children, on its own, is not determinative.

[ 53 ] These prohibitions are to be distinguished from DNA orders, which have been found not to constitute punishment under s. 11 ( i ): see Rodgers , at para. 65 . As discussed, the objective of DNA orders is primarily to facilitate the investigation of future crimes, rather than to achieve deterrence, denunciation, separation, or rehabilitation in connection with a past offence: see Rodgers , at para. 64 .

[ 54 ] Turning to the impact of the amendments, both s. 161(1) (c) and (d) can have a significant impact on the liberty and security of offenders — potentially for the rest of their lives. This Court has recognized that living in the community under restrictions can attract a considerable degree of stigma ( R. v. Proulx , 2000 SCC 5 , [2000] 1 S.C.R. 61, at para. 105 ). Further, a prohibition under s. 161(1) (c) on having any contact with persons under the age of 16 could potentially curtail the types of employment an offender can pursue, and an offender’s ability to interact with people (including adults in the company of children) in public and private spaces. And depriving an offender under s. 161(1)(d) of access to the Internet is tantamount to severing that person from an increasingly indispensable component of everyday life:

The Internet has become a hub for every kind of human activity, from education to recreation to commerce. It is no longer merely a window to the world. For a growing number of people, the Internet is their world — a place where one can do nearly everything one needs or wants to do. The Web provides virtual opportunities for people to shop, meet new people, converse with friends and family, transact business, network and find jobs, bank, read the newspaper, watch movies, and attend classes. [Emphasis in original; footnotes omitted.]

(B. A. Areheart and M. A. Stein, “Integrating the Internet” (2015), 83 Geo . Wash. L. Rev. 449, at p. 456)

For many Canadians, membership in online communities is an integral component of citizenship and personhood. In my view, retrospectively excluding offenders from these virtual communal spaces is a substantial consequence that implicates the fairness and rule of law concerns underlying the s. 11 ( i ) right.

[ 55 ] The significant impact the 2012 amendments can have on the liberty and security of offenders is another way in which these sanctions are distinguishable from DNA orders. I agree with Doherty J.A. that “a sentencing provision requiring an accused to provide a DNA sample upon conviction . . . does not meaningfully impair the accused’s liberty or security of the person and would not be regarded as punishment” ( Hooyer , at para. 45 ).

[ 56 ] I also note that the text of s. 161(1) (“in addition to any other punishment” or “ en plus de toute autre peine ”), while certainly not determinative, is nonetheless informative. As Groberman J.A. observed in dissent at the Court of Appeal, “Parliament itself appears to have considered that the sanctions set out in s. 161(1) come within the ordinary meaning of the word ‘punishment’” (para. 78) or “ peine ”. [3]

[ 57 ] In sum, the prohibitions found in the 2012 amendments to s. 161(1) constitute punishment for the purposes of s. 11( i ) of the Charter . They are a consequence of conviction, imposed in furtherance of the purpose and principles of sentencing, and they can have a significant impact on the liberty and security of offenders. Clearly, the 2012 amendments constitute greater punishment than the previous prohibitions: under the new s. 161(1) (c), a judge can prohibit all contact with children, no matter the means (not just contact involving a computer system); and under the new s. 161(1) (d), a judge can prohibit an offender from using the Internet or other digital network for any purpose (not just for the purpose of contacting children). Accordingly, the retrospective operation of these provisions limits the s. 11 ( i ) right as it deprives the appellant of the benefit of the less restrictive community supervision measures captured in the previous version of s. 161 — that is, the “lesser punishment”.

B. Is the Limitation of Section 11(i) Justified Under Section 1 of the Charter ?

[ 58 ] Section 1 of the Charter provides as follows:

1. The Canadian Charter of Rights and Freedoms guarantees the rights and freedoms set out in it subject only to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society.

To establish that the limitation on the appellant’s s. 11 ( i ) right is reasonable and demonstrably justified, the government must show that the 2012 amendments have a sufficiently important objective “and that the means chosen are proportional to that object[ive]” ( Carter v. Canada (Attorney General) , 2015 SCC 5 , [2015] 1 S.C.R. 331, at para. 94 ). A law is proportionate if (1) there is a rational connection between the means adopted and the objective; (2) it is minimally impairing in that there are no alternative means that may achieve the same objective with a lesser degree of rights limitation; and (3) there is proportionality between the deleterious and salutary effects of the law ( R. v. Oakes , 1986 46 (SCC) , [1986] 1 S.C.R. 103; Carter , at para. 94 ). The proportionality inquiry is a normative and contextual one, which requires courts to examine the broader picture by “balanc[ing] the interests of society with those of individuals and groups” ( Oakes , at p. 139).

[ 59 ] Unfortunately, s. 1 was not dealt with in the courts below. This means we do not have the benefit of a full record, including expert testimony. But the parties urged us to consider s. 1 on the record before us. This Court therefore deals with this issue, on consent, as a court of first instance.

[ 60 ] The Crown adduced fresh evidence attached to two affidavits, consisting of statistics and social science articles relating to the issue of the recidivism of sexual offenders. The appellant did not oppose the admission of this evidence and I am satisfied it would be appropriate to receive it. Accordingly, in assessing whether the Crown has discharged its justificatory burden, I will consider the Crown’s fresh evidence as “supplemented by common sense and inferential reasoning”, in addition to the jurisprudence and legislative debates proffered by the parties ( R. v. Sharpe , 2001 SCC 2 , [2001] 1 S.C.R. 45, at para. 78 ).

(1) Do the 2012 Amendments Have a Sufficiently Important Objective?

[ 61 ] A law that limits a constitutional right must do so in pursuit of a sufficiently important objective that is consistent with the values of a free and democratic society. This examination is a threshold requirement that is undertaken without considering the scope of the right infringement, the means employed, or the relationship between the positive and negative effects of the law.

[ 62 ] The appellant correctly submits that the relevant objective is that of the infringing measure: see Toronto Star Newspapers Ltd. v. Canada , 2010 SCC 21 , [2010] 1 S.C.R. 721, at para. 20 . Here, the infringing measure is the retrospective operation of the impugned law. However, the more general purpose behind the enactment of the 2012 amendments informs the specific rationale for applying the amendments retrospectively.

[ 63 ] The appellant argues that the objective of the retrospective operation of the 2012 amendments is to increase the punishment imposed on offenders who committed their offences prior to 2012 so as to more effectively further the purpose and principles of sentencing. In my view, this articulation of the law’s purpose is not sufficiently precise and is essentially a description of the means the legislature has chosen to achieve its purpose: see Carter , at para. 76 ; see also R. v. Moriarity , 2015 SCC 55 , [2015] 3 S.C.R. 485, at para. 28 .

[ 64 ] As discussed above, the legislative history, judicial interpretation, and design of s. 161 all confirm that the overarching goal of the section is to protect children from sexual violence perpetrated by recidivists. And there is ample evidence in the legislative record surrounding the enactment of the new s. 161(1) (c) and (d) to show that enhancing child protection motivated the impugned amendments as well. To highlight but one example, at the debate accompanying the second reading of the Bill, the Parliamentary Secretary to the Minister of Justice said the amendments “see[k] to prevent . . . child sex offenders from having the opportunity to facilitate their offending. Finding access to a child or the opportunity to be alone with a child is a key for many child sex offenders” ( House of Commons Debates , vol. 145, No. 110, 3rd Sess., 40th Parl., December 3, 2010, at p. 6787).

[ 65 ] Accordingly, the overarching objective of the prospective operation of the 2012 amendments is to enhance the protection s. 161 affords to children against the risk of harm posed by convicted sexual offenders. It follows naturally that the objective of the retrospective operation of these amendments — the infringing measure — is to better protect children from the risks posed by offenders like the appellant who committed their offences before, but were sentenced after, the amendments came into force. This latter objective anchors the s. 1 analysis.

[ 66 ] Obviously, this objective is sufficiently important to warrant further scrutiny. As Laskin J.A. wrote in R. v. Budreo (2000), 2000 5628 (ON CA) , 46 O.R. (3d) 481 (C.A.), “Children are among the most vulnerable groups in our society. The sexual abuse of young children is a serious societal problem, a statement that needs no elaboration” (para. 37). Providing enhanced protection to children from becoming victims of sexual offences is vital in a free and democratic society.

(2) Are the Means Adopted Proportional to the Law’s Objective?

[ 67 ] In assessing the proportionality of a law, a degree of deference is required. As this Court recently wrote in Carter :

At this stage of the analysis, the courts must accord the legislature a measure of deference. Proportionality does not require perfection: Saskatchewan (Human Rights Commission) v. Whatcott , 2013 SCC 11 , [2013] 1 S.C.R. 467, at para. 78 . Section 1 only requires that the limits be “reasonable”. [para. 97]

(a) Rational Connection

[ 68 ] At this first step of the proportionality inquiry, the government must demonstrate that the means used by the limiting law are rationally connected to the purpose the law was designed to achieve. “To establish a rational connection, the government need only show that there is a causal connection between the infringement and the benefit sought ‘on the basis of reason or logic’” ( Carter , at para. 99 , citing RJR-MacDonald Inc. v. Canada (Attorney General) , 1995 64 (SCC) , [1995] 3 S.C.R. 199, at para. 153 ).

