COURT OF APPEAL FOR ONTARIO
CITATION: Cycle Toronto v. Ontario (Attorney General), 2026 ONCA 582
DATE: 20260814
DOCKET: COA-25-CV-1047
Fairburn A.C.J.O., Huscroft and Zarnett JJ.A.
BETWEEN
Cycle Toronto, Eva Stanger-Ross and Narada Kiondo
Applicants (Respondents)
and
Attorney General of Ontario and Minister of Transportation
Respondents (Appellants)
and
The Canadian Public Health Association, Greenpeace Canada and For Our Kids Toronto
Interveners (Interveners)
Josh Hunter, Cara Zwibel and Elizabeth Guilbault, for the appellants
Andrew Lewis, Greta Hoaken, Braxton Murphy, Catherine Dunne, Lindsay Beck and Bronwyn Roe, for the respondents
George Avraam, Jennifer Bernardo and Ryan Carroll, for the intervener Canadian Constitution Foundation
Nikolas De Stefano, for the intervener Canadian Public Health Association
Louis Century, Amy Chen and Umaiyahl Nageswaran, for the intervener Greenpeace Canada
Vilko Zbogar, Meaghan Daniel and Vibhu Sharma, for the intervener For Our Kids – Toronto
Megan Savard and Cheryl Milne, for the intervener David Asper Centre for Constitutional Rights
Heard: January 28, 2026
On appeal from the judgment of Justice Paul B. Schabas of the Superior Court of Justice, dated July 30, 2025, with reasons reported at 2025 ONSC 4397, 177 O.R. (3d) 341.
OVERVIEW
[1] The Ontario legislature exercised its law-making authority over highway traffic in the province to require the removal of bicycle lanes from three downtown Toronto streets. This is the sort of decision legislatures make routinely without attracting any concerns about the constitutionality of their actions.
[2] The application judge saw things differently. He concluded that the removal of the bicycle lanes constituted a serious violation of the most basic and fundamental rights in the Canadian Charter of Rights and Freedoms: it deprived bicyclists not only of security of the person but also of the right to life itself, and it did so in a manner that violated the principles of fundamental justice. He found, in particular, that the purpose of removing the bicycle lanes – reducing traffic congestion – was not supported by the government’s internal advice or expert evidence tendered in the application and would not be achieved by the legislation. The application judge went on to conclude that this was a violation of s. 7 of the Charter that could not be saved by s. 1.
[3] The application judge’s decision presents as a straightforward application of orthodox s. 7 principles. But the application of those principles, divorced from a broader understanding of the constitution of which they are a part, has led to a surprising outcome: the conclusion that routine legislation regulating road usage is unconstitutional.
[4] Something has gone wrong.
[5] The application judge erred in law in interpreting and applying s. 7. But his decision is wrong in a more profound sense, for it subverts a fundamental principle of our democratic constitutional order: a legislature cannot bind its successors. Legislatures in the Westminster tradition are free to undo legislation by amending or repealing it as they see fit. That is the nature of democratic lawmaking in our constitutional order. Of course, all laws are subject to the Charter, but this does not preclude the amendment or repeal of legislation the legislature was under no constitutional obligation to have passed in the first place. Amendment or repeal of legislation may breach the Charter only if the breach can be established independent of the prior state of the law.
[6] The Charter does not require the legislature to establish bicycle lanes. Thus, the Ontario legislature is as free to establish bicycle lanes as it is to remove them or authorize their removal. The Charter has nothing to say about whether bicycle lanes are a good or a bad idea – wise or unwise policy. These judgments are no business of the courts. The government is not constitutionally obligated to make policy decisions that accord with policy advice it receives from its advisors, nor is the legislature obligated to legislate in accordance with that advice – let alone the advice of academic experts.
[7] The bottom line is this: there is no constitutional obligation to regulate the use of roads in any particular manner – no constitutional obligation to build bicycle lanes or to retain any that have been built. The legislature is free to legislate to remove a bicycle lane from a road – indeed, it may remove a road entirely – without implicating the Charter, and in particular, without implicating anyone’s rights under s. 7 of the Charter.
[8] The appeal must be allowed.
BACKGROUND
[9] Bicycle lanes have been established by the City of Toronto on some roads pursuant to the City’s bylaw-making authority. On November 25, 2024, the Ontario legislature passed the Reducing Gridlock, Saving You Time Act, 2024, S.O. 2024, c. 25. Among other things, the Act added s. 195.6 to the Highway Traffic Act, R.S.O. 1990, c. H.8 (“HTA”), requiring the Minister to remove bicycle lanes from three major streets in Toronto: Bloor Street, University Avenue, and Yonge Street.
[10] In January 2025, the respondents, Cycle Toronto and two individual users of the bicycle lanes, Eva Stanger-Ross and Narada Kiondo, brought an application for a declaration that s. 195.6 of the HTA violates s. 7 of the Charter.
