Dans cette affaire, la défense tentait de faire déclarer inconstitutionnels les articles du Code criminel qui régissent les ordonnances relatives au registre des délinquants sexuels, qui dans certains cas, s’appliquent à perpétuité
La Cour d’appel de l’Ontario en vient à la conclusion qu’une ordonnance rendue à perpétuité en vertu de l’art. 490.013(2.1) « is not arbitrary, overbroad, or grossly disproportionate and thus does not violate s. 7 of the Charter. » :
(1) Section 7 of the Charter
Everyone has the right to life, liberty and security of the person and the right not to be deprived thereof except in accordance with the principles of fundamental justice.
 The s. 7 analysis is concerned with laws that are inherently bad because they take away life, liberty and security of the person in a way that runs afoul of our basic values, embodied in the expression “principles of fundamental justice”: Bedford, at para. 96. Those principles, as expressed by the Supreme Court, include the requirement that laws must not be arbitrary, overbroad or grossly disproportionate.
 In order to demonstrate a breach of s. 7, the appellant must first show that the law deprives him of his life, liberty or security of the person. Once he establishes that s. 7 is engaged, he must then show that the deprivation in question is not in accordance with the principles of fundamental justice: Carter, at para. 55. While it is difficult to justify a s. 7 violation in light of the fundamental rights it protects, the Crown may be able to demonstrate that the infringement is justified under s. 1 of the Charter in some situations: Bedford, at para. 113, Carter, at para. 95; R. v. Michaud, 2015 ONCA 585 (CanLII), 127 O.R. (3d) 81, at para. 83.
 As I have noted, the Crown concedes that the impugned provision engages the appellant’s s. 7 liberty interest. Liberty protects the right to make fundamental personal choices free from state interference.
 There can be no doubt that the impugned provision, which requires lifetime registration and supervision, interferes with the appellant’s liberty rights and thus engages s. 7 of the Charter. The effect and severity of this breach are considered in more detail below.
 The s. 7 analysis does not end here. As the Supreme Court observed in Carter, at para. 71, laws regularly interfere with life, liberty and security of the person. The question is whether the law does so in a way that violates our basic values, described as the principles of fundamental justice. That is the central issue in this case.
 All three of these principles of fundamental justice – arbitrariness, overbreadth and gross disproportionality – hinge on the connection between the purpose of the legislation and its effect on the individual. I now turn to the application of these principles to this case.
 While the appellant did not pursue his argument that s. 490.013(2.1) is arbitrary, I will briefly address it for the sake of completeness.
 Arbitrariness asks whether the effect of the law on the individual bears some relation to the law’s purpose. An arbitrary law is one that is not capable of fulfilling its objectives: Carter, at para. 83. In Bedford, McLachlin C.J. observed, at para. 111:
There must be a rational connection between the measure that causes the s. 7 deprivation and the limits it imposes on life, liberty, or security of the person. A law that imposes limits on those interests in a way that bears no connection to its objective arbitrarily impinges on those interests. [Citations omitted. Emphasis in original.]
 As noted above, the SCAJ found that the purpose of SOIRA is to further public safety by enabling police to keep track of sex offenders who, by virtue of their past convictions, could be suspects in future crimes. It subjects those who are convicted of more than one offence to a longer registration period because, as a group, they have a greater propensity to re-offend. He found a rational connection between the purpose of the legislation and the means it employs, and thus concluded that the impugned provision was not arbitrary. I agree.
 I will begin by setting out the parties’ submissions with respect to overbreadth. I will then outline the Supreme Court’s approach to overbreadth before turning to my analysis of the issue in this case.
(a) The parties’ submissions
 The appellant submits that the structure of s. 490.013 demonstrates that the objective of the impugned provision is to tie the length of registration to the seriousness of the underlying offence and the moral culpability of the offender. Effectively, he submits, Parliament has decided that the more serious the offence, the greater the risk of recidivism – hence, the need to keep the offender on the registry for a longer period of time.
 The appellant argues that where the Crown exercises its discretion to charge an accused with more than one offence for several relatively minor sexual transgressions within the same transaction, the imposition of a longer period of registration is not connected to the seriousness of the underlying conduct or the offender’s moral culpability. There is no evidence that, in those circumstances, the offender necessarily poses a higher risk of recidivism. In contrast, there may be situations where the underlying conduct is more serious, but the offender is only charged with one offence and, consequently, is subject to a shorter registration period.
 Therefore, says the appellant, it is possible to imagine scenarios where s. 490.013(2.1) subjects offenders to lifetime registration where the offender is not at a higher risk of recidivism. The provision therefore captures reasonable hypothetical offenders who are not connected to the object of the legislation.
