R. v. Parranto, 2021 SCC 46, [2021] 3 S.C.R. 366

I. Overview

[ 1 ] In these appeals, the appellants ask this Court to abolish the starting‑point method of sentencing. Like sentencing ranges, starting points are a form of appellate guidance that provide a place to begin the exercise of reaching a fit and proportionate sentence.

[ 2 ] The appellants, Patrick Douglas Felix and Cameron O’Lynn Parranto, pleaded guilty to various offences arising out of unrelated drug trafficking operations, including fentanyl trafficking contrary to s. 5(1) and s. 5(2) of the Controlled Drugs and Substances Act , S.C. 1996, c. 19 (“ CDSA ”). Both appellants were found to be operating at the “wholesale” commercial level. At sentencing, Mr. Felix received a global 7‑year sentence ( 2019 ABQB 183 ), and Mr. Parranto received a global sentence of 11 years ( 2018 ABQB 863 ). The Crown appealed the sentences to the Court of Appeal of Alberta, where a five‑member panel heard the appeals jointly for the express purpose of setting a “starting point” for wholesale fentanyl trafficking. In separate decisions released concurrently, the Court of Appeal set a 9‑year starting point for wholesale fentanyl trafficking and increased Mr. Felix’s sentence to 10 years ( 2019 ABCA 458 , 98 Alta. L.R. (6th) 136), and Mr. Parranto’s to 14 years ( 2019 ABCA 457 , 98 Alta. L.R. (6th) 114).

[ 3 ] The appellants and several interveners sought to discredit the starting‑point approach by arguing that it has undesirable results, including higher rates of incarceration for Indigenous and other offenders. These criticisms lose their force, however, if starting points are properly treated as non‑binding guidance by both sentencing and appellate courts. Further, such criticisms speak to the risks inherent in using any form of quantitative sentencing guidance, including sentencing ranges. But these risks can be avoided if appellate courts adhere to the deferential standard of review in sentencing appeals, and if this Court provides clear direction on how appellate courts should account for starting points when reviewing sentences for errors in principle and demonstrable unfitness. What follows, therefore, is not an endorsement of starting points as they have sometimes been enforced at the Court of Appeal of Alberta, but rather a revised understanding, bringing them into conformance with the standard of appellate review and principles and objectives of sentencing.

[ 4 ] Accordingly, there is no need to disavow the starting‑point approach to sentencing. Sentencing ranges and starting points are simply different tools that assist sentencing judges in reaching a proportionate sentence. It is not for this Court to dictate which of these tools can or cannot be used by appellate courts across the country. Provincial appellate courts should be afforded the respect and latitude to provide their own forms of guidance to sentencing judges, as long as that guidance comports with the principles and objectives of sentencing under the Criminal Code , R.S.C. 1985, c. C‑46, and with the proper appellate standard of review.

[ 5 ] Nor would we interfere with the sentences imposed upon Mr. Felix and Mr. Parranto at the Court of Appeal. In our respectful view, the sentences imposed on these offenders by the respective sentencing judges were demonstrably unfit. The Court of Appeal’s intervention was therefore appropriate.

[ 6 ] Therefore, and for the reasons that follow, we would dismiss the appeals, affirm the orders of the Court of Appeal, and confirm that starting points, when properly understood and applied, represent a valid form of sentencing guidance.

II. Parties’ Submissions

[ 7 ] The appellants submit that the Court of Appeal erred in intervening to impose its own views over those of the sentencing judge in each case. The foundation of this error, they say, is the starting‑point approach to sentencing. The appellants raise several criticisms of the starting‑point approach and argue that sentencing ranges are a preferable means of providing quantitative guidance to sentencing judges. They submit that, in light of these concerns, this Court must abolish starting‑point sentences.

[ 8 ] The Crown, on the other hand, submits that there is no reason to abolish starting‑point sentences. Starting points are merely guidelines and functionally no different from sentencing ranges. The Crown argues that the Court of Appeal correctly intervened as the sentences imposed at first instance were demonstrably unfit and, in the case of Mr. Parranto, the trial judge’s sentencing decision contained errors in principle that affected the sentence.

III. Analysis

A. Standard of Review

(1) Basic Sentencing Principles

[ 9 ] This Court has repeatedly expressed that sentencing is “one of the most delicate stages of the criminal justice process in Canada” ( R. v. Lacasse , 2015 SCC 64 , [2015] 3 S.C.R. 1089, at para. 1 ). More of an art than a science, sentencing requires judges to consider and balance a multiplicity of factors. While the sentencing process is governed by the clearly defined objectives and principles in Part XXIII of the Criminal Code , it remains a discretionary exercise for sentencing courts in balancing all relevant factors to meet the basic objectives of sentencing ( Lacasse , at para. 1 ).

[ 10 ] The goal in every case is a fair, fit and principled sanction. Proportionality is the organizing principle in reaching this goal. Unlike other principles of sentencing set out in the Criminal Code , proportionality stands alone following the heading “Fundamental principle” ( s. 718.1 ). Accordingly, “[a]ll sentencing starts with the principle that sentences must be proportionate to the gravity of the offence and the degree of responsibility of the offender” ( R. v. Friesen , 2020 SCC 9 , at para. 30 ). The principles of parity and individualization, while important, are secondary principles.

[ 11 ] Despite what would appear to be an inherent tension among these sentencing principles, this Court explained in Friesen that parity and proportionality are not at odds with each other. To impose the same sentence on unlike cases furthers neither principle, while consistent application of proportionality will result in parity (para. 32). This is because parity, as an expression of proportionality, will assist courts in fixing on a proportionate sentence (para. 32). Courts cannot arrive at a proportionate sentence based solely on first principles, but rather must “calibrate the demands of proportionality by reference to the sentences imposed in other cases” (para. 33).

[ 12 ] As to the relationship of individualization to proportionality and parity, this Court in Lacasse aptly observed:

Proportionality is determined both on an individual basis, that is, in relation to the accused him or herself and to the offence committed by the accused, and by comparison with sentences imposed for similar offences committed in similar circumstances. [para. 53]

Individualization is central to the proportionality assessment. Whereas the gravity of a particular offence may be relatively constant, each offence is “committed in unique circumstances by an offender with a unique profile” (para. 58). This is why proportionality sometimes demands a sentence that has never been imposed in the past for a similar offence. The question is always whether the sentence reflects the gravity of the offence, the offender’s degree of responsibility and the unique circumstances of each case (para. 58).