[11] Shortly after the application was filed, the legislature repealed and replaced s. 195.6 of the HTA. The new s. 195.6 directs the Minister to “reconfigure” the target bicycle lanes rather than “remove” them.[1] The HTA was also amended to grant the Lieutenant Governor in Council the power to make regulations concerning the removal of bicycle lanes pursuant to s. 195.6.
THE APPLICATION JUDGE’S DECISION
[12] The application judge found that the amendment to s. 195.6 simply involved a “slight tweaking of the language” and that the change from “remove” to “reconfigure” was a “distinction without a difference”. He also noted that any reconfiguration would inevitably require removal of at least some of the physical separation of bicycle lanes from vehicular traffic and that a live controversy continued between the parties concerning the application of s. 7. Accordingly, the application judge concluded that the application was not moot but added that he would have exercised his discretion to hear the application in any event.
[13] The application judge stated that the respondents were not asserting a positive right to bicycle lanes; they were, instead, asking the court to strike down legislation that would remove a safety feature on roads. He rejected the argument that the government was being held to a constitutional “baseline” established by prior legislation. He stated that although the legislature is entitled to remove a benefit once established, in doing so it cannot violate the Charter.
[14] The application judge found that removal of the bicycle lanes would make roads less safe for cyclists and for other users of the road, resulting in an increased likelihood of collisions, injuries, and death. On this basis, he concluded that s. 195.6 engaged both the right to life and security of the person in s. 7 of the Charter. The application judge rejected the argument that the risk to life and security of the person was caused by the cyclists’ choices rather than s. 195.6, finding the choice to cycle was not unconstrained and that, for many, there were no suitable alternatives to riding their bicycles on the relevant streets. He also noted that cycling is a lawful activity and that the real question is whether the law makes that lawful activity more dangerous.
[15] Furthermore, the application judge found that removal of the bicycle lanes would not achieve the stated objective of the legislation – the alleviation of traffic congestion – but would in fact worsen it. This finding was based on expert evidence proffered by the respondents as well as advice the government had received prior to proposing s. 195.6. The application judge concluded that there was therefore no connection between the purpose of the impugned provision and its true effects, and as a result the law was arbitrary. The application judge also concluded that the impact of the removal of the bicycle lanes was grossly disproportionate to the asserted benefit of “saving some drivers of cars a few minutes of travel time.” Thus, s. 7 was violated.
[16] The application judge concluded that this violation could not be saved under s. 1. He accepted that the objective of reducing traffic congestion was pressing and substantial, but found that the law was not rationally connected to its objective because restoring lanes for cars would not reduce congestion or further the objective of the legislation. The application judge also found that the law did not minimally impair the rights of cyclists: “under the guise of reducing congestion” it caused people to ride in more dangerous traffic, putting them in “harm’s way”. Finally, he found that the benefits of the law did not outweigh its negative impacts.
[17] In the result, the application judge declared the original version of s. 195.6 to be unconstitutional. He also declared that any steps taken to “reconfigure” the target bicycle lanes to install a lane for motor vehicles in order to reduce congestion would violate s. 7 of the Charter and would not be saved under s. 1.
ISSUES ON APPEAL
[18] The appellants argue that the application judge erred in the following respects:
(1) In failing to recognize that s. 7 is only concerned with deprivations of life and security of the person caused by the coercive power of the state;
(2) In failing to recognize that s. 7 does not create a positive right to state-provided harm reduction measures;
(3) In finding s. 195.6 to be arbitrary and grossly disproportionate;
(4) In finding that s. 195.6 could not be justified under s. 1; and
(5) In assuming jurisdiction to judicially review hypothetical future action taken under an unchallenged statutory provision.
[19] As I will explain, the appeal can be resolved on the basis of the first three issues. As a result, it is unnecessary to address the fourth and fifth issues.
DISCUSSION
The role of the courts on Charter challenges
[20] It is important to avoid approaching judicial review under the Charter in a technical or mechanical fashion. The Charter is part of a larger democratic constitutional order and must be interpreted and applied in the context of that order and all that it entails.
[21] The starting point is this: ours is a Westminster parliamentary democracy with well-defined constitutional roles. As the preamble to the Constitution Act, 1867 states, Canada has “a Constitution similar in Principle to that of the United Kingdom”, a constitution that by 1867 had been directing the political branches of government for centuries. In that system, the executive proposes legislation but the decisions whether to legislate and how to do so belong to the legislature. Courts play no role in the policy development process, no role in the choices made by the government in proposing legislation, and no role in the legislature’s decision to pass a law. The role of the courts is to adjudicate disputes arising under the law once it is enacted. This separation of powers is a bedrock principle of constitutional law and it is uncontroversial in our legal tradition: see Ontario v. Criminal Lawyers’ Association of Ontario, 2013 SCC 43, [2013] 3 S.C.R. 3, at paras. 27-31 ; The Honourable Malcolm Rowe, Chris Puskas and Allyse Cruise, “The Separation of Powers in Canada” (2024) 1 S.C.L.R. (3d) 323.