[ 69 ] As the appellant concedes, there is clearly a rational connection between providing enhanced protection to children from the risks of sexual violence presented by offenders who committed their offences before the 2012 amendments came into force (the objective) and retrospectively giving sentencing judges the discretionary power to limit those offenders who pose a continuing risk to children in contacting children in person or online, and in engaging with online child pornography (the means chosen). Although the Crown’s fresh evidence, which I discuss below, assists in solidifying this causal link, at this stage, I am satisfied that reason and logic suffice to establish that Parliament proceeded rationally in opting to give s. 161(1) (c) and (d) retrospective effect in order to better protect children from recidivism risks posed by offenders who committed their offences before the 2012 amendments came into force.

(b) Minimal Impairment

[ 70 ] The question at this second stage is whether the 2012 amendments are minimally impairing, in the sense that “the limit on the right is reasonably tailored to the objective” ( Carter , at para. 102 ). It is only when there are alternative, less harmful means of achieving the government’s objective “in a real and substantial manner” that a law should fail the minimal impairment test ( Alberta v. Hutterian Brethren of Wilson Colony , 2009 SCC 37 , [2009] 2 S.C.R. 567, at para. 55 ).

[ 71 ] I am satisfied that the retrospective operation of the prohibitions contained in the 2012 amendments is minimally impairing of s. 11( i ).

[ 72 ] The amendments were enacted within the context of a highly discretionary provision that is tailored to its objective. Prohibitions listed in s. 161(1) are to be imposed only when a judge is satisfied that the specific offender poses a continued risk to children upon his release into the community and that the specific terms of the order are a reasonable attempt to minimize the risk. The law is therefore not “drafted in a way that unnecessarily catches [conduct] that has little or nothing to do with the prevention of harm to children” ( Sharpe , at para. 95 ). In other words, the retrospective use of s. 161(1) (c) and (d) is available only when a judge is satisfied that the prohibitions will advance the enhanced child-protection goal of the amendments. No risk, no retrospective order.

[ 73 ] Further, s. 161(1) permits a sentencing judge to impose any conditions or exemptions that correspond to the circumstances of a particular offender. Section 161(1)(c) provides that offenders may have contact with persons under the age of 16 if “the offender does so under the supervision of a person whom the court considers appropriate”. Similarly, s. 161(1)(d) permits offenders to use the Internet if “the offender does so in accordance with conditions set by the court”. Finally, the prohibition order can be limited in duration (s. 161(2)) and reviewed periodically to ensure it continues to correspond to an offender’s circumstances (s. 161(3)).

[ 74 ] Despite the highly discretionary and tailored nature of s. 161 , the appellant argues that the impugned amendments are not minimally impairing because the Crown has failed to demonstrate that a purely prospective application of the amendments would undermine its objective. [4] Although I will discuss the potential gaps in the evidentiary record more fully below when I weigh the deleterious and salutary effects of the law, I would not give effect to this submission at the minimal impairment stage, for a few reasons.

[ 75 ] It is widely accepted (and the record confirms) that a non-trivial percentage of sex offenders will reoffend. If the amendments operated only prospectively, a sentencing judge would be unable to impose the prohibitions in s. 161(1)(c) and (d) on offenders who committed their crimes before 2012 even if the judge were satisfied that the prohibitions were required to minimize the risk to a child that a sex offender will recidivate. I therefore accept that a purely prospective application of the amendments would have prevented Parliament from fully realizing its objective of enhancing the protection s. 161 affords to children from offenders who committed their offences before the coming into force of the 2012 amendments. Further, accepting the appellant’s argument would fail to accord sufficient deference, at this stage of the analysis, to the government’s choice of legislative means. And questions pertaining to the extent of the efficacy of the retrospective operation of the 2012 amendments are best left to the next step of the analysis: proportionality of effects.

[ 76 ] In sum, given the discretionary and tailored nature of s. 161 and the fact that a purely prospective operation of the amendments would have compromised Parliament’s full objective, I conclude that the retrospective operation of s. 161(1)(c) and (d) impairs the s. 11 ( i ) right as little as reasonably possible. [5] The more difficult issue is whether the benefits achieved from imposing the 2012 amendments retrospectively outweigh the deleterious effects.

(c) Proportionality of Effects

[ 77 ] At this final stage of the proportionality analysis, the Court must “weig[h] the impact of the law on protected rights against the beneficial effect of the law in terms of the greater public good” ( Carter , at para. 122 ). [6] This final stage is an important one because it performs a fundamentally distinct role. As a majority of this Court observed in Thomson Newspapers Co. v. Canada (Attorney General) , 1998 829 (SCC) , [1998] 1 S.C.R. 877:

The focus of the first and second steps of the proportionality analysis is not the relationship between the measures and the Charter right in question, but rather the relationship between the ends of the legislation and the means employed. . . . The third stage of the proportionality analysis provides an opportunity to assess, in light of the practical and contextual details which are elucidated in the first and second stages, whether the benefits which accrue from the limitation are proportional to its deleterious effects as measured by the values underlying the Charter . [para. 125]

[ 78 ] It is for this reason that Aharon Barak, former President of the Supreme Court of Israel, has described this final step as “the very heart of proportionality” (“Proportional Effect: The Israeli Experience” (2007), 57 U.T.L.J. 369, at p. 380). And in Hutterian Brethren , Abella J. wrote: “. . . most of the heavy conceptual lifting and balancing ought to be done at the final step — proportionality. Proportionality is, after all, what s. 1 is about” (para. 149).

[ 79 ] I agree. While the minimal impairment test has come to dominate much of the s. 1 discourse in Canada, this final step permits courts to address the essence of the proportionality enquiry at the heart of s. 1. [7] It is only at this final stage that courts can transcend the law’s purpose and engage in a robust examination of the law’s impact on Canada’s free and democratic society “in direct and explicit terms” (J. Cameron, “The Past, Present, and Future of Expressive Freedom Under the Charter ” (1997), 35 Osgoode Hall L.J. 1, at p. 66). In other words, this final step allows courts to stand back to determine on a normative basis whether a rights infringement is justified in a free and democratic society. Although this examination entails difficult value judgments, it is preferable to make these judgments explicit, as doing so enhances the transparency and intelligibility of the ultimate decision. Further, as mentioned, proceeding to this final stage permits appropriate deference to Parliament’s choice of means, as well as its full legislative objective.

[ 80 ] In this case, there are important differences between the effects of the two impugned amendments. I will therefore consider the two provisions separately.

(i) Balancing the Deleterious and Salutary Effects of the Retrospective Operation of Section 161(1)(c) of the Criminal Code

[ 81 ] The deleterious effects flowing from the retrospective operation of s. 161(1)(c) are substantial. At the individual level, in depriving offenders of the benefit of the lesser punishment, s. 161(1)(c) prevents the appellant and other offenders from freely participating in society following their release into the community. Before the new s. 161(1)(c) was introduced, outside the digital realm, judges could prohibit offenders only from attending public parks, public swimming pools, daycare centres, schoolgrounds, playgrounds, and community centres, or from seeking employment or volunteer opportunities involving children. The new s. 161(1)(c) potentially goes much further and prohibits “ any contact — including communicating by any means — with a person who is under the age of 16 years” in a public or private space. For example, offenders might be prohibited from conversing with younger members of their family, or from freely moving about certain private and public spaces where children are present. This expanded prohibition, relative to the more limited prohibitions that existed previously, constitutes a substantial intrusion on the liberty and security of certain offenders.

[ 82 ] The deleterious effects experienced by specific offenders translate into broader societal harms. By impacting people like the appellant with a punishment of which they had no notice, the retrospective operation of s. 161(1)(c) undermines fairness in criminal proceedings and compromises the rule of law. These are core tenets of our justice system.

[ 83 ] The adverse impact the retrospective operation of s. 161(1)(c) has on fairness and the rule of law is particularly acute because, in broadening the scope of prohibited conduct, Parliament does not appear to have been responding to an emerging threat, or an evolving social context. Unfortunately, sexual offences against children have persisted for centuries. Setting aside for the moment the use of technology to contact young people, which is captured by s. 161(1)(d), why was additional protection required in 2012? In terms of sexual offences resulting from physical proximity, on this record, there appears to have been little change in the nature and degree of risk facing children since the last time s. 161(1) was amended. The dearth of a compelling temporal justification for imposing s. 161(1)(c) retrospectively enhances the damage the provision does to fairness and the rule of law, and thus undermines public confidence in the criminal justice system.

[ 84 ] The Crown submits that the benefit of retrospectively applying s. 161(1)(c) is that more children will be protected from sexual violence. In advancing this claim, the Crown chiefly relies on social science articles and statistics relating to recidivism of sexual offenders in order to clarify the risk children face when sexual offenders are released into the community.