(2) Role of Sentencing Courts and Appellate Courts

[ 13 ] Appellate and sentencing courts each have distinct roles in sentencing, based on the Code and their respective institutional capacities. In s. 718.3(1), Parliament has “explicitly vested sentencing judges with a discretion to determine the appropriate degree and kind of punishment under the Criminal Code ” ( R. v. M. (C.A.) , [1996] 1 S.C.R. 500, at para. 90 (emphasis deleted)). Sentencing courts are best‑positioned to craft a fit sentence for the offenders before them. Sentencing is a “profoundly subjective process”, and the sentencing judge “has the advantage of having seen and heard all of the witnesses whereas the appellate court can only base itself upon a written record” ( R. v. Shropshire , [1995] 4 S.C.R. 227, at para. 46 ). The sentencing judge also has “unique qualifications of experience and judgment from having served on the front lines” and “will normally preside near or within the community which has suffered the consequences of the offender’s crime” ( M. (C.A.) , at para. 91 ).

[ 14 ] In sentencing appeals, provincial appellate courts play two roles “in ensuring the consistency, stability and permanence of the case law” ( Lacasse , at para. 36 ). First, the court of appeal must “consider the fitness of the sentence appealed against” and is empowered to vary the sentence ( Code , s. 687). In fulfilling this role, appellate courts safeguard against errors of law made by sentencing courts while reviewing the reasonableness of the judge’s exercise of discretion. They also ensure that sentencing courts state the law correctly and apply it consistently ( Lacasse , at para. 36 ).

[ 15 ] Secondly, provincial appellate courts must promote stability in the development of the law while providing guidance to lower courts to ensure the law is applied consistently in a particular jurisdiction ( Lacasse , at para. 37 ). In carrying out this role, appellate courts may provide guidance to assist sentencing judges in reaching a proportionate sentence that properly balances parity and individualization (para. 2). Appellate courts are well‑positioned to provide such guidance because of their appreciation of “overall sentencing practices, patterns and problems” in their jurisdiction ( R. v. Arcand , 2010 ABCA 363 , 40 Alta. L.R. (5th) 199, at para. 153 ). Appellate guidance may take the form of quantitative tools (such as sentencing ranges and starting points), non‑quantitative guidance explaining the harms entailed by certain offences, or a mix of both (see, e.g., Friesen , at paras. 42‑105 and 114‑16 ; R. v. Williams , 2019 BCCA 295 , at paras. 64‑66 and 71 ; R. v. Sandercock (1985), 1985 ABCA 218 , 40 Alta. L.R. (2d) 265 (C.A.), at pp. 270-71).

(3) Forms of Quantitative Appellate Guidance

[ 16 ] Quantitative appellate guidance generally takes one of two forms: starting points, or sentencing ranges. These tools are best understood as “navigational buoys” that operate to ensure sentences reflect the sentencing principles prescribed in the Criminal Code . Busy sentencing judges face a challenging task; the Code often provides for a wide range of possible sentences and the factual circumstances of each case vary infinitely. Sentencing must begin somewhere, and both starting‑point and range methodologies assist sentencing judges by providing a place to start in the form of either a single number or a range. As this Court has recognized, however, “there is no such thing as a uniform sentence for a particular crime” ( M. (C.A.) , at para. 92 ). Neither tool relieves the sentencing judge from conducting an individualized analysis taking into account all relevant factors and sentencing principles.

[ 17 ] Sentencing ranges generally represent a summary of the case law that reflects the minimum and maximum sentences imposed by trial judges in the past ( Lacasse , at para. 57 ; Friesen , at para. 36 ). They “provide structure and guidance and can prevent disparity”, while leaving judges space to “weigh mitigating and aggravating factors and arrive at proportional sentences” ( R. v. Smith , 2019 SKCA 100 , 382 C.C.C. (3d) 455, at para. 126 ). The range, therefore, “reflects individual cases, but does not govern them” (C. C. Ruby, Sentencing (10th ed. 2020), at § 23.7, citing R. v. Brennan and Jensen (1975) , 11 N.S.R. (2d) 84 (C.A.)).

[ 18 ] Contrary to the comments of the Court of Appeal in Arcand , at para. 148 , this Court has subsequently and expressly held that starting points are an “alternative” to ranges, with “[s]imilar principles” applying to both ( Friesen , at para. 36 ). The starting‑point methodology has three stages: (1) defining the category of an offence to which the starting point applies; (2) setting a starting point; and (3) individualization of the sentence by the sentencing court ( Sandercock , at p. 269). Both ranges and starting points are attached to “a category of offences that share enough common features that it is useful to judge them by the same rubric” ( Friesen , at para. 39 ). At the second stage, the court of appeal must choose a “reasonable” number which “may be viewed as the mid‑point in the traditional range of sentences for a particular sort of crime” ( R. v. McDonnell , [1997] 1 S.C.R. 948, at para. 60 , per McLachlin J.; contra , Arcand , at paras. 150‑56 ). At the final stage, the sentencing judge considers whether aggravating or mitigating factors justify moving the sentence up or down from the starting point ( McDonnell , at paras. 79‑80 ).

[ 19 ] The quantitative guidance set by appellate courts can be distinguished from informal or “discerned” ranges relied on by sentencing courts. Since starting points and appellate sentencing ranges — which may be described as “formal” or “established” ranges — do not exist for every offence, sentencing judges must often attempt to discern the appropriate range from the case law with the assistance of counsel. As with appellate guidance, discerned ranges are a tool for reaching a fit and proportionate sentence. After hearing the parties’ submissions on analogous case law and on the appropriate disposition in the circumstances, the exercise of discerning a range requires the sentencing judge to consider comparator cases that best represent the gravity of the offence and the circumstances of the offender.