 The intervener supports this position. The legislative scheme uses the seriousness of the offence as a proxy for future risk. The intervener says there is a fundamental mismatch between s. 490.013(2.1) and this future risk-based criterion for setting the mandatory term of the SOIRA order. There are reasonable hypotheticals and real cases in the jurisprudence where lifetime registration makes no sense because the fact of more than one conviction is not a justifiable proxy for what the legislation is trying to accomplish. Conversely, where there is a very serious course of sexual abuse over numerous years but the Crown only charges one count, the offender would only be subject to a 20-year registration period.
 The Crown in response says that the purpose of SOIRA is set out in s. 2 of that Act: to prevent and investigate crimes of a sexual nature. Section 490.013(2.1) is meant to assist with this purpose by gathering information from persons convicted of sexual offences, irrespective of the level of seriousness of the offence. The commission of even a single summary offence is a reliable indicator of a heightened risk of re-offending. That will remain true over time. There will thus always be a presumptive risk that a person who is convicted of a sexual offence is at a higher risk of re-offending than a person in the general population.
 The Crown submits that there is a common sense inference to be drawn from this: an individual who is convicted of more than one sexual offence, even relatively minor ones closely connected in time, has a greater propensity to commit sexual crimes than an individual who goes no further than the first incident. The former is not a recidivist in the traditional sense (i.e. a person who re-offends after being convicted of an offence); however, as long as the offender commits legally discrete sexual offences that do not trigger the application of the principle in R. v. Kienapple, 1974 CanLII 14 (SCC),  1 S.C.R. 729, an inference may be drawn that the offender poses additional risk because the offender persists in acting in a particular manner. The fact that another offender may be treated differently as a result of prosecutorial discretion to charge only one count does not make the law overbroad or arbitrary for constitutional purposes.
 In other words, the number of criminal acts bear some relationship to the risk of re-offending. Some individuals in the group may be at greater risk of re-offending than others, but there is no way to know which ones will re-offend and which will not. All one can say is that, as a group, they are at enhanced risk to re-offend. Parliament is entitled to apply that gradation and categorization to offenders as long as there is some rational basis on which to do so. Where there is a rational basis to conclude there is some additional propensity that would not exist in the case of only one sexual assault, the imposition of the added period of registration does not make the law overbroad.
(b) The Supreme Court’s analytical approach to overbreadth
 The Supreme Court explained the overbreadth analysis in Bedford, at paras. 101-119, and in Carter, at paras. 85-88. A law is described as “overbroad” when it “goes too far and interferes with some conduct that bears no connection to its objective”: Bedford, at para. 101. This addresses situations where there is no rational connection between the purposes of the law and some, but not all, of its effects. As the Supreme Court explained it in Bedford, at para. 113:
Overbreadth allows courts to recognize that the law is rational in some cases, but that it overreaches in its effect in others. Despite this recognition of the scope of the law as a whole, the focus remains on the individual and whether the effect on the individual is rationally connected to the law’s purpose. For example, where a law is drawn broadly and targets some conduct that bears no relation to its purpose in order to make enforcement more practical, there is still no connection between the purpose of the law and its effect on the specific individual. [Emphasis in original.]
 A law will be unconstitutionally overbroad if its effect is to deprive some persons of liberty for reasons unrelated to its purpose: R. v. Safarzadeh-Markhali, 2016 SCC 14 (CanLII),  1 S.C.R. 180, at para. 22. Therefore, the question this court must consider is whether s. 490.013(2.1), which imposes mandatory lifetime registration, “goes too far and interferes with some conduct that bears no connection to its objective”: R. v. Appulonappa, 2015 SCC 59 (CanLII),  3 S.C.R. 754, at para. 71; Bedford, at para. 101.
 In Bedford, the Supreme Court emphasized that there is a heavy onus on the party challenging the legislation to establish that there is no connection between the effect of the law and its purpose. The standard is not easily met. Chief Justice McLachlin stated, at para. 119:
As noted above, the root question is whether the law is inherently bad because there is no connection, in whole or in part, between its effects and its purpose. This standard is not easily met. The evidence may, as in Morgentaler, show that the effect actually undermines the objective and is therefore “inconsistent” with the objective. Or the evidence may, as in Chaoulli, show that there is simply no connection on the facts between the effect and the objective, and the effect is therefore “unnecessary”. Regardless of how the judge describes this lack of connection, the ultimate question remains whether the evidence establishes that the law violates basic norms because there is no connection between its effect and its purpose. This is a matter to be determined on a case-by-case basis, in light of the evidence.
(c) The purpose of s. 490.013(2.1)
 The first step in the overbreadth analysis is to identify the purpose and the effect of the impugned provision itself, in this case s. 490.013(2.1), in the context of the legislative scheme of which it is a part: R. v. Moriarity, 2015 SCC 55 (CanLII),  3 S.C.R. 485, at para. 24.