[22] Now, courts learn a great deal about a matter in the course of judicial review proceedings, and judges may be tempted to comment on matters of policy. The temptation to do so may be great; judges may suppose that they have much to offer to public policy debate. But the temptation must be resisted. It is not just that courts have no expertise in policy making – no expertise in making the sorts of polycentric decisions that legislatures make routinely. It is illegitimate for courts to involve themselves in matters of legislative policy. Policy is for politics. Whether courts think legislation is wise or unwise, is irrelevant so far as its constitutionality is concerned, and constitutionality is the only question with which courts are legitimately concerned: see e.g., Quebec (Attorney General) v. Canada (Attorney General), 2015 SCC 14, [2015] 1 S.C.R. 693, at para. 3 ; R. v. Malmo-Levine; R. v. Caine, 2003 SCC 74, [2003] 3 S.C.R. 571, at para. 173 ; and Reference re ss. 193 and 195.1(1)(c) of the Criminal Code (Man.), 1990 CanLII 105 (SCC), [1990] 1 S.C.R. 1123, at p. 1142, per Dickson C.J.
[23] With respect, the application judge lost sight of this principle when, in the course of outlining the background to the application, he weighed in on policy and political considerations. He commented extensively on the political background to the legislation, recounting remarks by the Premier and Minister of Transportation, contested data concerning road use, and exchanges at legislative committee proceedings. He quoted at length from members of the public and organizations opposed to passage of the legislation, as well as the Mayor of Toronto and the City Manager – all of this prior to addressing the only matter with which the court was properly concerned: whether the legislation violated the Charter. Indeed, the application judge’s decision is replete with criticism not only of the legislation but also of the government’s conduct in promoting its passage. He went so far as to express concern that the legislation was passed despite the government’s internal policy advice and public criticism of the bill.
[24] None of this was relevant to the task before him.
[25] Lest there be any doubt: in formulating policy or legislating, neither the government nor the legislature is under any obligation to accept policy advice, whether from bureaucrats, experts, or members of the public. The choices they make are the stuff of politics. We are governed by those whom we elect and they are accountable to the people for the choices they make, not to the courts.
[26] It is important to remember that judicial independence is a two-way street: the government and legislature must obviously respect the independence of the courts, but courts must respect the roles and responsibilities of the other branches of government. Judicial review under the Charter is not an occasion for courts to comment on or critique policy or political decisions, as Tulloch C.J.O. explained in Dorceus v. Ontario, 2026 ONCA 321, at paras. 23-24:
[T]he cornerstone of the separation of powers is mutual respect. …
Because respect is reciprocal, the judiciary must remain mindful not to intrude into the domains of the political branches. The courts of law exist to resolve legal disputes. The courtroom is not a political forum. Nor is it a proxy for the public inquiry process in which the policy judgments of the legislative and executive branches can be second-guessed and re-weighed. These differences between the judicial function and the political arena must be preserved. (Citations omitted.)
1. Drover and the threshold for the application of s. 7
[27] The scope of s. 7 of the Charter – specifically, whether it applies beyond matters involving the administration of justice – has not been resolved definitively by the Supreme Court. In Drover v. Canada (Attorney General), 2025 ONCA 468, 177 O.R. (3d) 561, leave to appeal granted, [2025] S.C.C.A. No. 392, a majority of this court held that s. 7 applies beyond individuals’ interactions with the administration of justice, at least in the context of the liberty interest. In that case, the majority held that a legislative requirement that electoral returning officers reside in the ridings in which they were appointed – a requirement that clearly did not implicate the administration of justice – deprived the applicant of his liberty in a manner that was not in accordance with the principles of fundamental justice.
[28] The appellants did not ask this court to reconsider its decision in Drover but sought instead to confine its holding to the liberty right in s. 7. If s. 7 applies to deprivations of life or security of the person only in the context of the administration of justice, then the application judge’s decision is plainly erroneous, for there is no question that the impugned legislation does not involve the administration of justice. It established no offence or sanction, civil or criminal. It merely required the removal of bicycle lanes from three Toronto streets.
[29] However, it is not clear that it is possible to interpret Drover as the appellants suggest. The proper scope of s. 7 will ultimately have to be decided by the Supreme Court, and I note that since this appeal was argued, the court has granted leave to appeal in Drover.
[30] In these circumstances, I will proceed on the assumption that s. 7 may be engaged beyond the administration of justice context. As I will explain, even assuming that s. 7 may be engaged in the wider context that Drover suggested, it is not engaged in this case. The impugned legislation does not effect a deprivation of life, liberty, or security of the person. Even if it did, any such deprivation would be consistent with the principles of fundamental justice.