[ 85 ] The Crown’s social science articles endeavour to quantify rates of recidivism of sexual offenders. One article pegged the recidivism rates for “child molesters” at 13% 5 years following the commission of the offence, 18% after 10 years, and 23% after 15 years. [8] The authors found that the recidivism rate for sexual offenders who victimize extra familial young boys (35% after 15 years) is significantly higher than the average recidivism rate for all sexual offenders (24% after 15 years) (p. 8). These recidivism rates were confirmed by another article adduced by the Crown, which asserts that “[s]exual interest in children was a significant predictor of sexual recidivism”. [9] That is, “[t]hose individuals with identifiable interests in deviant sexual activities were among those most likely to continue sexual offending. The evidence was strongest for sexual interest in children” (p. 15). The authors further observed that these figures “should be considered to underestimate the real recidivism rates” because sexual crimes are significantly underreported (p. 8).

[ 86 ] These recidivism rates are significant. I accept that a non-trivial number of sexual offenders commit further sexual crimes after being released into the community. And the odds of this occurring appear to increase in the context of sexual offences against children. This is the harm the 2012 amendments are aimed at mitigating.

[ 87 ] The Crown also seeks to demonstrate the beneficial effects of making these enhanced prohibitions available retrospectively through statistics relating to the number of offenders potentially impacted by the 2012 amendments. Since the amendments came into force and as of May 14, 2015, 157 s. 161 orders have been imposed in British Columbia on offenders who committed their offences prior to August 9, 2012. And as of that same date there were 239 accused persons in British Columbia charged with offences captured by s. 161 that were committed prior to the coming into force of the 2012 amendments. On a national scale, these numbers would clearly be much higher. These statistics suggest that if the 2012 amendments cannot operate retrospectively, sentencing judges will be unable to consider imposing the enhanced prohibitions found in s. 161(1)(c) and (d) on many hundreds of sex offenders across the nation.

[ 88 ] I accept that the Crown’s fresh evidence assists in identifying recidivism rates and the number of offenders who stand to be impacted by the retrospective operation of the 2012 amendments. Real risks to children are certainly present. And I accept that a provision prohibiting contact between sexual offenders and children will, to some extent, assist in mitigating these risks.

[ 89 ] However, the appellant correctly points out that the Crown has failed to lead much, if any, evidence to establish the degree of enhanced protection s. 161(1)(c) provides in comparison to the previous version of the prohibition. It is therefore unclear what effect the retrospective operation of s. 161(1)(c) would have on the recidivism rates identified by the Crown. And there is no evidence demonstrating that the risks s. 161(1)(c) are directed at have changed quantitatively or qualitatively, such that the fundamental fairness and rule of law concerns would be mitigated. Even in the passages of the legislative record that the Crown put before this Court, it is striking that there was almost no discussion of why the amendments to s. 161(1)(c) were required to better protect children.

[ 90 ] Put simply, the precise benefits of the retrospective operation of s. 161(1)(c) remain unclear. It can be difficult to prove a negative, which is why reason and logic are important complements to tangible evidence. And, to some extent, these evidentiary difficulties may be unavoidable. After all:

Public policy is often based on approximations and extrapolations from the available evidence, inferences from comparative data, and, on occasion, even educated guesses. Absent a large-scale policy experiment, this is all the evidence that is likely to be available. Justice La Forest offered an observation in McKinney which rings true: “[d]ecisions on such matters must inevitably be the product of a mix of conjecture, fragmentary knowledge, general experience and knowledge of the needs, aspirations and resources of society”.

(S. Choudhry, “So What Is the Real Legacy of Oakes ? Two Decades of Proportionality Analysis under the Canadian Charter ’s Section 1” (2006), 34 S.C.L.R. (2d) 501, at p. 524, quoting McKinney v. University of Guelph , 1990 60 (SCC) , [1990] 3 S.C.R. 229, at p. 304.)

[ 91 ] Nonetheless, s. 1 mandates that the limitation on the right be demonstrably justified. As Dickson C.J. wrote in Oakes , this is a “stringent standard of justification” (p. 136). The retrospective operation of the impugned measure adversely impacts the liberty and security of offenders (relative to the previous version of s. 161 ), and, importantly, the fairness of criminal proceedings and the rule of law. Although this adverse impact will be experienced only when a judge concludes it is necessary to alleviate the risk the offender poses to children, it remains the case that the deleterious effects of the impugned measure are significant and tangible.

[ 92 ] In comparison, the benefits society stands to gain are marginal and speculative. While the Crown’s evidence regarding recidivism of sexual offenders begins to paint the picture (particularly since it shows that sex offenders who victimize children are more likely to reoffend), the rendering remains largely incomplete. In particular, the Crown has provided no temporal justification for the retrospective limitation, nor much evidence to establish the degree of enhanced protection s. 161(1)(c) provides. For example, the record suggests that many sexual assaults committed against children are perpetrated by family members or acquaintances. But surely this reality did not just recently come to Parliament’s attention. In the context of a s. 11 ( i ) infringement, one expects the Crown to better explain why retrospective penal laws were required.

[ 93 ] Temporal considerations are relevant in this content because, at its root, s. 11 ( i ) is about the timing of changes to penal laws. In this case, it is not Parliament’s decision to increase the punishment for sexual offenders that has, by itself, triggered Charter scrutiny — rather, it is Parliament’s decision to reach back in time to impose these enhanced prohibitions on offenders who had no notice of them that offends s. 11( i ) . Thus, temporal factors that may help explain Parliament’s rationale for circumventing a basic tenet of our criminal law are relevant to the s. 11 ( i ) inquiry. When it comes to s. 11 ( i ), timing can be everything.

[ 94 ] Evidence related to the risks of recidivism is generally insufficient, on its own, to discharge the Crown’s justificatory burden. To hold otherwise would be to potentially eviscerate the s. 11 ( i ) right for the simple reason that retrospectively increasing punishment in order to curtail the risk of recidivism is a rationale that could apply to a broad range of crimes.

[ 95 ] It may be tempting to conclude that mitigating the risk of sexual violence to even one child is worth the costs. However, there can be no broad exception to the protection of s. 11 ( i ) whenever the victim is a child. Such an approach ascribes almost no value to the right. Section 11 ( i ) protects fundamental interests that can be overridden only in demonstrably compelling circumstances. In my view, the Crown has failed to show that the largely speculative salutary effects of the retrospective operation of s. 161(1)(c) outweigh its tangible and substantial drawbacks.

[ 96 ] The retrospective operation of s. 161(1)(c) therefore cannot be justified under s. 1. As a result, s. 161(1)(c) applies only prospectively — that is, only to offenders who committed their offences after the 2012 amendments came into force ( s. 52(1) , Constitution Act, 1982 ).

[ 97 ] I note that there are other prohibition orders under the Criminal Code that may assist the Crown to some extent in filling the gap left by the lack of any retrospective application of s. 161(1)(c), such as those that can be imposed pursuant to ss. 810, 810.1, and 810.2. However, I make no further comment on those provisions since they were not meaningfully raised or argued by any of the parties before us.

(ii) Balancing the Deleterious and Salutary Effects of the Retrospective Operation of Section 161(1)(d) of the Criminal Code

[ 98 ] The deleterious effects resulting from the retrospective operation of s. 161(1)(d) are also significant. A complete ban on “using the Internet or other digital network” — an indispensable tool of modern life and an avenue of democratic participation — is more intrusive than the previous ban on “using a computer system . . . for the purpose of communicating” with young people. This constitutes a significant deprivation of liberty. Therefore, the retrospective operation of s. 161(1)(d) can erect massive barriers to an offender’s full participation in society, which may result in substantial consequences both socially and economically.

[ 99 ] As with the retrospective operation of s. 161(1)(c), the imposition of punishment without notice translates into broader societal harms, including compromising the fairness of criminal proceedings and challenging the rule of law. Clarity and predictability are central to the proper functioning of the criminal justice system, and are at the core of s. 11 ( i )’s purpose. Respect for the law and public confidence in the administration of justice are threatened when laws are changed retrospectively, without notice.

[ 100 ] Turning to the salutary effects, the Crown’s evidence relating to the risk of harm from recidivism of sexual offenders, discussed above, applies equally here; however, when it comes to s. 161(1)(d), this evidence is buttressed by other important considerations.

[ 101 ] As I shall explain, in brief, the record before this Court demonstrates that s. 161(1)(d) is directed at grave, emerging harms precipitated by a rapidly evolving social and technological context. This evolving context has changed both the degree and nature of the risk of sexual violence facing young persons. As a result, the previous iteration of s. 161 became insufficient to respond to the modern risks children face. By closing this legislative gap and mitigating these new risks, the benefits of the retrospective operation of s. 161(1)(d) are significant and fairly concrete.

[ 102 ] The rate of technological change over the past decade has fundamentally altered the social context in which sexual crimes can occur. Social media websites (like Facebook and Twitter), dating applications (like Tinder), and photo-sharing services (like Instagram and Snapchat) were all founded after 2002, the last time prior to the 2012 amendments that substantial revisions to s. 161(1) were made. These new online services have given young people — who are often early adopters of new technologies — unprecedented access to digital communities. At the same time, sexual offenders have been given unprecedented access to potential victims and avenues to facilitate sexual offending.