 The Supreme Court’s jurisprudence has emphasized that the identification of the purpose of the legislation must be undertaken with precision. It summarized the considerations in Safarzadeh-Markhali, at paras. 24-29:
Whether a law is overbroad within the meaning of s. 7 turns on the relationship between the law’s purpose and its effect. It is critically important, therefore, to identify the purpose of the challenged law at the outset of the s. 7 inquiry.
Moriarity summarizes the considerations that guide the task of properly characterizing Parliament’s purpose in a s. 7 analysis into overbreadth.
First, the law’s purpose is distinct from the means used to achieve that purpose. A law’s means may be helpful in determining its objective, but the two must be treated separately.
Second, the law’s purpose should be characterized at the appropriate level of generality, which “resides between the statement of an ‘animating social value’ – which is too general – and a narrow articulation” that amounts to a virtual repetition of the challenged provision, divorced from its context.
Third, the statement of purpose should be both precise and succinct. Precision requires that courts focus on the purpose of the particular statutory provision subject to constitutional challenge.
Fourth, the analysis is not concerned with the appropriateness of the legislative purpose. The court must take the legislative objective “at face value” and assume that it is appropriate and lawful. The appropriateness of a legislative objective may be relevant to its constitutionality under other Charterprovisions. But it has no place in the s. 7 analysis of overbreadth. [Citations omitted.]
 In determining the purpose of the legislation in the s. 7 analysis, a reviewing court may look to three things: (a) statements of purpose contained in the legislation; (b) the text, content and scheme of the legislation; and (c) extrinsic evidence such as the history and evolution of the legislation: Safarzadeh-Markhali, at para. 31; see also Moriarity, at para. 31.
 Turning to this case, the impugned provision is s. 490.013(2.1), which I repeat for ease of reference:
(2.1) An order made under subsection 490.012(1) applies for life if the person is convicted of, or found not criminally responsible on account of mental disorder for, more than one offence referred to in paragraph (a), (c), (c.1), (d), (d.1) or (e) of the definition designated offence in subsection 490.011(1). [Italics in original.]
Statements of purpose in the legislation
 The purpose of s. 490.013(2.1) is informed by the provisions of the Criminal Code incorporating the SOIRA legislation, as well as the SOIRA legislation itself. Neither SOIRA nor the Code contains any specific statement of the purpose of s. 490.013(2.1). However, SOIRA contains a general purpose clause.
 SOIRA, as originally enacted, identified its purpose as “to help police services investigate crimes of a sexual nature by requiring the registration of certain information relating to sex offenders.” Amendments that came into force in 2011 added prevention to the purposes described in s. 2(1):
(1) The purpose of this Act is to help police services prevent and investigate crimes of a sexual nature by requiring the registration of certain information relating to sex offenders.
(2) This Act shall be carried out in recognition of, and in accordance with, the following principles:
(a) in the interest of protecting society through the effective prevention and investigation of crimes of a sexual nature, police services must have rapid access to certain information relating to sex offenders;
(b) the collection and registration of accurate information on an ongoing basis is the most effective way of ensuring that such information is current and reliable; and
(c) the privacy interests of sex offenders and the public interest in their rehabilitation and reintegration into the community as law-abiding citizens require that
(i) the information be collected only to enable police services to prevent or investigate crimes of a sexual nature, and
(ii) access to the information, and use and disclosure of it, be restricted.
The text, context and scheme of the legislation
 Part of the context giving rise to the creation of a federal sex offender registry was a growing public concern, exemplified in Ontario by the enactment of Christopher’s Law, about protecting children and other vulnerable victims from sex offenders through a system of careful monitoring of convicted sex offenders. In this court’s decision in Dyck, Blair J.A. explained, at paras. 99 and 100, the logic behind the creation of a sex offender registry:
The valid state objective of Christopher’s Law is to address this harm and protect the community from recidivist sex offenders. The means developed by the Act to accomplish this objective is the creation of a sex offender registry that can track the whereabouts of past sex offenders.
…The trial judge found (i) that sex offenders are prolific offenders who are likely to re-offend; and (ii) that a prior conviction for a sex offence is a reliable indication of risk and a proper method of assessing that risk. He also accepted the evidence of Dr. Collins that, although the numbers are uncertain, “as a group sex offenders are always going to be at risk” and that “if a person has offended sexually they are going to be a greater risk than the rest of the population for engaging in that activity”. There is therefore a presumptive risk of harm to potential victims of all ages posed by persons convicted of sex offences, even though there may be numerous offenders amongst that group who ultimately do not re-offend at all. These findings underpin the conclusion that it was open to the legislature to enact Christopher’s Law as a logical response to the risk caused by recidivist sex offenders.