2. The Charter does not preclude the repeal of legislation
[31] The appellants argue that, just as s. 7 does not mandate that the state take action to reduce the risk of harm caused by third parties or to safeguard citizens from the hazards of life, it does not require the state to continue measures it has previously taken to achieve those ends.
[32] This is undoubtedly correct, but the point can be stated more broadly.
[33] The Charter is part of the supreme law of Canada, and any legislation inconsistent with it is of no force or effect to the extent of the inconsistency: Constitution Act, 1982, s. 52. Thus, legislation that undoes prior legislation is subject to the Charter, just as the prior legislation was. But there is a significant difference: as long as the prior legislation was not mandated by the Charter, undoing it by amendment or repealing it altogether is necessarily Charter-compliant, regardless of how beneficial or important the prior legislation may be thought to have been. This is so “even where those measures accord with or enhance Charter values”: Lalonde v. Ontario (Commission derestructuration des services de santé) (2001), 2001 CanLII 21164 (ON CA), 56 O.R. (3d) 505 (C.A.) at para. 94. To hold otherwise would deny one of the most fundamental rules of our democratic constitutional order: legislatures cannot bind their successors. Amendment or repeal of legislation may breach the Charter only if the breach can be established independent of the prior state of the law.
[34] The Supreme Court has made this point clear on several occasions, most recently in R. v. Sharma, 2022 SCC 39, [2022] 3 S.C.R. 147. There the court said “Parliament is not bound by its past policy choices … legislation must be assessed on its own to determine whether it is constitutionally compliant, without having regard to the prior legislative scheme”: at para. 61, citing Quebec (Attorney General) v. Alliance du personnel professionnel et technique de la santé et des services sociaux, 2018 SCC 17, [2018] 1 S.C.R. 464, at para. 33. The Charter “does not bind the legislature to its current policies”: Sharma, at para. 63. It is an “established constitutional principle that Parliament cannot, through ordinary legislation, fetter itself by inhibiting future legislative action”: Quebec v. Canada, at para. 25, citing Reference re Securities Act, 2011 SCC 66, [2011] 3 S.C.R. 837, at para. 119.
[35] Legislation is, after all, an experiment: it may or may not work, as intended or at all. It may work well for a time, but circumstances may change; legislation may outlive its usefulness and require significant reform to remain relevant. But ultimately, there is no question that it may be amended or repealed by a future legislature.
[36] Thus, rights, obligations, interests, and privileges can be created by ordinary legislation and altered or removed by the same means. This is a bedrock principle of parliamentary sovereignty, a principle that was not altered by passage of the Charter. Whether legislation is consistent with the Charter is not determined having regard to the state of the law prior to passage of that legislation. The prior legislation is irrelevant to the constitutionality of the amending or repealing legislation.
There is no Charter right to bicycle lanes
[37] The claim before the application judge was that bicycle lanes could not be removed to alleviate traffic congestion without violating s. 7 of the Charter. This, according to the application judge, was not a positive rights claim; it was, instead, a request to strike down legislation that put people at risk by removing a safety feature on roads.
[38] With respect, this distinction is not tenable. An order under the Charter prohibiting removal of bicycle lanes can only be understood as establishing a Charter right to those bicycle lanes. It is logically ineluctable: if the Charter prevents the removal of bicycle lanes because of the protection they are thought to afford, it can only be because there was a constitutional right to that protection – and hence, the bicycle lanes – in the first place.
[39] Thus, despite the application judge’s denial, an order prohibiting removal of the bicycle lanes would ensconce the prior legislation. In effect, it would hold the legislature to a constitutional baseline created by prior legislation, denying the authority of the legislature to change policy course – the very approach that, as we have seen, has long been rejected by the Supreme Court. In Toronto (City) v. Ontario (Attorney General), 2021 SCC 34, [2021] 2 S.C.R. 845, at para. 30, the court characterized a claim to restoration of a prior legislative framework – in that case, a 47-ward electoral structure – as a “straightforward positive claim”, describing the prior legislative framework as being “of no legal significance”. Citing Baier v. Alberta, 2007 SCC 31, [2007] 2 S.C.R. 673, at para. 36, the court equated a claim to restore the status quo with a demand to legislate a framework for the first time, adding that this approach was necessary in order to avoid fettering the legislature’s law-making authority.
[40] The application judge acknowledged the appellants’ concern that granting the application would turn many traffic issues into questions of constitutional law but dismissed it as “speculative” – a “floodgates concern”, he said, that is based on circumstances that “rarely arise”:
Road design and traffic engineering decisions are, as one would expect, decisions that are based on data and “highway engineering” advice. … [I]t is safe to conclude that most traffic and road design decisions are not made arbitrarily nor do they increase the risk of harm.
[41] The suggestion here is that there is no reason to be concerned about the constitutionalizing of traffic regulation that the application judge’s decision entails; Charter challenges to legislative change will normally fail because traffic regulation decisions will be made based on data and highway engineering advice.