[ 103 ] The legislative record before this Court speaks to this rapid evolution and shows that, in enacting s. 161(1)(d) and giving it retrospective effect, Parliament was attempting to keep pace with technological changes that have substantially altered the degree and nature of the risks facing children. For example, at the second reading of the Bill, the Parliamentary Secretary to the Minister of Justice said, “An increasing number of child sex offenders also use the Internet and other new technologies to facilitate the grooming of victims or to commit other child sex offences” (p. 6787). At a Committee debate, the Acting General Counsel, Criminal Law Policy Section, Department of Justice testified:

. . . what Bill C-54 recognizes is that offenders use the Internet computer systems for all sorts of reasons. Yes, they use it to communicate directly with a young person, and we catch that already, but they use it also to offend, in their offending pattern, whether it’s to access child pornography, for example . . . .

So the idea with Bill C-54 is to require a court to turn its mind to this each time it is sentencing a person who is convicted of one of these child sex offences and to consider whether in that instance, with the offender before them, given the nature of the offending pattern and the conduct before the court, there should be a restriction on that individual’s access to the Internet or other technology that would otherwise facilitate his or her reoffending.

(Standing Committee on Justice and Human Rights, Evidence , No. 50, 3rd Sess., 40th Parl., February 28, 2011, at p. 4)

[ 104 ] As well, a Statistics Canada Director (who was testifying before the Committee) said, “What we can say based on those data is that the number of charges of child luring via the Internet is increasing” ( Evidence , No. 49, 3rd Sess., 40th Parl., February 16, 2011, at p. 7). The legislative record contains other similar passages.

[ 105 ] In addition to this testimony concerning the evolving risks children face, others testified that controlling an offender’s access to the Internet is an effective means of curbing these risks. For example, during other Committee debates, the Executive Director of BOOST Child Abuse Prevention and Intervention testified that “[t]he emerging research connecting online offences to hands-on sexual offences emphasizes the importance of the court’s ability . . . to permit the offender use of the Internet only when supervised” ( Evidence , No. 46, 3rd Sess., 40th Parl., February 7, 2011, at p. 6). [10]

[ 106 ] The Crown’s social science literature also addresses the unique role the Internet plays in facilitating sexual crimes against children. For example:

The number of detected online sex offenders has drastically increased since the early 2000s . . . .

. . . Indeed, the rates of online sexual crimes, and child pornography offences in particular, have increased substantially with the increasing use of the internet . . . .

. . . Specifically, the ease of access to online child pornography may contribute to a new group of offenders who succumb to temptations that they would have otherwise controlled.

(K. M. Babchishin, R. K. Hanson and H. VanZuylen, “Online Child Pornography Offenders are Different: A Meta-analysis of the Characteristics of Online and Offline Sex Offenders Against Children” (2015), 44 Arch. Sex. Behav. 45, at p. 46)

[ 107 ] New and qualitatively different opportunities to harm young people exist. The Internet is a portal to accessing and distributing child pornography, a crime that itself victimizes children. As this Court observed in Sharpe :

. . . possession of child pornography contributes to the market for child pornography, a market which in turn drives production involving the exploitation of children. Possession of child pornography may facilitate the seduction and grooming of victims and may break down inhibitions or incite potential offences. [para. 28]

Further, the Internet can be used to contact other adults for the purposes of planning and facilitating criminal behaviour — pursuits not captured by the previous version of s. 161 . [11]

[ 108 ] What emerges from the Crown’s materials is that the proliferation of new technologies has altered the nature and degree of risk facing children, which, in turn, created a legislative gap in s. 161 . The previous iteration of s. 161 — which allowed sentencing judges to prohibit offenders only from using computer systems to contact children directly — was incapable of precluding sexual offenders from participating in other kinds of harmful behaviour. And, as the record and common sense suggest, monitoring an offender’s use of the Internet can limit an offender’s opportunities to offend and prevent this harmful behaviour.

[ 109 ] This unique social and technological context leads me to the conclusion that the benefits occasioned by retrospectively imposing the Internet prohibition contained in s. 161(1)(d) are greater and more certain than those stemming from s. 161(1)(c).

[ 110 ] The fact that Parliament enacted s. 161(1)(d) as a means of closing a legislative gap created by rapid social and technological change does not just enhance the salutary effects of the law: it mitigates the provision’s deleterious effects, too. From the perspective of public confidence in the criminal justice system, the retrospective operation of a law that was enacted to respond to a swiftly changing social context and emerging threats seems less unfair and less inconsistent with the rule of law than the retrospective operation of a law that was not enacted for a compelling temporal reason. As Professor C. Sampford writes in his book, Retrospectivity and the Rule of Law (2006), “Retrospective laws which close ‘loopholes’ and ‘unexpected interpretations and consequences’ reinforce the guidance of primary laws” and can therefore advance the fairness of the legal system as a whole (p. 81).

[ 111 ] Thus, while fairness and the rule of law are compromised by laws that retrospectively undermine a citizen’s liberty and security, these broader societal harms are mitigated by Parliament’s compelling temporal justification for giving s. 161(1)(d) retrospective effect.

[ 112 ] I now must balance the deleterious and salutary effects of the law. As discussed, s. 161(1)(d) constitutes a significant impact on an offender’s liberty and security. The impugned measure also has negative ramifications for society as a whole. Fairness and the rule of law are compromised by laws that retrospectively undermine a citizen’s liberty and security, although these broader societal harms are less acute given the context in which the government legislated. In addition, the adverse impact the provision has on offenders will be experienced only when there is good reason: in circumstances where a judge finds that doing so will mitigate the risk an offender poses to children.

[ 113 ] As for the salutary effects, the record demonstrates that the Internet is increasingly being used to sexually offend against young people and that sex offenders who target children are more likely to reoffend. This is not simply about changing technology or general risks associated with recidivism, broad factors that can relate to many offences. Rather, the nature and degree of the risks facing some of the most vulnerable members of our society have changed drastically since 2002, the last time s. 161(1) was substantially amended. Technology and the proliferation of social media cyber communities have increased the degree of risk facing young persons. This has created new triggers, and new avenues for offenders to pursue in committing further offences. The previous prohibition was insufficient to address these evolving risks. But the enhanced prohibition in s. 161(1)(d) can restrict the viability of these routes. While it remains difficult to quantify the precise benefits the retrospective operation of s. 161(1)(d) may create, it seems to me that the salutary effects associated with s. 161(1)(d) are quite tangible and compelling.

[ 114 ] On balance, in my view, Parliament was justified in giving s. 161(1)(d) retrospective effect in the unique context within which it was legislating. A variety of factors support this conclusion. The harms at stake (sexual offending against young people) are particularly powerful. The statutory regime is highly tailored and discretionary. An Internet prohibition, while invasive, is not among the most onerous punishments, such as increased incarceration. And, significantly, the rapidly evolving technological and social context surrounding the enactment of s. 161(1)(d) has created new and emerging risks that make the law’s salutary effects more concrete — while mitigating the adverse impact the law has on fairness and the rule of law. Although any one of these factors may have been insufficient in isolation, taken together, they create a compelling case. The benefits of the law outweigh its deleterious effects.

VI. Disposition

[ 115 ] I find that the retrospective operation of s. 161(1) (c) of the Criminal Code limits the right protected by s. 11( i ) of the Charter and that this limit is not justified under s. 1. Accordingly, I would allow the appeal with respect to s. 161(1) (c). As a result, the provision does not apply retrospectively to offenders who committed their offences prior to the coming into force of the 2012 amendments.

[ 116 ] I also find that the retrospective operation of s. 161(1) (d) of the Criminal Code limits the s. 11 ( i ) right. However, I conclude that this is a reasonable constitutional compromise under s. 1. I would therefore dismiss the appeal with respect to s. 161(1) (d).

The following are the reasons delivered by

[ 117 ] A BELLA J. (dissenting in part) — I agree with Justice Karakatsanis that both ss. 161(1) (c) and 161(1)(d) of the Criminal Code , R.S.C. 1985, c. C-46, violate s. 11( i ) of the Canadian Charter of Rights and Freedoms and that s. 161(1)(c) cannot be justified under s. 1. With great respect, however, I do not share the view that s. 161(1) (d) is justified.

[ 118 ] From 2008 to 2011, when K.R.J. committed the offences for which he was eventually convicted, s. 161(1) of the Criminal Code stated:

161. (1) When an offender is convicted . . . of an offence referred to in subsection (1.1) in respect of a person who is under the age of 16 years, the court that sentences the offender . . . in addition to any other punishment that may be imposed for that offence . . . shall consider making and may make, subject to the conditions or exemptions that the court directs, an order prohibiting the offender from

( a ) attending a public park or public swimming area where persons under the age of 16 years are present or can reasonably be expected to be present, or a daycare centre, schoolground, playground or community centre;

( b ) seeking, obtaining or continuing any employment, whether or not the employment is remunerated, or becoming or being a volunteer in a capacity, that involves being in a position of trust or authority towards persons under the age of 16 years; or

( c ) using a computer system within the meaning of subsection 342.1(2) for the purpose of communicating with a person under the age of 16 years.

[ 119 ] Under this scheme, K.R.J. could be subjected to geographic, work-related, and “virtual” restrictions. He could be prohibited from attending a wide variety of venues such as pools and schools, and from using a computer for the purpose of communicating with anyone under 16 years of age . He would still, however, have been entitled to engage in online activities with adults.