[ 20 ] Starting points and sentencing ranges set by appellate courts, on the other hand, often reflect judicial consensus on the gravity of the offence, helping to advance parity and “prevent any substantial and marked disparities” in sentencing ( Lacasse , at para. 2 ; see also R. v. Stone , [1999] 2 S.C.R. 290, at para. 244 ; R. v. Nasogaluak , 2010 SCC 6 , [2010] 1 S.C.R. 206, at para. 44 ; Smith (2019), at para. 126). The primary rationale for the use of starting points is to “reduce idiosyncratic decision‑making” ( Arcand , at para. 102 ; R. v. Hajar , 2016 ABCA 222 , 39 Alta. L.R. (6th) 209, at para. 72 ; R. v. McCowan , 2010 MBCA 45 , 251 Man. R. (2d) 295, at para. 11 ; R. v. Lemaigre , 2018 SKCA 47 , at para. 20 ). In Sandercock , for instance, a three‑year starting point was established to address the “wide, and unjustified, divergence amongst judges about appropriate sentences for rape and equally serious forms of sexual assault” ( Arcand , at para. 102 ). The starting point stated in Sandercock was intended to ensure that the harms caused by a particular category of offence — “major sexual assaults” — are consistently accounted for in sentencing.

[ 21 ] Quantitative appellate guidance — whether in the form of a range or a starting point — may draw on sentences imposed in past cases, or it may be intended to change the existing approach to sentencing for a particular offence. As we have explained, sentencing ranges generally represent a “historical portrait” of past decisions. This portrait reflects “all the principles and objective of sentencing” ( Lacasse , at para. 57 ) as well as the “collective wisdom of the appellate courts” (M. A. Crystal, “Are the Days of Range Sentencing and Starting Points Numbered? The Cases of R. v. Felix and R. v. Parranto ” (2021), 15 J.P.P.L. 125, at p. 139).

[ 22 ] In some cases, however,

an appellate court must also set a new direction, bringing the law into harmony with a new societal understanding of the gravity of certain offences or the degree of responsibility of certain offenders ( R. v. Stone , [1999] 2 S.C.R. 290, at para. 239 ). . . . as a general rule, appellate courts should take the lead in such circumstances and give sentencing judges the tools to depart from past precedents and craft fit sentences. [Emphasis added.]

( Friesen , at para. 35 )

This reflects the reality that “[i]t is a common phenomenon . . . for sentences to increase and decrease as societal and judicial knowledge and attitudes about certain offences change” ( R. v. Smith , 2017 BCCA 112 , at para. 36 (emphasis deleted), citing R. v. Nur , 2011 ONSC 4874 , 275 C.C.C. (3d) 330, at para. 49 ).

[ 23 ] The same applies to starting points. While some jurisprudence from the Court of Appeal of Alberta suggests that starting points are established through an independent policy‑laden inquiry rather than by looking to precedent (e.g., Arcand , at para. 104 ), an appellate court may derive a starting point in whole or in part from past sentencing decisions. It may also choose to depart from past trends to recalibrate how the gravity of the offence is weighed in the proportionality analysis. Like established ranges, then, starting points may reflect “collective court experience” by drawing on an overview of the case law for a range of sentences imposed, but they may also represent a departure based on a “consensus view of [all] the social values and policy considerations relating to the category of crime in question” ( Arcand , at para. 104 ).

[ 24 ] While each jurisdiction tends to prefer one of these two methodologies, describing a jurisdiction as either a “range jurisdiction” or “starting‑point jurisdiction” represents an incomplete view of appellate guidance. Contrary to the references contained in submissions from many parties on these appeals, the division between “range jurisdictions” and “starting‑point jurisdictions” cannot be described in absolute or dichotomous terms. Even in so‑called “starting‑point jurisdictions”, appellate courts have established starting points only for a limited number of offences, with other forms of guidance — including established and discerned sentencing ranges — used to reach a proportionate sentence for other offences. This flexibility in sentencing modalities also applies in “range jurisdictions”. In practice, courts that have either rejected or failed “to fully endorse the starting point” approach have, in effect, adopted something similar to the starting‑point methodology by either setting sentencing ranges without a top end ( Smith (2017)), or by incorporating mitigating factors such as prior good character into the sentencing range, an approach usually associated with starting points ( R. v. H. (C.N.) (2002) , 62 O.R. (3d) 564 (C.A.), at para. 52 ; R. v. Voong , 2015 BCCA 285 , 374 B.C.A.C. 166; R. v. Cunningham (1996) , 27 O.R. (3d) 786 (C.A.), at p. 790; see also P. Moreau, “In Defence of Starting Point Sentencing” (2016), 63 Crim. L.Q. 345, at pp. 356 and 365‑66).

[ 25 ] This flexibility in sentencing modalities is reflective of the primary goal of sentencing: to impose a fit sentence. In our view, irrespective of the preferred sentencing methodology, the purpose of the modality is to assist the sentencing judge in achieving the objectives and principles of sentencing, primarily proportionality. Ranges and starting points are simply different paths to the same destination: a proportionate sentence. Courts of appeal have discretion to choose which form of guidance they find most useful and responsive to the perceived needs of their jurisdiction, which may vary across the country. As long as that guidance conforms to the principles and objectives of sentencing in the Code , this Court should respect the choices made by appellate courts. Both sentencing ranges and starting points, where properly applied and subject to the correct standard of review on appeal, are consistent with the Code . It is not this Court’s role to decide which form of guidance is superior, nor would it be desirable to confine appellate courts to giving one or another form of quantitative guidance.

[ 26 ] A final point that merits discussion is how a range or starting point is modified once it has been established. The Court of Appeal for Ontario has described how ranges vary:

“Ranges” are not embedded in stone. Given their nature as guidelines only, I do not view them as being fixed in law, as is the case with binding legal principles. They may be altered deliberately, after careful consideration, by the courts. Or, they may be altered practically, as a consequence of a series of decisions made by the courts which have that effect. If a range moves by virtue of the application of individual cases over time, it is not necessary to overrule an earlier range that may once have been in vogue; it is only necessary to recognize that the courts have adapted and the guidelines have changed.