[42] But this answer begs the question about the legitimate scope of constitutional judicial review. It suggests that the constitutionality of legislation regulating traffic depends ultimately on whether the legislation is made in accordance with expert advice, and as I have said, there is no such requirement. No regulatory legislation would be safe from a finding of unconstitutionality if there were.
[43] There is no Charter right to bicycle lanes – neither specifically nor as inference from any of the rights specifically enumerated in the Charter. The existence of bicycle lanes depends on legislation alone. Thus, their removal does not effect a deprivation within the meaning of s. 7, and the application judge erred in concluding otherwise.
- The removal of bicycle lanes does not violate the principles of fundamental justice
[44] Even if removal of the bicycle lanes were properly understood as having effected a deprivation of life or security of the person, any such deprivation would be in accordance with the principles of fundamental justice.
[45] The principles of fundamental justice are found in “the basic tenets of our legal system”: Re B.C. Motor Vehicle Act, 1985 CanLII 81 (SCC), [1985] 2 S.C.R. 486, at p. 503. Lamer J. (as he then was) went on to say that the principles of fundamental justice “do not lie in the realm of general public policy but in the inherent domain of the judiciary as guardian of the justice system.” The principles of fundamental justice are, in other words, legal principles that underlie the operation of a fair and just legal system – basic values inherent in the Canadian constitutional order.
[46] The removal of bicycle lanes does not conflict with these principles.
[47] I will discuss the nature of arbitrariness and gross disproportionality before applying them in this case.
a. The nature of arbitrariness
[48] It is common to impugn decisions with which one disagrees as arbitrary, but the force of this criticism depends on what is meant by arbitrariness. At its worst, arbitrariness connotes autocratic, capricious rule divorced from reason – in short, a wholesale departure from the most basic requirements of the rule of law. The principles of fundamental justice are a bulwark against this sort of arbitrariness in the context of executive action, but their protection is unlikely to be required against legislation: Canadian legislatures rarely, if ever, pass legislation that is arbitrary in this sense. They may well pass legislation that engenders significant political opposition, but the law is not arbitrary on this account.
[49] The rule against arbitrariness has been described as a principle of “instrumental rationality”, a concept that is concerned not with the normative desirability of the legislation’s purpose but with the means chosen to achieve that purpose: see e.g., Canada (Attorney General) v. Bedford, 2013 SCC 72, [2013] 3 S.C.R. 1101, at para. 107; Canadian Council for Refugees v. Canada (Citizenship and Immigration), 2023 SCC 17, 481 D.L.R. (4th) 581, at para. 124 (“CCR”). Instrumental rationality entails a “means-ends assessment”: CCR, at para. 124.
[50] This idea is sometimes expressed in terms of the relationship between purpose and effect, or purpose and deprivation. In Carter v. Canada (Attorney General), 2015 SCC 5, [2015] 1 S.C.R. 331, at para. 83, the court stated:
The principle of fundamental justice that forbids arbitrariness targets the situation where there is no rational connection between the object of the law and the limit it imposes on life, liberty or security of the person: Bedford, at para. 111. An arbitrary law is one that is not capable of fulfilling its objectives. It exacts a constitutional price in terms of rights, without furthering the public good that is said to be the object of the law.
[51] However, it is clear from the case law that a law is not arbitrary if there is a rational connection between the purpose of the law and the means chosen to achieve that purpose. Thus, in Carter, at para. 84, the court concluded that legislation prohibiting assisted suicide was not arbitrary:
The object of the prohibition on physician-assisted dying is to protect the vulnerable from ending their life in times of weakness. A total ban on assisted suicide clearly helps achieve this object. Therefore, individuals’ rights are not limited arbitrarily.
[52] Arbitrariness speaks to the absence of any connection between the purpose of a law and the means chosen to achieve that purpose: Bedford, at paras. 111, 119; Carter, at para. 83. As the court put it in Bedford, at para. 119, this standard is “not easily met”. After all, legislatures do not pass laws for no reason. And having chosen to legislate, they are unlikely to do so in a manner that is not rationally capable of at least contributing to achieving their purpose.
[53] I emphasize this point in order to highlight what it means to conclude that legislation is arbitrary. It means that the legislature has taken steps that do not rationally contribute to achieving its goal at all – not even in a small way. The implication of this conclusion is that the legislature has failed utterly in its task.
[54] A failure of rationality goes far beyond mere disagreement as to the efficacy of the law. As the court states in Bedford, arbitrariness “do[es] not look to how well the law achieves its object”: at para. 123. Disagreement concerning the efficacy of a law may rise to the level of a “legitimate ongoing debate”, but this does not establish irrationality in the requisite sense: Malmo-Levine, at paras. 135-36, 178.