[ 120 ] By the time K.R.J. was sentenced, Parliament amended the provision. While s. 161(1)(a) and (b) were left unchanged, s. 161(1)(c) was amended and s. 161(1)(d) was added, giving sentencing judges authority to prohibit offenders from:

(c) having any contact — including communicating by any means — with a person who is under the age of 16 years, unless the offender does so under the supervision of a person whom the court considers appropriate; or

(d) using the Internet or other digital network, unless the offender does so in accordance with conditions set by the court.

[ 121 ] The amendments expanded the restrictions K.R.J. could be placed under. Rather than being banned from certain venues, s. 161(1)(c) could be used to prohibit him from attending any place where children are present. And rather than being prohibited from using the internet for the purpose of communicating with children , s. 161(1)(d) could be used to prohibit him from using the internet for any purpose.

[ 122 ] I agree with the majority that these potential restrictions would significantly affect K.R.J.’s liberty and security interests, and would, as a result, constitute punishment under s. 11( i ) of the Charter , which states:

11. Any person charged with an offence has the right

( i ) if found guilty of the offence and if the punishment for the offence has been varied between the time of commission and the time of sentencing, to the benefit of the lesser punishment.

[ 123 ] The wording in this provision is unequivocal. As noted by Prof. Don Stuart, the intention behind this text is “crystal clear”: Charter Justice in Canadian Criminal Law (6th ed. 2014), at p. 523.

[ 124 ] In my view, the absolutist language used by the drafters of the Charter in s. 11 must colour the s. 1 analysis by demanding the most stringent of justifications. That was the approach taken by this Court in Canada (Attorney General) v. Whaling , 2014 SCC 20 () , [2014] 1 S.C.R. 392. The issue was the retrospective repeal of the accelerated parole review under s. 11( h ) of the Charter , which protects individuals from being punished twice for the same offence. Because the Crown had failed to adduce “compelling evidence” demonstrating that its objectives would be “significantly undermined” unless the repeal was applied on a retrospective as well as prospective basis, this Court concluded that the infringement was not justified under s. 1.

[ 125 ] The repeal of the accelerated parole review was subsequently also found to be unconstitutional by the British Columbia Court of Appeal, but from the perspective of s. 11 ( i ), the provision at issue in this appeal. In Liang v. Canada (Attorney General) (2014), 2014 BCCA 190 () , 311 C.C.C. (3d) 159, the British Columbia Court of Appeal concluded that the Crown’s concern that it could take years to phase out the program if it could not be applied retrospectively, did not justify overriding the right:

. . . the Charter specifically requires that if punishment has changed between offence commission and sentencing, the offender is entitled to the lesser punishment. . . . [T]he fact the offender will receive a lesser punishment, and perhaps one that does not meet the objectives of the present sentencing regime, is exactly what s. 11(i) contemplates. . . .

. . . to meet the burden under s. 1 in this case, something more must be asserted than that the objective of the increased punishment is important, and therefore those who are constitutionally entitled to the lesser punishment must forego their rights. [Emphasis added; paras. 59 and 61.]

[ 126 ] Both Whaling and Liang are clear that s. 11 imposes a singularly onerous evidentiary burden on the Crown to justify a violation under s. 1. To apply a lesser burden transforms s. 11 ( i ) from being practically an air-tight right into a porous one. In this case, that means that the Crown has the highest possible evidentiary burden, namely, to demonstrate through “compelling evidence” that the previous provisions so “significantly undermined” the government’s objectives, that the retrospective application of greater punishment was justified.

[ 127 ] As the majority notes, the Crown’s evidentiary record consisted largely of statistics about s. 161(1) orders in British Columbia, and studies on recidivism rates pertaining to sexual offenders in general, including two that suggested a link between recidivism and online activities. The Crown also argued that the language shift from “computer system” to “Internet and digital network” in s. 161(1)(d) was designed to reflect advancements in technology. I agree with the majority that this evidence is insufficient to justify s. 161(1)(c) because “the Crown has failed to lead much, if any, evidence to establish the degree of enhanced protection . . . in comparison to the previous version of the prohibition” such that “the precise benefits of the retrospective operation of s. 161(1)(c) remain unclear”: paras. 89-90 (emphasis in original).

[ 128 ] But unlike my colleagues, I find that this same reasoning is fatal to s. 161(1)(d). Far from offering compelling evidence, the Crown offered no evidence in the context of s. 161(1)(d) to show that the former provisions so significantly undermined its objectives, that the retroactive application of greater restrictions was justified. If all that is needed to justify a breach of s. 11 ( i ) is the suggestion of a possible reduction in recidivism rates, whether based on changes in technology or otherwise, the state could, in theory, justify the retrospective application of more stringent punishments so routinely that s. 11( i ) is written out of the Charter .

[ 129 ] In fact, there was no evidence about how the retrospective application of s. 161(1)(d) was expected to, or would, reduce recidivism rates any more than those under the former s. 161(1)(c) “computer” restrictions. I see no reason to bridge the significant empirical gaps in the evidence with inferences, particularly in the context of s. 11.

[ 130 ] I would therefore allow the appeal in connection with both ss. 161(1)(c) and 161(1)(d).

The following are the reasons delivered by

Brown J. (dissenting in part) —

I. Introduction

[ 131 ] As my colleague Karakatsanis J. aptly notes for the majority, sexual offences against children have “persisted for centuries” (para. 83). Their legacy is toxic. They are notorious for their devastating impact, often ruining the lives of their victims, and of those whose lives intersect with those victims as they move into adulthood. Trauma from childhood sexual abuse may reverberate for generations, creating pernicious cycles of abuse.

[ 132 ] My colleague recounts how, in response to this persistent grave misconduct and its consequent social harms, Parliament amended s. 161(1) of the Criminal Code , R.S.C. 1985, c. C-46, in 2012, augmenting the conditions which a sentencing judge may, in his or her discretion, impose upon an offender convicted of designated sexual offences, where the sentencing judge considers such conditions appropriate to prevent the offender from committing sexual offences against children in the future. Specifically, the sentencing judge’s discretion was expanded from prohibiting offenders from “using a computer system . . . for the purpose of communicating with a person under the age of 16 years” to the following:

161 (1) . . .

(c) having any contact — including communicating by any means — with a person who is under the age of 16 years, unless the offender does so under the supervision of a person whom the court considers appropriate; or

(d) using the Internet or other digital network, unless the offender does so in accordance with conditions set by the court.

Significantly, these amendments apply to all offenders being sentenced for a designated offence, irrespective of when the offender committed that offence.

[ 133 ] I agree with Karakatsanis J. that these conditions constitute “punishment” within the meaning of s. 11( i ) of the Canadian Charter of Rights and Freedoms , and I endorse the test by which she makes that determination. I also agree that their retrospective application infringes s. 11( i ) . My point of departure is at the s. 1 stage of the analysis. Whereas my colleague concludes that the Crown has met its burden of justifying its infringement of s. 11 ( i ) only in respect of the conditions relating to Internet use contained in s. 161(1) (d), in my view the Crown has also done so in respect of the conditions imposable under s. 161(1) (c) relating to contact with children. I would therefore uphold both conditions, dismiss the appeal, and affirm the s. 161 order made by the Court of Appeal.

II. Section 1

[ 134 ] It is worth bearing in mind that s. 11( i ) of the Charter deals with the retrospective application of laws which are punitive in nature. At issue under s. 11 ( i ), then, is not the punishment itself, but rather the means by which it is imposed. In my view, this means-based quality of the s. 11 ( i ) protection affects the analysis to be applied under s. 1, since the Oakes analysis considers the proportionality between a legislative objective and the Charter -infringing effects resulting from its pursuit, not the choice of means that, by itself, constitutes a Charter infringement. The s. 1 analysis should be sensitive to this, in keeping with Dickson C.J.’s direction in Oakes : “. . . the nature of the proportionality test will vary depending on the circumstances” ( R. v. Oakes , 1986 46 (SCC) , [1986] 1 S.C.R. 103, at p. 139). The Oakes test is not, and should not be treated as, a technical inquiry, as it is “dangerously misleading to conceive of s. 1 as a rigid and technical provision”: R. v. Keegstra , 1990 24 (SCC) , [1990] 3 S.C.R. 697, at p. 735, per Dickson C.J. As La Forest J. (dissenting, but not on this point) stated in RJR-MacDonald Inc. v. Canada (Attorney General) , 1995 64 (SCC) , [1995] 3 S.C.R. 199:

In Oakes , this Court established a set of principles, or guidelines, intended to serve as a framework for making this determination. However, these guidelines should not be interpreted as a substitute for s. 1 itself. It is implicit in the wording of s. 1 that the courts must, in every application of that provision, strike a delicate balance between individual rights and community needs. Such a balance cannot be achieved in the abstract, with reference solely to a formalistic “test” uniformly applicable in all circumstances. The s. 1 inquiry is an unavoidably normative inquiry, requiring the courts to take into account both the nature of the infringed right and the specific values and principles upon which the state seeks to justify the infringement. [Emphasis added; para. 62.]