(a) That sexual offences are particularly prevalent in our society and victimize mostly women and children and can have substantial physical, psychological and emotional impact on the victims, even occasionally leading to murder;
(b) That sex offenders as a group are prolific in their offending and are likely to re-offend, and that a prior conviction for such offences is a reliable indication of risk and a proper method of assessing that risk; and
(c) That the creation of a sex offender registry requiring those convicted of a designated sexual offence to register with the police and provide particulars of address and other information, including a photograph, and to keep that information current, creates a valuable investigative tool to aid the police when a sexual offence occurs, including the ability of the police to conduct a radius search and to investigate by means of geographic profiling.
 SOIRA and the sex offender information provisions of the Criminal Code (ss. 490.011 to 490.032), together with the federal Sex Offender Information Registration Regulations (see, for example, the Ontario Sex Offender Information Registration Regulations, SOR/2004-306), set out a comprehensive scheme for the registration of information about sex offenders. The scheme is plainly designed to ensure that the information is complete, current and accurate, so that police are able to identify and locate a convicted sex offender when seeking to prevent or investigate a sex crime.
 This court, and other appellate courts, have emphasized that while one of the underlying rationales for the enactment of SOIRAwas to facilitate the investigation of sexual offences by keeping track of sexual predators, the regime is not limited to “likely recidivists or sexual predators”: see R. v. Debidin, 2008 ONCA 868 (CanLII), 94 O.R. (3d) 421, at paras. 70, 77. See also: S.S.C., at para. 43; R. v. Redhead, 2006 ABCA 84 (CanLII), 56 Alta. L.R. (4th) 15, at paras. 36-38. The scheme of the legislation evinces a concern about the apprehension of sex offenders and prevention of sexual offences, particularly, but by no means exclusively, in relation to offences against children and other vulnerable groups.
History and evolution of the legislation
 Initially, the Criminal Code did not make SOIRA orders mandatory. The Crown had to apply for an order after sentencing and was not obliged to do so (s. 490.012). Moreover, the court had discretion to refuse the order. Section 490.012(4) of the Codeprovided:
The court is not required to make an order under this section if it is satisfied that the person has established that, if the order were made, the impact on them, including on their privacy or liberty, would be grossly disproportionate to the public interest in protecting society through the effective investigation of crimes of a sexual nature, to be achieved by the registration of information relating to sex offenders under the Sex Offender Information Registration Act.
 In 2010, the Minister of Justice, the Honourable Vic Toews, appeared before the Senate Standing Committee on Legal and Constitutional Affairs to discuss proposed amendments to SOIRA and further amendments to the Criminal Code. He noted that the existing discretionary system meant that 42 percent of those convicted of sexual offences had not been entered in the registry, resulting in inconsistent application of the law across the country and impairing the investigation of sexual offences because police were unaware of some offenders as potential suspects. He expressed the concern that due to the difficulty in determining which sex offenders will re-offend and which will not, the discretionary nature of the registration system meant that some recidivists would fall through the cracks. Other witnesses at committee hearings spoke of the need for an expeditious response in the prevention and investigation of sexual crimes.
 In 2011, the Act to Amend the Criminal Code and other Acts (Protecting Victims from Sex Offenders Act), S.C. 2010, c. 17, came into force. This Act amended both the Code and SOIRA and made a SOIRA order mandatory in all cases, removing any prosecutorial or judicial discretion to forego adding an offender convicted of a designated offence to the registry.
 The 2011 amendments also added the provision at issue in this appeal, s. 490.013(2.1), which imposes mandatory lifetime registration on persons convicted of more than one designated offence, subject to a provision permitting review after 20 years. In doing so, Parliament extended the circumstances in which lifetime registration was mandatory: see ss. 490.013(2)(c), (4) and (5).
 The Parliamentary record before us is somewhat sparse. The Crown has, however, provided excerpts from Hansard, including proceedings of the Senate Standing Committee on Legal and Constitutional Affairs.
 In submissions to the Senate Standing Committee, a lawyer speaking on behalf of the Criminal Lawyers’ Association expressed concern about the absence of an exemption in the proposed legislation for what he described as “relatively minor” sexual offences or sexual assaults. He gave the example of a teenager taking a sexual picture of herself and sending it to her boyfriend, making him guilty of possession of child pornography. He also referred to a “woman who is angry with her boyfriend for cheating on her and grabs him by the crotch to show her displeasure”.
The concern is that, in many cases, being convicted of more than one count does not necessarily correspond with the seriousness of the offence. The number of counts on an indictment, for example, is often arbitrary. A complainant who describes being sexually assaulted ten times during 2009 can result in a person being charged with one count of sexual assault covering January 1 to December 31, 2009, or ten separate counts depending on a variety of factors. The charge can go either way.
Our concern is that by arbitrarily saying a conviction for two counts requires putting the person on the registry for life has a disproportionate effect. I do not think that is what is intended in the bill. Imagine a situation where a person is charged with two counts and the Crown elects to proceed summarily because it is a less serious offence. If the person is convicted on one count, they are on the registry for ten years. If they are convicted on both counts, they are on the registry for life. In contrast, a person charged with one count of a serious sexual assault and the Crown proceeds indictably will be on the registry for twenty years.