[55] Disagreement and debate invariably accompany changes in legislative policy and must be allowed to play out in the political process. It is not for the court to end the debate by declaring a winner under the banner of arbitrariness. “A legislative policy is not ‘arbitrary’ just because [the court] may disagree with it”: Chaoulli v. Quebec (Attorney General), 2005 SCC 35, [2005] 1 S.C.R. 791, at para. 169, per Binnie & LeBel JJ.
[56] In most cases, instrumental rationality can be established without difficulty. Common sense connections and inferences will normally suffice. The test for arbitrariness under s. 7 is in this way similar to the test for rationality under s. 1 of the Charter. As the court stated in Alberta v. Hutterian Brethren of Wilson Colony, 2009 SCC 37, [2009] 2 S.C.R. 567, in discussing the rational connection requirement under the Oakes test, at para. 48, “[t]he rational connection requirement is aimed at preventing limits being imposed on rights arbitrarily. The government must show that it is reasonable to suppose that the limit may further the goal, not that it will do so.” See also R. v. Lucas, 1998 CanLII 815 (SCC), [1998] 1 S.C.R. 439, at para. 53, citing RJR‑MacDonald Inc. v. Canada (Attorney General), 1995 CanLII 64 (SCC), [1995] 3 S.C.R. 199, at p. 352. Expert evidence may well call into doubt the efficacy of legislation, but as Carter and Malmo-Levine demonstrate, it will rarely establish that legislation is irrational. As noted above, in Carter, the common-sense inference that a criminal prohibition on physician-assisted suicide helped to protect the vulnerable from ending their lives in times of weakness was sufficient to establish a rational connection between purpose and means: at para. 84. And in Malmo-Levine, Parliament’s common-sense conclusion that prohibiting marijuana would deter its use was logical despite evidence that the extent to which a prohibition would deter usage was unclear: at paras. 135-36, 176-78. As the court noted, it has “exercised caution in accepting arguments about the alleged ineffectiveness of legal measures”, and while different considerations may be relevant in Charter analysis, “it remains important that some deference be accorded to Parliament in assessing the utility of its chosen responses to perceived social ills”: at para. 177.
[57] In summary, instrumental rationality does not establish a constitutional means to challenge the wisdom or efficacy of legislation. Law is not arbitrary simply because it is not designed to the satisfaction of expert critics.
The removal of bicycle lanesis not arbitrary
[58] Section 7 analysis proceeds from identification of the purpose of the impugned provisions: Sharma, at para. 87; R. v. Kloubakov, 2025 SCC 25, 450 C.C.C. (3d) 34, at para. 59. The application judge described the ostensible purpose of the law as the reduction of traffic congestion. The sense in which this purpose is “ostensible” is not clear. The parties do not contest the purpose of the law, which must be “taken at face value” for the s. 7 analysis: Bedford, at para. 125. The respondents argue that the appellants have sought to recharacterize the purpose of the legislation on appeal, but nothing hinges on any alleged imprecision of the purpose identified.
[59] The question, then, is this: is there a rational connection between the purpose of reducing traffic congestion and the decision to increase the number of lanes for motor vehicles by removing bicycle lanes? The respondents were required to establish that there was no such connection. The appellants were not required to establish that the legislation will prove effective in reducing traffic congestion.
[60] Common sense provides the answer here: increasing the number of lanes available for motor vehicles must, at least to some extent, and for some period of time, be capable of contributing to reducing traffic congestion. The legislation is therefore rational. Assumptions grounded in everyday experience and common sense need not be established empirically: Gosselin v. Québec (Attorney General), 2002 SCC 84, [2002] 4 S.C.R. 429, at para. 56.
[61] The application judge acknowledged that “[i]ncreasing the number of lanes available to motor vehicles may allow more cars to travel on a road” and that the respondents’ expert “recognized the possibility of a short-term benefit” from the removal of the bicycle lanes. But instead of accepting that a rational connection had been established and leaving the policy debate about the efficacy of the law to the political process, he went on to consider the expert’s predictions of what would happen to motor vehicle traffic “over time”. Based on these predictions, the application judge found that the law would not contribute to achieving its purpose and would instead achieve the opposite result – an increase in traffic congestion.
In reaching these conclusions, the application judge relied on evidence proffered by Dr. Shoshanna Saxe, whom he described as an expert on transportation infrastructure, including cycling. Dr. Saxe is a proponent of cycling. Her evidence, and that of the respondents’ experts more generally, was impugned by the appellants on this basis. I accept that Dr. Saxe was not biased in a legal sense but repeat what this court said in Fair Voting BC v. Canada (Attorney General), 2025 ONCA 581, 177 O.R. (3d) 401, at paras. 19-20, leave to appeal refused, [2025] S.C.C.A. No. 384: academic expert evidence – social science research in particular – must be approached by courts with considerable caution. It is often geared towards supporting preferred policy outcomes rather than discovering the objective truth of a matter under investigation. And courts are prone to missing the normative and policy dimensions of issues if they present in purely technical terms, ostensibly resolvable on an objective and uncontroversial basis by technical experts.