[ 135 ] In other words, a technical and inflexible application of the Oakes test risks reducing what ought to be a rich, contextual inquiry under s. 1 into a form of “mechanical jurisprudence”, where “[c]onceptions are fixed”, “[t]he premises are no longer to be examined”, and “[p]rinciples cease to have importance”: R. Pound, “Mechanical Jurisprudence” (1908), 8 Colum. L. Rev. 605, at p. 612. The moral nuances inherent in the question of justifiable limits on fundamental rights cannot be reduced to “technical questions of weight and balance”: G. C. N. Webber, The Negotiable Constitution: On the Limitation of Rights (2009), at p. 104. Yet, and despite its statements to the contrary, the majority in this case has in my respectful view done precisely that. Its rigid and acontextual application of Oakes and its subsequent jurisprudence causes it to lose sight of the broader context and overall goal sought by Parliament. It reads the purpose of the legislation in an excessively narrow fashion, which results in an application of the Oakes test in a way that is ill-suited to deal with punitive laws which apply retrospectively. It holds Parliament to an exacting standard of proof, thereby denying Parliament the room necessary to perform its legislative policy-development role when addressing a chronic social problem. And it also insists on direct evidence of anticipated benefits which, given that chronic nature of the harm, is likely impossible to obtain.

[ 136 ] The insight of Dickson C.J. and La Forest J. in our jurisprudence is that the s. 1 analysis must account for the broader picture. The issue is not, as La Forest J. put it, whether a particular “formalistic ‘test’” has been satisfied. The “unavoidably normative inquiry” must remain focussed on the broader picture: has the state demonstrated that the impugned law prescribes a reasonable limit, demonstrably justified in a free and democratic society? To be clear, I do not suggest that Oakes is incorrect. Rather, I echo Dickson C.J.’s and La Forest J.’s warnings about its rigid, acontextual application. We should not lose the proportionality forest for the Oakes trees.

A. Objective of the Measure

[ 137 ] The means-based quality of s. 11( i )’s protection should therefore inform the characterization of the objective anchoring the s. 1 proportionality analysis. The majority says that the relevant objective for the purpose of a proportionality analysis is that of the Charter -infringing measure — which, in this case, is the retrospective operation of the amendments to s. 161(1). I agree, but only to a point. The relevant objective for this purpose is indeed the objective of the measure. However, as I will explain, the measure to be considered here comprises the amendments as a whole, and not merely their retrospectivity.

[ 138 ] Considering retrospectivity in isolation from the broader provision of which it forms a part skews the Oakes analysis by making several of its elements largely redundant. If, as the majority says, Parliament’s objective was to “better protect children from the risks posed by offenders like the appellant” (para. 65) — i.e., offenders who committed a designated offence before, but were sentenced after, the amendments came into force and who pose a risk to reoffend sexually against children — then the application of such orders to offenders like the appellant is obviously rationally connected to this objective. And, there would be no possible less-impairing means of achieving this objective: simply put, the only way Parliament can apply the protective aspect of s. 161(1) orders to such offenders retrospectively is to apply s. 161(1) orders to such offenders retrospectively. Indeed, under the majority’s approach, the minimal impairment inquiry becomes otiose. Of course, were such orders to be applied retrospectively as to offenders unlike the appellant (i.e., those who do not pose a risk to reoffend sexually against children), the rational connection and minimal impairment steps would then have some work to do under the Oakes analysis. By narrowly construing Parliament’s purpose as the majority has, however, considerations of the rational connection and minimal impairment elements of the proportionality analysis are limited to determining whether the Charter- infringing measure captures the individuals which it targets, not whether the measure is rationally connected to the objective and minimally impairing of the Charter rights of those who legitimately fall within its ambit.

[ 139 ] A broader examination of Parliament’s purpose is therefore necessary in order to anchor a useful proportionality analysis. The measure that gave rise to the Charter infringement, and which should anchor the proportionality analysis, comprises the amendments to s. 161 as a whole. And, as to that measure, I agree with the majority’s characterization of its objective as being to “enhance the protection s. 161 affords to children against the risk of harm posed by convicted sexual offenders” (para. 65). The retrospective application of these amendments is rationally connected to that protective purpose, since the risk an offender poses to reoffend sexually against children is not affected by whether the offence occurred before or after the measure’s enactment. And, given Parliament’s objective of enhancing the protections that s. 161 affords to children, there is no less-impairing alternate measure that would allow for s. 161(1) ’s protections to be realized in respect of an offender who committed his or her offence before the amendments came into force and who poses a risk to reoffend.

B. Balancing Salutary and Deleterious Effects

[ 140 ] I agree with the majority that the final stage of the s. 1 analysis allows courts to “transcend the law’s purpose and engage in a robust examination of the law’s impact on Canada’s free and democratic society” (para. 79). But a robust examination of this impact takes us only so far because, after all, the impact of a provision on a free and democratic society is hardly a measurable thing. The question we are trying to answer is whether “the deleterious effects are out of proportion to the public good achieved by the infringing measure”: Alberta v. Hutterian Brethren of Wilson Colony , 2009 SCC 37 , [2009] 2 S.C.R. 567, at para. 78 . Neither criterion is amenable to demonstrative proof. The final proportionality analysis is tied to the practical impacts and benefits of the law, but what is ultimately being weighed is much more abstract and philosophical: the detriment to Charter -protected rights against the public benefit sought. We must therefore be careful to avoid insisting upon too strict an evidentiary burden.

[ 141 ] With these general comments in mind, I turn to the majority’s proportionality analysis. It suffers, in my respectful view, from several flaws. First, it imposes an evidentiary burden on the state that is impossible to satisfy, especially in the murky area of recidivism risks and criminal law policy. Second, it overstates the deleterious effects of s. 161(1) (c) while understating its salutary effects. Further, the majority’s reasons for upholding the retrospective application of s. 161(1) (d) are, in principle, equally applicable to the retrospective application of s. 161(1) (c). In other words, if the majority’s reasoning on s. 161(1) (d) is accepted, then the retrospective application of s. 161(1) (c) must also be a proportionate limit on the appellant’s s. 11 ( i ) right.

(1) The Evidentiary Burden

[ 142 ] The majority stresses — almost to a determinative extent — shortcomings it sees in the Crown’s social science evidence, concluding that while it sufficiently demonstrates that the sought-after “degree of enhanced protection” for children will be achieved by the retrospective operation of s. 161(1) (d), “the rendering remains largely incomplete” in respect of s. 161(1) (c) (para. 92).

[ 143 ] This reasoning is troubling in several respects. First, it departs significantly from this Court’s approach to social science evidence and the evidentiary burden borne by the state under s. 1. Social science evidence used to establish legislative facts should ordinarily be adduced through expert witnesses in order to allow its truth to be tested: Public School Boards’ Assn. of Alberta v. Alberta (Attorney General) , 2000 SCC 2 , [2000] 1 S.C.R. 44, at paras. 4-5 , per Binnie J. This social science evidence, however, was adduced through a “Brandeis brief”, and is untested by the ordinary truth-seeking processes of a trial. Considerable care should therefore be taken in examining this evidence and drawing inferences — whether favourable or adverse from the state’s standpoint — from it: M. v. H. , 1999 686 (SCC) , [1999] 2 S.C.R. 3, at para. 296 , per Bastarache J., writing separate but concurring reasons.

[ 144 ] Further, given the complex social context in which Parliament often develops policy — of which the prevention of recidivism in cases of sexual offences against children is clearly an instance — it will sometimes be difficult, if not impossible, for the state to provide reliable and direct evidence of the benefit its measures will achieve. Recidivism rates are derived from statistical extrapolation, psychology, and other elements of social science, which will not always translate easily into proof to the standard of demonstrable justification. As this Court has recognized, “social claims are not always amenable to proof by empirical evidence”: Mounted Police Association of Ontario v. Canada (Attorney General) , 2015 SCC 1 , [2015] 1 S.C.R. 3, at para. 144 . As a result, “public policy is often made on the basis of incomplete knowledge”: S. Choudhry, “So What Is the Real Legacy of Oakes ? Two Decades of Proportionality Analysis under the Canadian Charter ’s Section 1” (2006), 34 S.C.L.R. (2d) 501, at p. 524 . The proportionality analysis should therefore be sensitive to policy-makers’ need for a measure of latitude to consider and try previously untried alternatives, particularly when confronting persistent and complex public policy concerns.

[ 145 ] This is not to say that these evidentiary difficulties compel acceptance of the Crown’s claims. This Court has held that a rigorous s. 1 analysis may also be accomplished by employing “logic [and] reason” in assessing justifiable limits on Charter rights: Harper v. Canada (Attorney General) , 2004 SCC 33 , [2004] 1 S.C.R. 827, at para. 78 ; see also R. v. Butler , 1992 124 (SCC) , [1992] 1 S.C.R. 452, at p. 503-4, per Sopinka J.; Keegstra , at p. 776, per Dickson C.J.; Thomson Newspapers Co. v. Canada (Attorney General) , 1998 829 (SCC) , [1998] 1 S.C.R. 877, at para. 107 , per Bastarache J.; R. v. Sharpe , 2001 SCC 2 , [2001] 1 S.C.R. 45, at paras. 85-94 , per McLachlin C.J.; R. v. Bryan , 2007 SCC 12 , [2007] 1 S.C.R. 527, at para. 20 , per Bastarache J., and paras. 100-103, per Abella J., dissenting. By applying this approach here (instead of demanding empiricism where none can exist), the salutary effects of s. 161(1) (c) become clear, as does the true scope of its deleterious effects.