( R. v. Wright (2006) , 83 O.R. (3d) 427 (C.A.), at para. 22 )

[ 27 ] In contrast, in Alberta, the Court of Appeal has required parties to formally apply for reconsideration of a starting point under r. 16.27 of the Alberta Rules of Court , A.R. 124/2010. Our colleague Rowe J. views this procedure as a judicial mandate and an expression of the Court of Appeal’s “aggressive enforcement” of starting points (para. 164). The reconsideration procedure, however, is a legislative requirement of general application ( Alberta Rules of Court , r. 16.27). The process is a judicial one whereby the Court of Appeal weighs several criteria in deciding whether to “reconsider” the starting point, such as whether the starting point is new or old, has been disapproved of or is contrary to decisions of other courts of appeal, was created by overlooking binding statute or authority, contains some “simple, obvious, demonstrable flaw”, or was established in Reasons for Judgment Reserved or a Memorandum of Judgment ( Arcand , at para. 199 ). With the exception of the final criterion, all of these considerations are relevant in determining whether the starting point for a particular offence should be recalibrated.

[ 28 ] While our colleague finds it difficult to reconcile the reconsideration procedure with the standard of review, the answer, in light of this Court’s direction on the standard of review and the non‑binding nature of starting points, is clear. Parties seeking to challenge a starting point need not have resort to the procedure, because starting points are not binding precedents. If the procedure is used, the effect of a successful reconsideration application is merely to re‑establish the point from which trial judges begin their thinking. Such applications provide the Court of Appeal with the opportunity to consider the sentencing landscape to determine if the starting point still provides relevant appellate guidance and accurately reflects the gravity of the offence. To the extent the reconsideration procedure is perceived to be necessary to change starting points because they are binding precedents ( Arcand , at para. 199 ; R. v. Melnyk , 2014 ABCA 313 , 580 A.R. 389, at para. 2 ), this perception cannot survive Lacasse and Friesen , as it is mistaken in law.

(4) Basis for Appellate Intervention

[ 29 ] It is trite law that appellate courts cannot interfere with sentencing decisions lightly (see R. v. Suter , 2018 SCC 34 , [2018] 2 S.C.R. 496, at para. 23 , citing Shropshire , at para. 48 ; R. v. L.F.W. , 2000 SCC 6 , [2000] 1 S.C.R. 132, at para. 25 ; R. v. L.M. , 2008 SCC 31 , [2008] 2 S.C.R. 163, at para. 14 ; Nasogaluak , at para. 46 ; Lacasse , at para. 39 ; and Friesen , at para. 25 ). Sentencing judges are to be afforded wide latitude, and their decisions are entitled to a high level of deference on appeal ( Lacasse , at para. 11 ). It remains the case that, where a judge deviates from a sentencing range or starting point, no matter the degree of deviation, this does not in itself justify appellate intervention.

[ 30 ] It bears emphasizing that the sentencing judge’s discretion includes the choice of a sentencing range or of a category within a range, and that this exercise of discretion cannot in itself constitute a reviewable error ( Lacasse , at para. 51 ). It is an error of law for an appellate court to intervene merely on the ground that it would have placed the offence in a different range or category. Unless a sentence is demonstrably unfit or the sentencing judge made an error in principle that impacts the sentence, an appellate court must not vary the sentence on appeal (paras. 11 and 67). The focus of the demonstrable unfitness inquiry is on whether the sentence is proportionate, not whether the sentencing judge applied the correct starting point, sentencing range or category within a range ( Lacasse , at paras. 51 and 53 ; Friesen , at para. 162 ).

[ 31 ] Following the recent judgments of this Court in Lacasse and Friesen , we are of the view that these appeals do not require the Court to chart a new path, but rather requires us to reiterate and reinforce the standard for appellate intervention. In particular, the Court must clarify the role that sentencing ranges or starting points play in appellate review of sentences.

[ 32 ] The relationship between quantitative appellate guidance and the standard of review on appeal had led to exchanges between the various levels of court. But Lacasse and Friesen have brought finality to the matter, and they leave no room for doubt, interpretation or equivocation. In Friesen , this Court made it clear that “[r]anges of sentence and starting points cannot be binding in either theory or practice , and appellate courts cannot interpret or apply the standard of review to enforce them, contrary to [ Arcand ], at paras. 116‑18 and 273 ” (para. 37 (emphasis added)). The Attorney General of Alberta intervened on these appeals and argued that this passage is open to interpretation, as the paragraphs it cites from Arcand closely mirror this Court’s guidance in Lacasse . In Lacasse , at para. 67 , this Court stated that “a deviation from such a range or category is not an error in principle and cannot in itself automatically justify appellate intervention unless the sentence that is imposed departs significantly and for no reason from the contemplated sentences ” (emphasis added).

[ 33 ] While we agree that the Court of Appeal’s comments in Arcand do closely mirror this Court’s direction in Lacasse , there is an important distinction between Arcand and this Court’s sentencing jurisprudence. The clear direction from this Court is that “[a]ppellate courts cannot treat the departure from or the failure to refer to a range of sentence . . . as an error in principle” ( Friesen , at para. 37 (emphasis added)). The Court of Appeal in Arcand , however, suggested the contrary, stating “that sentencing judges will give due consideration to those starting points and the process that starting point sentencing entails” (at para. 273 (emphasis added)). To resolve any possible ambiguity that could have survived Friesen , we say this: the directions relating to the binding nature of starting points or the starting‑point approach as set out in Arcand have been overtaken by Lacasse and Friesen and no longer reflect the required standard of appellate review.

[ 34 ] While the goal of the starting‑point methodology has been framed as “uniformity of approach” ( Arcand , at para. 92 ; R. v. Johnas (1982), 1982 ABCA 331 , 41 A.R. 183 (C.A.), para. 31 ), the standard of review limits its function in this regard. Certainly, it is open to appellate courts to provide guidance to assist sentencing judges in minimizing idiosyncrasies and to promote consistency in approaches to sentencing. As R. v. Ipeelee , 2012 SCC 13 , [2012] 1 S.C.R. 433, and Friesen make clear, however, there is no one uniform approach to sentencing in Canada. Attempts to create a single uniform approach are therefore misguided. Different cases may require different methods, and selecting the method of sentencing is within the discretion of the sentencing judge.