[62] Dr. Saxe clearly favours bicycles. That was clear from the tenor of her evidence, which included the assertion that “investing in cycling infrastructure is one of the most powerful tools available to reduce congestion.” As the application judge noted, Dr. Saxe conceded that the removal of bicycle lanes might provide a benefit in the short term. But she expressed the opinion that, over the medium to long term, removing the bicycle lanes would make traffic congestion worse. She predicted that emergency response “will likely be worse” and that “gridlock becomes a possibility”.
[63] Dr. Saxe’s evidence was based on the economic concept of “induced demand”, which the application judge described as “well-accepted”. Induced demand is a theory that permits predictions to be made, all things being equal. But predictions may or may not prove to be sound. Induced demand theory offers no legitimate basis to countermand the legislative policy concerning road use – no legitimate basis to conclude that the legislation is irrational in the face of the obvious logical connection between the number of lanes on a road and the volume of traffic. The legislature may well be wrong; in time, the phenomenon of induced demand may operate to undermine the effects of the legislation as Dr. Saxe predicted. But many additional factors are and will continue to be relevant to traffic congestion: the price of gasoline; the availability of carpool lanes; and the availability of parking and parking costs, to name just a few. Then too, there are variables relevant to traffic congestion unrelated to road use and policy, including the ability of employees to work from home. No doubt there are many more relevant considerations.
[64] But whether the legislature turns out to be right or wrong, in the medium or long term, induced demand is no more than a prediction as to what might occur. Dr. Saxe’s evidence was not capable, in law, of establishing that the legislation was arbitrary. Her prediction that, over time, things would get worse, is concerned with the efficacy of the legislation. It does not establish a lack of a rational connection.
[65] The application judge erred by focusing on concerns about the efficacy of the legislation, rather than the rational connection between the purpose of the legislation and the means chosen by the legislature. The Charter does not require the legislature to act in a manner the court considers optimal. It does not require the legislature to legislate in accordance with predictions, expert or otherwise. If a rational connection exists between the legislature’s purpose and means, that is the end of the court’s inquiry. The wisdom and effectiveness of the legislature’s decision – the efficacy of the legislation – is for the political process.
[66] The application judge’s error in focusing on efficacy rather than rationality is clear from his discussion of the evidence and his criticism of the government. There was, he stated, “no evidence that [the government] based its decision on data, manuals or expert ‘highway engineering’, or that its decision would ‘contribute to highway safety.’” On the contrary, he stated, the law was passed “in the face of internal advice and public representations, including from organizations with highway engineering expertise, that the removal of the target bike lanes would make those roads less safe and would not reduce congestion.”
[67] But legislative policy is not arbitrary just because some or even many advisors, experts, and citizens predict that it will not work: Malmo-Levine, at paras. 135-36, 176-78; Chaoulli, at para. 169, per Binnie & LeBel JJ. This point is fundamental: the government is free to accept some, all, or none of the advice it receives. The legislature could have chosen to legislate in accordance with the views of Dr. Saxe had it so desired. But it was under no constitutional obligation to do so. The principles of fundamental justice neither require nor permit courts to countermand rational policy choices legitimately made by the government and legislature, regardless of the preferences of academic experts.
[68] In summary, the application judge conflated instrumental rationality and efficacy and erred in concluding that the legislation was arbitrary. Legislation is not unconstitutional for failing to satisfy the preferences of experts whose advice the government was not required to accept. The question for the court was whether there was a rational connection between the legislature’s purpose of reducing traffic congestion and the decision to increase the number of lanes for motor vehicles. This was easily established and the application judge erred in concluding otherwise.
b. The nature of gross disproportionality
[69] The rule against gross disproportionality is not the same as a rule against mere disproportionality: R. v. Safarzadeh‑Markhali, 2016 SCC 14, [2016] 1 S.C.R. 180, at paras. 67-73. The decision to recognize the rule against gross disproportionality as a principle of fundamental justice reflects an intention to elevate the standard considerably – well beyond the possibility of reasonable disagreement inherent in questions of mere disproportionality – in order to preserve “‘broad latitude … for legislative action’”: Bedford, at para. 103, quoting Malmo-Levine, at para. 175. Purported disproportionality short of that high threshold is a matter for public and legislative debate rather than judicial intervention, even where the law’s effects are serious: Malmo-Levine, at para. 175. This reinforces the idea that the principles of fundamental justice are basic legal principles inherent in the Canadian constitutional order – not instruments authorizing courts to declare winners and losers in public policy debates.
[70] Thus, in Bedford, at para. 120, the Supreme Court described gross disproportionality as follows:
Gross disproportionality asks a different question from arbitrariness and overbreadth. It targets the second fundamental evil: the law’s effects on life, liberty or security of the person are so grossly disproportionate to its purposes that they cannot rationally be supported. The rule against gross disproportionality only applies in extreme cases where the seriousness of the deprivation is totally out of sync with the objective of the measure. This idea is captured by the hypothetical of a law with the purpose of keeping the streets clean that imposes a sentence of life imprisonment for spitting on the sidewalk. The connection between the draconian impact of the law and its object must be entirely outside the norms accepted in our free and democratic society.