Conclusion regarding the purpose of s. 490.013(2.1)
 Both the appellant and the intervener argued that the purpose of the legislation was to link a longer period of mandatory registration to the seriousness of the underlying offence. In oral argument, they described s. 490.013(2.1) as the “black sheep” of the legislative scheme because the commission of two offences triggers lifetime registration regardless of their severity.
 The problem with this interpretation of the legislation is that it fails to afford any significance to the enactment of the very provision at issue in this appeal. Prior to 2011, subsections (a), (b) and (c) of s. 490.013(2) already linked lengthier periods of registration to offences subject to longer potential prison sentences – that is, longer periods of registration for more serious offences. Moreover, s. 490.013(4) and (5) already captured traditional recidivists. When Parliament amended the legislation to add subsection (2.1), it is presumed not to have done so for a redundant purpose. Subsection (2.1) was clearly intended to achieve something other than what the legislation already provided for.
 In my view, the legislative purpose of s. 490.013(2.1) is not tied to the seriousness of the underlying offence. Nor is it tied, as the appellant suggests, to the moral culpability of the offender, as a result of the offender having committed and been convicted of two or more designated offences. Rather, it is concerned with the offender who, by virtue of having been convicted of more than one offence, is a member of a group with an enhanced risk of re-offending.
 In light of the text of the provision, SOIRA’s statement of purpose and statement of principles, the scheme of the legislation, the context of the provision and its legislative history, I find that the animating social value of SOIRA is the protection of the public. The purpose of s. 490.013(2.1) of the Criminal Code is to further public safety by subjecting sex offenders who are at enhanced risk of re-offending to a longer period of registration. In the case of offenders like the appellant who fall within s. 490.013(2.1), this means registration for life, subject to the right to apply for termination of the order after 20 years, pursuant to s. 490.015(1)(c).
 Finally, the effect of the impugned provision is that offenders convicted of more than one designated offence will be subject to lifetime SOIRA registration, subject to the right to apply for termination after 20 years.
(d) Rational connection
 Having identified the purpose of s. 490.013(2.1), the next step in the overbreadth analysis is to determine whether it deprives individuals of liberty in cases that do not further that purpose: Appulonappa, at para. 27. In that inquiry, this court can consider reasonable hypothetical cases of the law’s application: Appulonappa, at para. 28.
 On this appeal, the intervener put forward the hypothetical of an 18-year-old accused who is convicted of two summary conviction sexual assaults for two brief incidents, minutes apart, of groping over clothing. The intervener argues that s. 490.013(2.1)is overbroad because it is irrational to impose a lifetime SOIRA order on the sole basis of an additional conviction in cases like this hypothetical, because the second conviction does not say anything meaningful about the risk of re-offending.
 In addition to considering this hypothetical, some analysis of the purpose and effect of the impugned provision is provided by the two conflicting decisions that have been put to us – the decision of this court in Dyck, on the one hand, and the decision of the Alberta Court of Queen’s Bench in R. v. Ndhlovu, 2016 ABQB 595 (CanLII), 44 Alta. L.R. (6th) 382, on the other.
R. v. Dyck
 As I have already noted, in Dyck this court dismissed a challenge to Christopher’s Law on both federalism and Chartergrounds. Mr. Dyck had been convicted of sexual interference and in 2002, he was charged with failing to report under Christopher’s Law.
 The court had before it a considerable evidentiary record adduced by both parties, including evidence as to the behaviour of sex offenders and their risk of recidivism, the operation of the registry and the legislative history.
 Mr. Dyck argued that the legislation was broader than necessary because it swept into the registry all offenders convicted of the designated offences without regard to whether they were likely to re-offend or not, or whether there were reasons in their particular case for an exemption. As under SOIRA, s. 7(1)(c) of Christopher’s Law imposes lifetime reporting requirements on offenders convicted of more than one sex offence. However, the challenge in Dyck was to the scheme of mandatory registration itself and that provision was not directly at issue.
 Blair J.A. held that the legislature acted rationally in pursuit of a legitimate state interest to protect the community from recidivist sex offenders by creating a mandatory registry to enable police to track past offenders. He concluded that the means used were rationally connected to the objective.
 He also held that the mandatory inclusion of offenders in the registry regardless of their risk of re-offending did not make the law overbroad, because there was no way of knowing which would re-offend and which would not. He stated, at paras. 122-124:
Finally, Christopher’s Law is not constitutionally flawed in terms of the scope of persons that are captured by its provisions, in my opinion. The trial judge concluded that Christopher’s Law failed on overbreadth grounds because it required all persons convicted of designated sex offences to register regardless of whether that particular offender posed a risk to re-offend…
I agree, however, with the Crown’s submission that individualized risk assessment is not a Charterprerequisite…
Once legislators have determined there is more than a minimal risk of harm, Parliament and the legislatures are entitled to make policy choices within a reasonable range of options.