[ 35 ] Moreover, it is not the role of appellate courts to enforce a uniform approach to sentencing through the application of the standard of review. Rather, the proper focus of appellate review is whether the sentence was fit and whether the judge properly applied the principles of sentencing. To be clear, those principles do not include an obligation to follow a particular uniform approach to sentencing. While promoting consistency in method of sentencing may have a role to play in setting appellate guidance, at the stage of appellate review, focusing on uniformity of approach is apt to mislead the reviewing court. Appellate courts must guard against undue scrutiny of the sentencing judge’s discretionary choice of method, as this may stray from the standard of review.

[ 36 ] The key principles are as follows:

1. Starting points and ranges are not and cannot be binding in theory or in practice ( Friesen , at para. 36 );

2. Ranges and starting points are “guidelines, not hard and fast rules”, and a “departure from or failure to refer to a range of sentence or starting point” cannot be treated as an error in principle ( Friesen , at para. 37 );

3. Sentencing judges have discretion to “individualize sentencing both in method and outcome”, and “[d]ifferent methods may even be required to account properly for relevant systemic and background factors” ( Friesen , at para. 38 , citing Ipeelee , at para. 59 ); and,

4. Appellate courts cannot “intervene simply because the sentence is different from the sentence that would have been reached had the range of sentence or starting point been applied” ( Friesen , at para. 37 ). The focus should be on whether the sentence was fit and whether the judge properly applied the principles of sentencing, not whether the judge chose the right starting point or category ( Friesen , at para. 162 ).

These principles settle the matter. Contrary to the Crown’s submission, it is not an open question whether sentencing judges are free to reject the starting‑point approach. Sentencing judges retain discretion to individualize their approach to sentencing “[f]or this offence, committed by this offender, harming this victim, in this community” ( R. v. Gladue , [1999] 1 S.C.R. 688, at para. 80 (emphasis in original)). There is no longer space to interpret starting points (or ranges) as binding in any sense .

[ 37 ] Having made clear that starting points and ranges are tools and not straitjackets, we turn to the role that a sentencing range or starting point plays in appellate sentence review.

[ 38 ] As we have already stated, sentencing is an individualized process, and parity is secondary to proportionality. Therefore, departures from the starting point or sentences above or below the range are to be expected. Even significant departures are not to be treated as a prima facie indication of an error or demonstrable unfitness. Fitness is assessed with reference to the principles and objectives of sentencing in the Code , not with reference to how far the sentence departs from quantitative appellate guidance.

[ 39 ] Ultimately, the sentencing judge’s reasons and the record must allow the appellate court to discern whether the sentence is fit in light of the principles and objectives in the Code . Section 726.2 requires the sentencing court to provide reasons for the sentence imposed. This is not a new standard in criminal law. In sentencing, as with any decision, the reasons must, when read in conjunction with the record, show why the judge reached a particular result.

[ 40 ] Most particularly, and irrespective of the modality used, the trial judge’s reasons and the record must demonstrate why the sentence is proportionate to the moral blameworthiness of the offender and the gravity of the offence. This may involve adopting appellate guidance such as this Court’s comments in Friesen on the harms of the offence. Quantitative appellate guidance may also be part of the jurisprudential background informing the gravity of the offence. As noted, it is not an error in principle for the sentencing judge to fail to refer to a starting point. Since starting points and ranges reflect the gravity of the offence, however, the sentencing judge’s reasons and the record must allow the reviewing court to understand why the sentence is proportionate despite a significant departure from the range or starting point. This applies regardless of whether the reasons refer to the starting point or not. At the very least, the appellate court must be able to discern from the reasons and the record why the sentence is fit in the circumstances of the offence and the offender. We emphasize, however, that it is inappropriate for appellate courts to “artificially constrain sentencing judges’ ability to impose a proportionate sentence” by requiring “exceptional circumstances” when departing from a range ( Friesen , at paras. 111‑12 ; R. v. Burnett , 2017 MBCA 122 , 358 C.C.C. (3d) 123, at para. 26 ). Departing from a range or starting point is appropriate where required to achieve proportionality.

[ 41 ] In Arcand , the Court of Appeal asked whether “starting point sentencing has a meaningful function in Canada” (para. 116). The answer is that it does, but that function is not to bind trial judges or to licence boundless appellate intervention. Quantitative sentencing tools need not be binding to provide useful guidance to busy sentencing courts. Starting‑point sentences, like ranges, assist sentencing judges in their day-to-day work by providing a frame of reference and a shorthand for judicial views on the gravity of the offence. This guidance is especially important where Parliament has left open the possibility of a wide range of sentences ( R. v. Proulx , 2000 SCC 5 , [2000] 1 S.C.R. 61, at para. 86 ).

B. Starting Points

[ 42 ] The Court heard many criticisms of the starting‑point approach to sentencing in the course of these appeals. So long, however, as appellate courts apply the appropriate standard of review to sentencing decisions, many of these criticisms fall away. Further, any risks associated with starting‑point sentencing are inherent in other forms of quantitative appellate guidance, including sentencing ranges. Such risks therefore do not serve to compel the abolition of starting‑point sentencing; they do, however, demonstrate the importance of individualizing sentences and applying the proper appellate standard of review, irrespective of the form of quantitative guidance used.

[ 43 ] Moreover, a caricatured version of the starting‑point approach as it is pronounced in Arcand should not be relied on to reject starting points altogether. Focussing on Arcand fails to account for both subsequent developments in the law in Alberta (see, e.g., R. v. Stewart , 2021 ABCA 79 , 21 Alta. L.R. (7th) 213; R. v. Gandour , 2018 ABCA 238 , 73 Alta. L.R. (6th) 26, at para. 55 ) and jurisprudence in other jurisdictions that use starting points. In Manitoba, for instance, the Court of Appeal has been clear that starting points are simply tools or guidelines, “not rigid tariffs that fetter the discretion of a sentencing judge to impose an individualized sentence” ( Burnett , at para. 10 ; see also R. v. Sidwell , 2015 MBCA 56 , 319 Man. R. (2d) 144, at para. 50 ). Other jurisprudence thus shows that the starting‑point approach can be compatible with the principles of sentencing and the standard of review on appeal. The starting‑point methodology ought not to be rejected wholesale by focusing on pre‑ Friesen cases decided at a time in which the Court of Appeal of Alberta still spoke of starting points as binding. As we have made clear, and as the Court of Appeal of Alberta has acknowledged in Stewart and Gandour , starting points are not binding and the jurisprudence of this Court prevails over any case law that suggests otherwise.