[71] Several things are remarkable in this passage in the court’s unanimous decision. Gross disproportionality is understood as a relationship between a law’s purpose and its effects that “cannot rationally be supported”. It is devoid of reason.
[72] This is a standard that should rarely be met. As I said in discussing arbitrariness, legislatures do not act for no reason. And having decided that it is necessary to act, they are unlikely to legislate so poorly that the law they create is devoid of reason. The Supreme Court emphasizes as much when it says that gross disproportionality arises only in “extreme” cases – cases in which “the seriousness of the deprivation is totally out of sync with the objective of the measure” (emphasis added). Such cases should be obvious – not even remotely arguable, as the court’s hypothetical example demonstrates. Gross disproportionality, says the court, is demonstrated by a law designed to keep the streets clean that imposes a sentence of life imprisonment for spitting on the sidewalk. And in Bedford itself, the court described the disparity grounding its finding of gross disproportionality in similarly extreme terms – a law that had exposed its subjects to a high risk of being murdered by a notorious serial killer merely to prevent nuisances associated with street prostitution: at paras. 135-36, 158.
[73] Almost nothing is like these examples. It is not surprising, therefore, that the Supreme Court has, since Bedford, repeatedly declined opportunities to declare deprivations of s. 7-protected rights to be grossly disproportionate, even in cases like Carter, where it perceived the law’s effects to be “severe”: at para. 90. And again, as further evidence of how high the standard is, in Bedford the court refers to laws that have a “draconian impact” – laws that are “entirely outside the norms accepted in our free and democratic society”: at para. 120. To say that gross disproportionality establishes an extremely high standard is an understatement.
The removal of bicycle lanes is not grossly disproportionate
[74] The application judge cites Bedford and purports to acknowledge the high standard it establishes. Yet in short order he concludes that high standard is met by the removal of bicycle lanes:
[T]he evidence is clear that restoring lanes of motor vehicle traffic by removing the bike lanes, or reconfiguring them so that they are no longer physically separated from cars, will lead to more accidents, injuries and deaths. … It is reasonable to conclude that people who cycle will be injured and killed when lanes for motor vehicles are installed and protected bike lanes are removed. That impact is grossly disproportionate to the asserted benefit, taken at face value, of saving some drivers of cars a few minutes of travel time.
[75] There is much that is wrong with this reasoning. It minimizes the purpose of the legislation, in effect trivializing it by characterizing it as saving some drivers a few minutes rather than reducing traffic congestion, which may in turn result in a wide range of positive effects. At the same time, it inflates the negative effects of removing the bicycle lanes, talking of injuries and deaths that will occur only if the bicyclists choose to ride on the roads that have no bicycle lanes – a situation that obtains on roads throughout the province. Having so characterized the purpose and effects of the legislation, the application judge rendered the outcome inevitable on a simple proportionality standard.
[76] But that is not the standard he was required to apply. The absurd hypothetical the court used in Bedford to demonstrate gross disproportionality and the extreme facts of Bedford itself illustrate this point. Removing bicycle lanes is in no way analogous to imposing a sentence of life imprisonment for spitting on the sidewalk or exposing citizens to a high risk of being murdered to prevent nuisances. It is mere traffic regulation and cannot be described as “draconian”.
[77] The application judge’s references to concerns raised by intervener groups do not support the conclusion that the law is grossly disproportionate. The assertion that the impact of removing bicycle lanes “will disproportionately fall on low-income and disabled individuals, increasing health inequities already experienced by these groups” may, if true, be relevant to the policy argument that the legislation is unwise or ill-conceived, but it is insufficient to establish that the deprivation the law is supposed to effect is grossly disproportionate to its purpose. The same is true of the assertion that removal of the bicycle lanes has a disproportionate effect on children. There was no basis for the submission that the legislation “put[s] children’s lives at risk in pursuit of folly”, and it was wrong for the application judge to cite this submission with apparent approval.
[78] That is sufficient to dispose of the appeal. There is no need to address the application of s. 1. In saying this, I am not to be taken to be in agreement with the application judge’s s. 1 analysis.
DISPOSITION
[79] I would allow the appeal.
[80] The parties may make costs submissions up to 5 pages in length, not including their cost outlines, within 21 days of the release of these reasons. No costs are awarded to or against the interveners.
Released: August 14, 2026 “J.M.F.”
“Grant Huscroft J.A.”
“I agree. Fairburn A.C.J.O.”
“I agree. B. Zarnett J.A.”
1The new s. 195.6 also added an additional bicycle lane to be reconfigured: the Avenue Road bicycle lane.