It is clear on the evidence called at trial that a conviction for a designated sex offence attests to a significant risk of recidivism. There is no way to know in advance which sex offenders will re-offend and which will not, and that is why all are included on the Registry. What these findings mean, in the context of the enactment of Christopher’s Law, is that the choice of the Legislature to include all persons convicted of designated sex offences in the Registry was logical, rational, and not grossly disproportionate to the state objective.
R. v. Ndhlovu
 The appellant also referred us to Ndhlovu, in which Moen J. of the Alberta Court of Queen’s Bench concluded that s. 490.012, which makes SOIRA registration mandatory, infringed s. 7 of the Charter because it was both overbroad and grossly disproportionate insofar as it removes judicial discretion to refuse to place on the registry offenders who present no risk of re-offending.
 In Ndhlovu, the accused pleaded guilty to two counts of sexual assault. The offences occurred at a party to which he had been invited by the complainant, RD. RD and a female friend, CB, reported that he had touched their buttocks while the three were posing for a picture and that he had touched RD’s thighs and buttocks during a conversation without their consent. In the early morning, RD awoke to find the accused’s fingers in her vagina. She told him to stop. He tried to touch her a second time at which point she pushed him away and he left. He was sentenced to six months in jail and three years’ probation. Mr. Ndhlovu was also subject to lifetime SOIRA registration under s. 490.013(2.1) because he was convicted of two designated offences.
 In sentencing Mr. Ndhlovu, Moen J. found that he was at a very low risk to re-offend. He was 19 years of age and had no criminal history. He had taken responsibility for his actions and shown remorse for what he had done. He had stopped drinking.
 Mr. Ndhlovu challenged the mandatory nature of SOIRA orders under s. 7 of the Charter. In considering the Charter issues, Moen J. found that the reporting requirements of SOIRA were “significant” and onerous, referring to Redhead, at para. 33; S.S.C.; R. v. Have, 2005 ONCJ 27 (CanLII), 194 C.C.C. (3d) 151; and the judgment of the SCAJ in this case.
 She found that the law was not arbitrary; there is a rational connection between requiring sex offenders to provide up-to-date information and the goal of investigating and preventing sex crimes. She agreed, at para. 91, with the SCAJ’s observation that “the information collected can make it easier to investigate sex crimes by identifying individuals who, by virtue of past convictions, may be considered suspects in similar crimes”, and that as a matter of common sense and experience, individuals who are convicted of sex crimes have an increased propensity to commit sex crimes in the future.
 While acknowledging that there was a statistical probability that a sex offender will re-offend, she concluded, at para. 119, that the law was overbroad because it captured individuals who have little or no chance of re-offending:
In my view, the mandatory registration for all sex offenders upon conviction of two or more offences, without regard to the seriousness of the offences or the offender’s propensity to reoffend is overbroad. The goal of the legislation is to assist police with investigating past crimes and preventing new ones. The Crown conceded that the Registry captures individuals who will never re-offend. In my view, including offenders on the Registry who have little to no chance of reoffending bears no relation to protecting the public. Subjecting all offenders, regardless of their future risk, to onerous reporting requirements, random compliance checks by the police, and internal stigma, goes further than what is necessary to accomplish the goal of protecting the public, and is therefore overbroad.
 In reaching this conclusion, Moen J. rejected the Crown’s argument that the mandatory inclusion of all offenders in the registry was connected to the law’s purpose because it was impossible to know in advance which offenders would re-offend and which would not. She concluded that this type of argument was more appropriately brought under s. 1 of the Charter.
 In light of her conclusion that s. 490.012 breached s. 7, Moen J. declined to make any SOIRA order with respect to Mr. Ndhlovu.
Relevance of Dyck and Ndhlovu
 Ultimately, neither case put to us deals with the issue before this court. Both dealt with different provisions and addressed a different concern – namely, the mandatory requirement that SOIRA orders be imposed on all offenders convicted of a designated offence. I turn now to my analysis of the provision at issue in this appeal.
Is s. 490.013(2.1) overbroad?
 The question before this court on the overbreadth analysis is whether there is some connection between the purpose of s. 490.013(2.1) and all of its effects. As noted above, for a law to be overbroad, there must be no connection between its effects and its purpose in some cases.
 The legislation at issue in this appeal involves the assessment of the risk of future harm, an exercise that is inherently imprecise. The onus is on the appellant to establish that, in some cases, there is no rational connection between the purpose and the effect of the law. The Supreme Court in Bedford described, at para. 119, the ultimate question as “whether the evidence establishes that the law violates basic norms because there is no connection between its effect and its purpose” (emphasis in original). As the Supreme Court observed, “[t]his standard is not easily met”: Bedford, at para. 119.