[ 44 ] While not binding, however, sentencing ranges and starting points are useful tools because they convey to sentencing judges an appreciation of the gravity of the offence. And, as we have already observed, they offer judges a place to begin their thinking. When applying these tools, sentencing judges must individualize the sentence in a way that accounts for both aspects of proportionality: the gravity of the offence and the offender’s individual circumstances and moral culpability. At the stage of individualizing the sentence, the sentencing judge must therefore consider “all of the relevant factors and circumstances, including the status and life experiences, of the person standing before them” ( Ipeelee , at para. 75 ). Those factors and circumstances may well justify a significant downward or upward adjustment in the sentence imposed.

[ 45 ] Starting points also do not relieve the sentencing judge from considering all relevant sentencing principles. The principles of denunciation and deterrence are generally built into starting points and reflected in ranges, but those objectives “cannot be allowed to obliterate and render nugatory or impotent other relevant sentencing objectives” ( R. v. Okimaw , 2016 ABCA 246 , 340 C.C.C. (3d) 225, at para. 90 ). When conducting an individualized analysis, sentencing judges are expected to account for other relevant sentencing objectives, including rehabilitation and restraint. Indeed, this Court has held that the 1996 sentencing reforms were intended to both ensure courts consider restorative justice principles and to address the problem of over‑incarceration in Canada ( Gladue , at para. 57 ; Proulx , at paras. 16‑20 ). Sentencing judges have discretion over which objectives to prioritize ( Nasogaluak , at para. 43 ; Lacasse , at para. 54 ) and may choose to weigh rehabilitation and other objectives more heavily than “built-in” objectives like denunciation and deterrence. Appellate courts should not lose sight of these principles — nor the deferential standard of review — when reviewing sentences that depart from a starting point or range.

[ 46 ] One objection to starting points advanced in these appeals is that they can easily harden into de facto minimums because they build in the mitigating factor of prior good character, thus preventing sentencing judges from relying on that factor to justify a downward departure (A. Manson, The Law of Sentencing (2001), at p. 72; R. v. Kain , 2004 ABCA 127 , 35 Alta. L.R. (4th) 5, at para. 32 , per Berger J.A.). But the same is sometimes true of ranges ( Cunningham ; Voong ). Importantly, neither tool incorporates other potential mitigating circumstances or Gladue factors, nor should it. Appellate sentencing guidance ought not to purport to pre‑weigh or “build‑in” any mitigating factors ( contra , Arcand , at para. 135 ). Likewise, since starting points are intended to reflect the gravity of the offence and the resulting need for deterrence and denunciation, they should not be viewed as incorporating sentencing principles such as restraint or rehabilitation, contrary to the suggestion in Arcand , at para. 293 . Where sentencing judges choose to refer to the starting point or range, they are not precluded from considering any factor that is “built in” as mitigating in the individual circumstances, and they retain the discretion to consider and weigh all relevant factors in their global assessment of a fit sanction. This comports with the principle that the sentencing judge must always consider all relevant individual circumstances in reaching a fit sentence tailored to the offender before the court.

[ 47 ] Moreover, when setting starting points and ranges, appellate courts must be mindful of what is built into those forms of guidance. Any inclusion of characteristics of an archetypal offender could impede individualization of sentences, rendering the guidance inconsistent with the standard of review ( M. (C.A.) , at para. 90 ; Nasogaluak , at para. 43 ) and with Parliament’s express choice to vest sentencing judges with the discretion to determine a fit sanction ( Code , s. 718.3(1)). While this Court’s role is not to dictate how provincial appellate courts should establish sentencing ranges and starting points, we emphasize that those tools are applicable only inasmuch as they solely speak to the gravity of the offence. By restricting starting points and ranges to strictly offence‑based considerations, they will continue to be useful to sentencing judges without fettering their discretion or impeding individualization in a way that could produce clustering of sentences.

[ 48 ] While we heard arguments regarding the clustering effect starting points may have on sentences, we were provided with no empirical data demonstrating this effect. Further, the literature relied on by the parties in support of this psychological “anchoring effect” relates to sentencing guidelines in other jurisdictions that involve ranges (M. W. Bennett, “Confronting Cognitive ‘Anchoring Effect’ and ‘Blind Spot’ Biases in Federal Sentencing: A Modest Solution for Reforming a Fundamental Flaw” (2014), 104 J. Crim. L. & Criminology 489; I. D. Marder and J. Pina‑Sánchez, “Nudge the judge? Theorizing the interaction between heuristics, sentencing guidelines and sentence clustering” (2020), 20 C.C.J. 399). This suggests that concerns about clustering apply equally when other forms of quantitative appellate guidance are used, not solely in the context of starting points. In any event, any risk of clustering — beyond what is appropriate in seeking a proportionate sentence that reflects the principle of parity — is properly addressed by ensuring sentencing judges consider all factors relevant to each individual offender, and by clarifying the proper standard of review on appeal. Further, knowledge of this risk will go far in mitigating the potential negative impacts of clustering by ensuring sentencing courts are on guard against it.

[ 49 ] We also heard submissions that starting points impede sentencing judges in applying the principles in Gladue to Indigenous offenders. As noted, sentencing judges have the authority to adopt a different and individualized method of sentencing to reach a fit sentence for Indigenous offenders ( Ipeelee , at para. 59 ). This does not mean that Gladue and starting‑point methodologies are mutually incompatible. When applied properly, starting points do not prevent judges from giving effect to s. 718.2(e) and the principles in Gladue , as demonstrated by R. v. Skani , 2002 ABQB 1097 , 331 A.R. 50, a case involving the three‑year starting point from Johnas , and R. v. Paul , 2016 ABPC 113 , a case engaging the eight‑year starting point from R. v. Matwiy (1996), 1996 ABCA 63 , 178 A.R. 356 (C.A.). In Paul , for example, Gladue factors played a central role in assessing the moral culpability of the offender, ultimately justifying a downward departure from the eight‑year starting point to a sentence of five years (para. 56). These examples show that starting‑point jurisdictions have been able to integrate Gladue principles into the starting‑point methodology (see also R. v. Beardy , 2017 MBPC 32 , at paras. 9 , 12 and 16 , aff’d 2018 MBCA 52 ).