(2) Salutary and Deleterious Effects of Section 161(1)(c)

[ 146 ] The majority says that the retrospective operation of s. 161(1)(c) creates serious deleterious effects at an individual and societal level. At an individual level, it views s. 161(1) (c) as going much further in its potential restrictions of an offender’s liberty than did its predecessor, since it “prohibits any contact — including communicating by any means — with a person who is under the age of 16 years” (para. 81 (emphasis in original)). It warns that this provision could have the effect of prohibiting offenders from conversing with younger members of his or her family, or that it could prohibit offenders from “freely moving about certain private and public spaces where children are present” (para. 81). At a societal level, the majority says that the retrospective operation of a punitive law “undermines fairness in criminal proceedings and compromises the rule of law” (para. 82) (although this can, of course, be said of any measure which infringes s. 11 ( i )).

[ 147 ] The general restriction on liberty or security of the person which results from retrospectively applied punishment is not, however, relevant to the inquiry under s. 11( i ) of the Charter . What is relevant when assessing the deleterious impact upon the offender of a retrospectively applied punitive law is the degree by which it increases punishment relative to the original law. For example, a retrospective increase in a mandatory minimum term of incarceration from one year to 14 years would have a greater deleterious impact on offenders and on the rule of law than would a retrospective increase in a fine from $100 to $101. But, again, this is because of the relative differences in the degree of increased punishment wrought by such measures, and not because of the general restrictions on liberty or security of the person that they impose. Again, s. 11( i ) is not concerned with the nature of the punishment, but with its retrospective increase.

[ 148 ] Further, the majority’s conclusion regarding the deleterious impact upon the offender’s liberty interests is, in my view, overstated.

[ 149 ] It is useful to return to the text of s. 161(1) (c):

161 (1) . . . the court that sentences the offender . . . shall consider making and may make, subject to the conditions or exemptions that the court directs, an order prohibiting the offender from

(c) having any contact — including communicating by any means — with a person who is under the age of 16 years, unless the offender does so under the supervision of a person whom the court considers appropriate . . . .

[ 150 ] Section 161(1)(c) contains two crucial qualifications which circumscribe its deleterious impact upon an offender’s liberty interest. First, the matter is left to the sentencing judge’s discretion, both as to whether to impose conditions (“shall consider making and may make”), and as to the tailoring of the conditions themselves (“subject to the conditions or exemptions that the court directs”). Second, a s. 161(1)(c) order — even when imposed without other conditions or exemptions — still contains the internal qualification that the prohibition of contact with a person under the age of 16 years only applies to such contact which occurs without the “supervision of a person whom the court considers appropriate”.

[ 151 ] In other words, an offender who seeks to interact with, for example, younger members of his or her family, may do so either by seeking an exemption or under the supervision of a person the court considers appropriate. Similarly — and assuming that, as the majority suggests, freely moving about in a public space where children are present is sufficient to constitute “contact” or to risk “contact” (a suggestion to which I return below) — were an offender to provide a legitimate reason for being in a public space where children are present, that offender may obtain an exemption for that particular place, or may be in that place under the supervision of a person the court considers appropriate. In determining whether such exemptions are appropriate, the sentencing court must of course consider the danger the offender poses to re-offend sexually against children. But the point is that s. 161(1) (c) gives a sentencing judge the tools to ensure that the offender’s liberty is not restricted more than is necessary to mitigate that offender’s risk.

[ 152 ] As to the meaning of “contact”, the majority’s assessment of the deleterious effects of s. 161(1)(c)’s retrospective application largely rests on an overly expansive interpretation of the meaning of “contact” in s. 161(1)(c). More to the point, the majority’s suggestion that merely “moving about” in a public space where children are present constitutes or risks “contact” represents a strained interpretation of the scope of the restriction on contact, and is directly at odds with the well-established principle that the criminal law’s prohibitions on conduct should be construed strictly: R. v. McIntosh , 1995 124 (SCC) , [1995] 1 S.C.R. 686, at paras. 38-39 , per Lamer C.J. To the extent, therefore, that the meaning of “contact” is ambiguous, it “must be interpreted in the manner most favourable to accused persons”: McIntosh , at para. 39 .

[ 153 ] While overstating the deleterious effects of s. 161(1) (c)’s retrospective operation, the majority also understates its salutary effects. The risk that some offenders pose to reoffend sexually against children simply cannot be mitigated by the original version of s. 161(1) . The appellant presents an example of this. Having committed several designated offences against his infant daughter, he was found by the sentencing judge to pose a “substantial” risk to reoffend sexually against children. While s. 161(1)(a) would have allowed the sentencing judge to restrict the offender’s presence in specified public places such as public parks and public swimming areas in which children are present or could reasonably be expected to be present, the sentencing judge could not tailor a s. 161(1) order to restrict the appellant’s ability to interact with children in private. But this is, of course, precisely where the appellant and other similar offenders pose the greatest risk to children. The evidence before Parliament showed that (1) of the children of the age of five years and less who were the victims of sexual offences in 2009, approximately 60% of boys and 70% of girls were victimized by family members; and (2) most victims under the age of 16 were victimized by family members or acquaintances. Far from “speculative” (para. 95), then, the salutary effects of s. 161(1)(c)’s retrospective operation seem manifest. It restricts an offender whose offences predate the amendments to s. 161(1)(c) from having unsupervised access to children, both in private and in public, where the sentencing judge determines that such a condition is necessary to address a risk that the offender will commit further sexual offences against children.

[ 154 ] The majority’s consideration of the deleterious effects of the retrospective operation of this provision also views as significant the “dearth of a compelling temporal justification” for s. 161(1)(c)’s retrospective operation, in the sense that “there appears to have been little change in the nature and degree of risk facing children since the last time s. 161(1) was amended” (para. 83). But with respect, and even assuming this concern could fairly be characterized as “temporal” in nature, this is not the sort of temporal concern that s. 11( i ) engages, being the retrospective application of punishment. The majority, is, in substance, questioning whether Parliament’s objective — which the majority has already found to have met the “pressing and substantial” objective requirement of Oakes — was pressing and substantial. Further, even if this “temporal justification” were an appropriate consideration at this stage of the analysis, it should not be virtually determinative when assessing the deleterious impact of a retrospective punishment. Bearing in mind that the record indicates that Parliament was responding to what it believed to be a grave social harm — which harm the majority acknowledges as persistent — it is worth recalling this Court’s statement in Keegstra (at p. 776, per Dickson C.J.) that it is “well accepted that Parliament can use the criminal law to prevent the risk of serious harms”. It does not matter whether that risk has remained constant or increased, or whether it is longstanding or emerging. This Court has never, for example, required the Crown to advance a compelling “temporal” justification to uphold Charter -infringing impaired driving legislation by showing that the persistent social harm of impaired driving has taken a turn for the worse: see, e.g., R. v. Orbanski , 2005 SCC 37 , [2005] 2 S.C.R. 3; R. v. St-Onge Lamoureux , 2012 SCC 57 , [2012] 3 S.C.R. 187 (upholding the presumption of identity in s. 258(1) (d.1) of the Criminal Code ). Parliament should be entitled, within constitutional limits, to innovate in finding a solution to chronic harms, irrespective of whether the incidence of such harms has remained stable, increased, or even declined.

[ 155 ] To be clear, nobody doubts that s. 11( i ) deals with temporal considerations, because, as the majority says, it is “about the timing of changes to penal laws” (para. 93 (emphasis in original)). But the “temporal” concern identified by the majority speaks more (if not exclusively) to the pressing and substantial nature of Parliament’s objective than it does to the deleterious effects of retrospective punishment on the rule of law (e.g. para. 93: “. . . temporal factors that may help explain Parliament’s rationale . . .”). All retrospective changes to the law derogate from the rule of law, irrespective of Parliament’s reasons for enacting them. All retrospective punishment is imposed without fair warning, denying a person “the opportunity to know what is expected of her and to decide what to do in light of that knowledge”: D. Lyons, Ethics and the rule of law (1984), at p. 75. In every such case, and even where the majority’s concern about whether there has been “change in the nature and degree of risk” (para. 83) is assuaged, the rule of law is harmed: see L. L. Fuller, The Morality of Law (rev. ed. 1969), at pp. 53-54; C. Sampford, Retrospectivity and the Rule of Law (2006), at p. 81. The relevance of this concern driving the majority’s assessment of the deleterious impacts on the rule of law in this case is therefore far from evident.

(3) Inconsistent Treatment of Paragraphs (c) and (d)

[ 156 ] I also observe that, apart from the matter of “temporal” justifications which I have just addressed, all the reasons identified by the majority in support of its conclusion that the limit imposed on the appellant’s s. 11( i ) right by the retrospective application of s. 161(1)(d) is justified are equally applicable to the retrospective application of s. 161(1)(c).