 The purpose of s. 490.013(2.1) of the Criminal Code, as I have found, is to further public safety by subjecting sex offenders who are at enhanced risk of re-offending to a longer period of registration.
 No issue was taken with the premise underlying SOIRA that sex offenders as a group present a greater risk of committing future sexual offences than the rest of the population: see Dyck, at paras. 100, 105; Redhead, at para. 38. While there is no evidence in the record to support the proposition that offenders who are convicted of more than one designated offence are at enhanced risk of re-offending, there is also no evidence to the contrary.
 Instead, we are faced with conflicting arguments appealing to “common sense”. The Crown, on the one hand, asks us to infer that an individual who is convicted of more than one designated sexual offence has a greater propensity to commit sexual crimes than an individual who is convicted of only one such offence.
 The appellant, on the other hand, supported by the intervener, advances reasonable hypothetical scenarios to argue that the appropriate inference to be drawn is this: in some cases, an individual who is convicted of committing two sexual offences, particularly two of which are minor and proximate in time, is at no greater risk of re-offending than an individual who commits only one such offence.
 I am not persuaded by the intervener’s suggestion that in the reasonable hypothetical of a young offender convicted of sexual assault for two brief incidents of groping over the clothing, the second sexual assault, proximate in time, does not say anything meaningful about the risk of re-offending. In my view, and in the absence of evidence to the contrary, the second sexual assault can reasonably be regarded as demonstrating the offender’s persistence and impulsiveness, and therefore his enhanced risk of re-offending. The same could be said of the appellant, who sexually assaulted the complainant three times over the course of two and a half hours.
 Parliament was entitled to draw the inference that conviction of more than one sexual offence is logically probative of an offender’s enhanced propensity to commit further sexual offences and is linked to an elevated risk of re-offending.
 The appellant has failed to displace the common sense inference that the criterion used by Parliament to assess future risk (i.e. more than one conviction for sexual offences) can reasonably be said to make the risk of future harm – re-offending – higher than it would have been absent the second or additional convictions. For that reason, there is some connection between the effect of the law, the lifetime SOIRA registration imposed on an offender, and the purpose of the provision.
 Further, I do not accept the appellant’s submission that the law is overbroad because prosecutorial discretion plays a role in determining which offenders are caught by the provision and which are not. Prosecutorial discretion is a well-established feature of the justice system. In the prosecution of sexual offences, prosecutors decide whether to charge at all, whether to charge by indictment or by summary conviction, and whether to charge a lesser included offence that may not be a designated offence. Each of these decisions may affect whether SOIRA registration is mandatory or the length of the registration. Prosecutorial discretion cannot save a law that is otherwise unconstitutional on its face: R. v. Nur, 2015 SCC 15 (CanLII). But a law does not become arbitrary or overbroad simply because similarly-situated accuseds are treated differently through the exercise of prosecutorial discretion. The fact that some offenders may escape lifetime registration as a result of prosecutorial discretion does not make the legislation overbroad in relation to those who are caught by the provision.
 I therefore agree with the Crown’s submission that simply because another offender may benefit from a decision by the Crown to charge only one count (thereby prosecuting multiple incidents as part of a single transaction), this does not support a conclusion that there is no rational basis to require a longer compliance period for an offender who has demonstrated a propensity to commit sexual offences.
 A similar distinction can be made regarding the Crown’s discretion to determine whether to proceed with certain charges summarily or by way of indictment. Under s. 490.013(2)(a), an offender is only required to register for a period of ten years if the offence is prosecuted summarily, rather than 20 years if convicted of an indictable offence.
 The legislative history of the impugned provision makes it clear that one of the purposes of the 2011 amendments was to remove prosecutorial and judicial discretion related to whether or not a SOIRA order was imposed on an offender because an individual offender’s risk of re-offending could not be predicted with precision. Subjecting an offender to a shorter period of registration based on a judge’s or prosecutor’s assessment of the individual’s risk of re-offending would undermine the integrity of the registry and, thereby, impede the ability of police to rapidly and effectively prevent or investigate a sex crime.
 Two judges examining the sexual assaults carried out by the appellant or by Mr. Ndhlovu or by the intervener’s hypothetical offender might well disagree on their likelihood of recidivism. One of the purposes of the legislative amendments was to remove discretion and, in its place, to substitute a reasonable proxy to assess the risk of future harm. Parliament determined that the stakes were too high to leave it to the discretion of the Crown or the court.
 In my view, this case is distinguishable from Safarzadeh-Markhali. The impugned provision in that case was broadly worded. McLachlin C.J. noted, at para. 47, that its purpose was to enhance public safety and security. The provision caught individuals denied bail primarily because of a prior conviction without specifying the nature or number of prior convictions that would warrant the denial of bail. McLachlin C.J. emphasized, at para. 53, that the law was overbroad because it resulted in bail being denied based on the person’s criminal record, without considering whether the nature of the past offences was indicative of a threat to public safety or security.