[ 50 ] When reviewing sentences imposed on Indigenous offenders, appellate courts must bear in mind that a “formalistic approach to parity” should not be allowed to “undermine the remedial purpose of s. 718.2( e )” ( Ipeelee , at para. 79 ). Whether starting points or ranges are used, sentencing judges are under an obligation to factor in the “unique circumstances of an Aboriginal offender which could reasonably and justifiably impact on the sentence imposed” (para. 72; see also R. v. Park , 2016 MBCA 107 , 343 C.C.C. 3(d) 347, at para. 24). As recognized by the Court of Appeal of Alberta in R. v. Swampy , 2017 ABCA 134 , 50 Alta. L.R. (6th) 240, at para. 36 , “[t]here can be no sound proportionality analysis in the case of an Aboriginal offender without considering the impact of the offender’s Aboriginal heritage on his moral culpability.” The Court of Appeal has also stated that it is an error in principle for a sentencing judge to refer to Gladue factors without considering the impact of those factors on moral blameworthiness ( R. v. Bird , 2021 ABCA 243 , at para. 20 ) .

[ 51 ] Another argument advanced before us is that starting points are inconsistent with s. 718.2(d) of the Code , which demands that trial judges consider sentences other than imprisonment in appropriate cases ( R. v. Drake (1997) , 151 Nfld. & P.E.I.R. 220 (P.E.I.S.C. (App. Div.)), at para. 5). Nothing about the nature of starting points or ranges compels this result. But this Court was clear in Lacasse that courts may have few options other than imprisonment where general or specific deterrence and denunciation must be emphasized, as in this case ( Lacasse , at para. 6 ). As we have discussed, ranges and starting points are properly understood as a reflection of the gravity of the offence. The reality is that the gravity of some offences may effectively remove the option of non‑custodial sentences.

[ 52 ] Further, in cases involving Indigenous offenders, starting points do not relieve sentencing judges from considering whether “different or alternative sanctions may more effectively achieve the objectives of sentencing in a particular community” ( Ipeelee , at para. 74 ). For example, in Skani , the starting‑point sentence of three years’ imprisonment was reduced to a sentence of 23 months served in the community, taking into account “the perspective of the aboriginal offender’s community” (para. 66). As with every offender, but especially Indigenous offenders, consideration of different sanctions reflects the second arm of the proportionality assessment. That is, while the gravity of the offence may dictate a custodial sentence, the offender’s individual situation must also support the sentence imposed.

[ 53 ] A final concern raised by these appeals is that starting‑point sentencing, through the use of judicially created categories, is a quasi‑legislative endeavour. Judicially created categories for sentencing are not, however, unique to starting points; both the range and starting‑point approaches rely on categorization of offences to ensure parity ( Lacasse , at paras. 2 and 51 ; Arcand , at para. 93 ). This Court has recognized that courts may use categories to situate conduct along a continuum for the purposes of sentencing ( Lacasse , at para. 67 ). Categorization can make sentencing more manageable, since many offences under the Code cover a wide range of conduct and may attract a broad range of sentences ( McDonnell , at para. 85 ). Categorization assists sentencing judges by breaking down a single offence under the Code based on factors such as the type of conduct at issue, the circumstances in which it was committed, and the consequences for the victim or community ( Arcand , at para. 95 ). Appellate courts are entitled to conclude that certain forms of conduct are generally more serious and thus should attract a higher range or starting point.

[ 54 ] The risk of incursion into the legislative sphere arises only where an appellate court departs from the standard of review by treating a sentencing judge’s failure to select the “correct” category as an error in principle. As always, the reviewing court is limited to questioning whether there is an error in principle that impacted the sentence or whether the sentence is demonstrably unfit. In some cases, mischaracterization of the offence will result in a sentence that is demonstrably unfit, but only if selecting the “wrong” category led the trial judge to misapprehend the gravity of the offence. An appellate court cannot conclude that the sentence is unfit simply because the judge failed to adhere to a judicially created category or range. Further, as in Friesen , a sentencing judge may make a “reasoned choice” — based on individualized factors — to place an offence in a sub‑category even where the established criteria for that sub‑category do not apply.

IV. Application: Sentences on Appeal

A. Setting a Starting Point

[ 55 ] In our view, the Court of Appeal did not err in setting a starting point for wholesale fentanyl trafficking. Relying heavily on para. 57 of Lacasse , in which this Court stated that sentencing ranges are “nothing more than summaries of the minimum and maximum sentences imposed in the past”, Mr. Felix argues that there was insufficient case law on sentencing for fentanyl trafficking to allow the Court of Appeal to create a “prospective” starting point (A.F. (F.), at paras. 86‑87). Relatedly, Mr. Felix urges this Court to reject the creation of a starting point in this case on the basis that it is beyond an appellate court’s purview to determine the gravity of an offence without a “broad jurisprudential base” (A.F. (F.), at para. 100). In effect, Mr. Felix takes the position that there must be a sufficient “historical portrait” before an appellate court may set a range or a starting point.

[ 56 ] It is true that appellate courts often ground quantitative guidance in judicial experience and past decisions. As noted in Friesen , at para. 33 , “[p]recedents embody the collective experience and wisdom of the judiciary.” It is unnecessary, however, for a provincial appellate court to wait for the development of a “historical portrait” of past sentences to give guidance on the gravity of the offence. As noted, this Court has affirmed that appellate courts must sometimes “set a new direction” that reflects a contemporary understanding of the gravity of the offence ( Friesen , at para. 35 ).