[ 157 ] In this regard, the majority observes in respect of s. 161(1)(d) that the harms at stake are “particularly powerful”; that the statutory regime “is highly tailored and discretionary”; and that the Internet prohibition is “not among the most onerous punishments, such as increased incarceration” (para. 114). But each of these reasons support the conclusion that the retrospective operation of s. 161(1)(c) is justified as well. Section 161(1)(c) addresses precisely the same “particularly powerful” concern as does s. 161(1)(d), being sexual offences against children. The condition in s. 161(1)(c), as I have explained, is also “highly tailored and discretionary”, since it is imposed only where the sentencing judge deems it necessary, and also since it is subject to such exemptions as the sentencing judge sees fit to allow. And the punishment imposed by s. 161(1)(c) is “not among the most onerous punishments, such as increased incarceration”, since it prohibits an offender only from having unsupervised contact with a child. It therefore follows that, if the retrospective operation of s. 161(1)(d) is a proportional and justified limit on an offender’s s. 11( i ) right, the retrospective operation of s. 161(1)(c) must be as well.

(4) The Proper Balancing

[ 158 ] I accept that the retrospective operation of the amendments to s. 161(1) works a relative increase in punishment that is not trivial. Section 161(1)(c)’s conditions on unsupervised contact with children regardless of location is more restrictive than the conditions imposable under the original provision. And s. 161(1)(d)’s restriction on Internet access goes much further in restricting an offender’s use of computers than did the original provision. I also accept that, like any other s. 11( i ) infringement, the retrospective operation of each has a deleterious impact on the rule of law and fairness in the criminal justice system, as each signifies an increase in possible punishment without notice to the individual.

[ 159 ] As for salutary effects, the evidence before Parliament and before this Court shows that a significant number of offenders convicted of designated sexual offences pose a risk to reoffend sexually against children. It also shows that most child victims are known to sexual offenders — they are not strangers taken from a public place, the victims of random chance. And it shows that Internet-based offending is rapidly increasing, which could realistically result in contact-based offences being committed against a child. Finally, it shows that the previous version of s. 161(1) could not address either of these issues — unsupervised contact with a child whether the child is known to the offender or not, and unsupervised access to the Internet for offenders who are likely to use the Internet to facilitate sexual offending.

[ 160 ] Balancing these deleterious and salutary effects at the proportionality stage of the s. 1 analysis entails, as the majority recognizes, “difficult value judgments” (para. 79). This is never a “neutral utilitarian calculus”: New Jersey v. T.L.O. , 469 U.S. 325 (1985), at p. 369, per Brennan J., dissenting in part. Despite claims to the contrary (see D. M. Beatty, The Ultimate Rule of Law (2004), at pp. 166-69; A. Barak, “Proportionality and Principled Balancing” (2010), 4 L. & Ethics Hum. Rts. 1 (abstract)), undertaking a proportionality analysis does not entail making a truly objective calculation, because it requires the court to weigh incommensurables — in this case, to weigh the deleterious impact on the sexual offender and on the rule of law against the possible benefit of protecting children from sexual offenders.

[ 161 ] Despite the impossibility of weighing incommensurables objectively, a reviewing court must nevertheless come to a reasoned conclusion. In my view, the salutary effects pursued are worth the cost in rights limitation: the harms sought to be addressed are grave, persistent, and worthy of Parliament’s efforts in the criminal law realm. The provisions are sufficiently tailored so that no offenders’ s. 11( i ) rights will be unduly limited — it is only those offenders who pose a risk to reoffend against children who will be subject to a s. 161(1) order, and it is only those offenders who pose a risk to reoffend either through unsupervised access to children or unsupervised use of the Internet who will be retrospectively subject to the impugned provisions. Neither of the impugned provisions works a drastic increase in the punishment imposed. On balance, the potential salutary effect of the retrospective operation of s. 161(1)(c) and s. 161(1)(d) of better protecting children from all sexual offenders who pose a risk to reoffend sexually against them, regardless of when the offender committed a designated offence, outweighs the modest impact on fairness and the rule of law.

III. Conclusion

[ 162 ] In my view, the retrospective operation of s. 161(1)(c) is a justified infringement on the appellant’s s. 11( i ) right. I would therefore dismiss the appeal and affirm the s. 161(1) order imposed by the majority of the Court of Appeal.

Appeal allowed in part, Abella and Brown JJ. dissenting in part.

Solicitors for the appellant: Eric Purtzki, Vancouver; Garth Barriere, Vancouver.

Solicitor for the respondent: Attorney General of British Columbia, Victoria.

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

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

Solicitors for the intervener Association des avocats de la défense de Montréal: Desrosiers, Joncas, Nouraie, Massicotte, Montréal.

Solicitors for the intervener the David Asper Centre for Constitutional Rights: John Norris, Toronto; University of Toronto Faculty of Law, Toronto.

Solicitors for the intervener the Criminal Lawyers’ Association (Ontario): Henein Hutchison, Toronto.

Solicitors for the intervener the British Columbia Civil Liberties Association: McCarthy Tétrault, Vancouver.

[1] In articulating this test, I do not decide whether s. 11( i ) would be infringed in circumstances akin to those in Whaling , in which accelerated parole review was retrospectively eliminated, thereby impacting the length of incarceration that was imposed as a sanction consequent to conviction.

[2] For example, the order imposed by McArthur J. in R. v. Levin , 2015 ONCJ 290 , at para. 113 (), illustrates how the Internet prohibition in s. 161(1)(d) can be crafted to fulfill the protective goals of the legislation while enhancing the offender’s rehabilitation process. See also the order imposed in R. v. Schledermann , 2014 ONSC 674 , at para. 13 ().

[3] The French text of s. 11( i ) reads as follows: “ Tout inculpé a le droit . . . i) de bénéficier de la peine la moins sévère, lorsque la peine qui sanctionne l’infraction dont il est déclaré coupable est modifiée entre le moment de la perpétration de l’infraction et celui de la sentence. ”

[4] It was not argued that other prohibition regimes in the Criminal Code (such as those found in ss. 810 , 810.1 , or 810.2 ) could have achieved the government’s objective in a real and substantial manner.

[5] It should be obvious from the above analysis that, had Parliament adopted a less tailored and discretionary regime, the 2012 amendments may very well have failed the minimal impairment test. It is accordingly unclear how my articulation of the purpose of the impugned amendments has rendered the minimal impairment analysis “redundant”, as my colleague Brown J. alleges (para. 138). On the contrary, the minimal impairment test remains an important part of assessing whether Parliament has discharged its burden under s. 1.

[6] In Oakes , this final stage of the proportionality analysis was initially conceived as a comparison between the deleterious effects of the limiting measure and the law’s objective. However, in Dagenais v. Canadian Broadcasting Corp ., 1994 39 (SCC) , [1994] 3 S.C.R. 835, Lamer C.J. reformulated the test to account for the “proportionality between the deleterious and the salutary effects of the measur[e]” because characterizing the final step “as being concerned solely with the balance between the objective and the deleterious effects of a measure rests on too narrow a conception of proportionality” (p. 889 (emphasis deleted)).

[7] See D. Grimm, “Proportionality in Canadian and German Constitutional Jurisprudence” (2007), 57 U.T.L.J. 383, at pp. 393-97; M. Zion, “Effecting Balance: Oakes Analysis Restaged” (2012-2013), 43 Ottawa L. Rev. 431; Barak, at pp. 380-82; F. Schauer, “Proportionality and the Question of Weight”, in G. Huscroft, B. W. Miller and G. Webber, eds., Proportionality and the Rule of Law: Rights, Justification, Reasoning (2014), 173, at pp. 181-85.

[8] Public Safety and Emergency Preparedness Canada, “Sex Offender Recidivism: A Simple Question”, by A. J. R. Harris and R. K. Hanson, March 2004 (online), at p. 7. This study used data from 10 follow-up studies of adult male sexual offenders with a combined sample of 4,724 offenders.

[9] Public Safety and Emergency Preparedness Canada, “Predictors of Sexual Recidivism: An Updated Meta-Analysis”, by R. K. Hanson and K. Morton-Bourgon, February 2004 (online), at p. 9. This article examined the research evidence of 95 different studies, involving more than 31,000 sexual offenders.

[10] Another individual, who had been involved with police training, testified as follows:

In 2010, I completed a pan-Canadian research project that examined the exponential increase of crimes of exploitation committed on or facilitated by the Internet against children in Canada and globally. Accessing images of child abuse — somewhat understated by the use of the term “child pornography” — child luring, trafficking, and travelling for the purpose of sexual offending are crimes increasingly facilitated by modern, ubiquitous technologies, especially the Internet, around the globe. . . .

. . . To prevent the ever-increasing numbers of crime, offenders must be disconnected from social networking sites through which they lurk and stalk.

( Evidence , No. 44, 3rd Sess., 40th Parl., January 31, 2011, at pp. 5-6)

[11] In one disturbing case summarized by an expert witness who testified before the parliamentary committee studying the Bill, two adults were chatting with each other in an online forum to set up an ‘exchange’ of children ( Evidence , No. 46, at p. 5, testimony of Lianna McDonald).