 In Safarzadeh-Markali it was possible for the court to say, as a matter of common sense, that the law was overbroad on its face, because it would capture an offender who had two or three previous convictions for failure to appear, or some other minor infraction, but did not pose any threat to public safety or security. The bail justice’s endorsement that bail was denied “primarily because of a previous conviction” was, in the words of the Chief Justice, “an inexact proxy for the danger that an offender poses to public safety or security.”
 In this case, by contrast, it cannot be said that as a matter of common sense the commission of two or more designated offences is not a reasonable proxy for an enhanced risk of re-offending. The appellant and the intervener ask us to infer that in some cases there is no connection between the commission of two or more offences and an enhanced risk of recidivism. I am not prepared to draw that inference. Bearing in mind that we are considering the inherently imprecise assessment of future risk, and there being no dispute that the commission of a sex offence is a reasonable proxy for the risk of re-offending, it was open to Parliament to conclude that the commission of two or more offences is a reasonable proxy for an enhanced risk of re-offending, warranting a longer registration period.
 In summary, the appellant has failed to discharge his onus to establish that the law is overbroad. In my view, being convicted of two sexual offences within a short period of time – even two “relatively minor” offences, in the case of the hypothetical – is rationally connected to the objective of s. 490.013(2.1), because it is a reasonable proxy for an enhanced risk of re-offending. This connects the effect of the law to its purpose, regardless of the severity of the underlying offence.
(4) Gross Disproportionality
 I turn now to the appellant’s argument that s. 490.013(2.1) is grossly disproportionate.
 A law will be found to be grossly disproportionate under s. 7 of the Charter “in extreme cases where the seriousness of the deprivation is totally out of sync with the objective of the measure”: Bedford, at para. 120. The Supreme Court affirmed the test for gross proportionality in Carter, at para. 89:
This principle is infringed if the impact of the restriction on the individual’s life, liberty or security of the person is grossly disproportionate to the object of the measure. As with overbreadth, the focus is not on the impact of the measure on society or the public, which are matters for s. 1, but on its impact on the rights of the claimant. The inquiry into gross disproportionality compares the law’s purpose, “taken at face value”, with its negative effects on the rights of the claimant, and asks if this impact is completely out of sync with the object of the law. The standard is high: the law’s object and its impact may be incommensurate without reaching the standard for gross disproportionality. [Emphasis in original. Citations omitted.]
 The task for this court is, therefore, to weigh the effects of a lifetime SOIRA registration on an individual offender against the purpose of furthering public safety by subjecting sex offenders who are at enhanced risk of re-offending to a longer period of registration. In this regard, the court’s task is to balance the negative effects with the purpose of the law, not its benefits to society more broadly: Bedford, at para. 121.
 The appellant submits that lifetime registration is out of sync with the objective of the legislation to the point that it is grossly disproportionate. One offender could be on the registry for life for two relatively minor sexual offences committed in the course of the same transaction, whereas an offender who commits a serious sexual assault would only be on the registry for 20 years if the Crown only charges one count. The gulf between the two offenders is so wide that it is not in keeping with the purpose of the legislation.
 The Crown submits that the length of a reporting requirement, by itself, can never justify a finding that the provision is grossly disproportionate. This court found in Dyck and Debidin that registration only impaired the offender’s liberty in a minimal way. Most of the impacts on the offender flow not from the reporting requirements, but from the convictions themselves.
 I agree with the Crown’s submission that the effect of registration on an offender’s liberty is modest. In Dyck, Blair J.A. observed, at para. 104, that the reporting requirements imposed by Christopher’s Law, which for the purposes of this appeal is analogous to SOIRA, are particularly modest given the fact that an offender has already been tried and convicted of a designated sexual offence. Blair J.A. noted, at para. 106, that “[t]he requirements are limited in their informational scope, do not prohibit the appellant from going anywhere or doing anything, are no more intrusive than other state-imposed registration requirements, impose minimal stigma on the offender, and are not publicly known.” Moreover, as noted by the Crown, any stigma flows from the conviction itself, not registration. The statute contains provisions designed to protect the privacy of registrants and to restrict access to the registry. There is no evidence on this appeal that would justify concluding that a lifetime registration order has anything other than a modest impact on an offender’s liberty interest.
 This impact must then be measured against the importance of the objective of the provision. There is no question that furthering public safety by subjecting offenders who are at enhanced risk of re-offending to a longer period of registration is a pressing and important objective. In light of the importance of the purpose pursued by the provision and its relatively modest impact on an offender’s liberty, I find that the provision is not grossly disproportionate.
 I conclude that s. 490.013(2.1) is not arbitrary, overbroad, or grossly disproportionate and thus does not violate s. 7 of the Charter.