[ 57 ] It follows that provincial appellate courts are not restricted to providing guidance only where a historical body of precedent has developed; they may also revise the sentencing landscape. Just as the Court in Friesen provided guidance to ensure sentencing judges would appreciate the harm of sexual offences against children, it was open to the Court of Appeal in this case to set out guidance conveying the gravity of wholesale trafficking in fentanyl. Regardless of the form of guidance at issue, appellate courts may consider the harms involved in a given offence and conclude that further guidance is needed to ensure lower courts appreciate the gravity of the offence. Accepting the appellant’s position would not only be contrary to this Court’s direction in Friesen , it would result in a perverse waiting game in which the court of appeal must delay providing quantitative guidance — such as a starting point or appropriate range reflecting the harmfulness of a new drug — until there are sufficient sentencing decisions (and sufficient instances of harm to victims and to society) to be considered a “historical portrait”.

[ 58 ] Moreover, the core offence in this case — drug trafficking — is not new. As the Crown points out, “[d]rug trafficking as an offence is easily quantifiable by reference to a variety of independent factors such as volume of drugs, price, and level of commerciality” (R.F., at para. 92). Another key factor in the categorization of drug offences, both in relation to criminality and sentencing, has always been the nature of the drug at issue. The composition and dangers of the drugs trafficked may change quickly. As the harms caused by the substance speak directly to the gravity of the offence, appellate courts may step in to provide guidance to ensure sentences reflect those harms, even where the drug is relatively new. We underscore the importance of this fact because harm‑based analyses are not an unfamiliar judicial exercise in the sentencing context ( Friesen , at para. 114 ).

[ 59 ] Further, the Court of Appeal was entitled to take the lead and consider the public health crisis in Alberta in the creation of the nine‑year starting point. It is noteworthy that Alberta has one of the highest rates of opioid‑related deaths and overdoses, relative to other provinces and territories (L. Belzak and J. Halverson, “The opioid crisis in Canada: a national perspective” (2018), 38 H.P.C.D.P.C . 224). As Lamer C.J. stated in M. (C.A.) , at para. 91 , a just and appropriate sentence may take into consideration “the needs and current conditions of and in the community”. Local conditions may enter into the assessment of the gravity of the offence and militate in favour of prioritizing certain sentencing goals ( Lacasse , at paras. 13 and 89 ). We stress that other jurisdictions are free to establish ranges or starting points that differ from that in Alberta, as any sentencing guidance should strive to reflect and be responsive to the local conditions in those jurisdictions.

[ 60 ] In addition to Mr. Felix’s concerns, Mr. Parranto urges the Court to reject the creation of a starting point in this case on the basis that there “is no evidence to support a contention that the [opioid] crisis was created or caused by overly lenient sentences for drug traffickers” (A.F. (P.), at para. 44). Even if “criminal justice responses alone cannot solve the problem”, however, the courts must use the tools Parliament has provided to address societal ills ( Friesen , at para. 45 ). Parliament has chosen to employ the mechanisms of criminal law and sentencing law to advance public safety, hold those who distribute drugs accountable, and communicate the wrongfulness of poisoning people and communities. This is perhaps most apparent in the maximum sentence for trafficking in a Schedule I drug, which is life in prison ( CDSA , s. 5(3)(a)). As stated in Friesen , “[m]aximum penalties are one of Parliament’s principal tools to determine the gravity of the offence” (para. 96, citing C. C. Ruby et al., Sentencing (9th ed. 2017), at § 2.18; R. v. Sanatkar (1981) , 64 C.C.C. (2d) 325 (Ont. C.A.), at p. 327; Hajar , at para. 75 ).

[ 61 ] We reiterate that the nine‑year starting point is just “one tool among others that [is] intended to aid trial judges in their work” ( Lacasse , at para. 69 ). Sentencing judges are free to depart from the starting point and move up or down from this marker based on the specific characteristics of the offender in order to meet the primary sentencing principle of proportionality.

B. Mr. Felix

[ 62 ] Mr. Felix entered guilty pleas to two counts of trafficking in fentanyl and two counts of trafficking in cocaine, both contrary to s. 5(1) of the CDSA . He entered his guilty pleas after exhausting his Charter challenges related to these offences. The following is a summary of the pertinent facts derived from an Agreed Statement of Facts.

[ 63 ] Mr. Felix was the directing mind of a drug trafficking operation in Fort McMurray, Alberta, that included selling drugs destined for Nunavut (sentencing reasons (Felix), at para. 15 ). It was structured as a “dial‑a‑dope” operation with “runners” filling orders from the stash location and remitting the proceeds to a “boss”, who then, in turn, remitted the money to Mr. Felix weekly. The structure functioned to insulate Mr. Felix from criminal exposure.

[ 64 ] The fentanyl trafficking convictions relate to five completed transactions and one incomplete transaction. The five completed transactions involved a total of 1398 fentanyl tablets and 19.75 ounces of cocaine sold for $76,000 (sentencing reasons (Felix), at para. 20). The one incomplete transaction involved 987 fentanyl tablets and 1974 g (69.63 ounces) of cocaine. In Fort McMurray in 2015, the 987 fentanyl tablets had an approximate street value of between $107,000 and $214,000 if sold by the tablet.

[ 65 ] The sentencing decision describes Mr. Felix, at the time of the offence, as a 34‑year‑old college‑educated individual with no prior criminal record and a successful owner‑operator of a non‑drug related business. Mr. Felix had no physical or mental impairments, had no relevant addiction or substance abuse issues, and provided the Court with a positive presentence report indicating that, among other things, he maintains a good parental relationship with his children. Mr. Felix also provided 17 letters of reference speaking to his good character.

[ 66 ] After considering all of these factors, the sentencing judge imposed a seven‑year global sentence: seven years on both counts of trafficking fentanyl to be served concurrently, and four years for each count of trafficking in cocaine to be served concurrently. The Crown appealed. The Court of Appeal found the sentence demonstrably unfit and indicated that it would have imposed a 13‑year sentence on each of the two counts of trafficking in fentanyl to run concurrently. In recognition of the fact that the Crown sought a 10‑year sentence at the sentencing hearing and that the parties’ sentencing positions were taken when the jurisprudence was still evolving, the Court of Appeal imposed a global 10‑year term of incarceration.

[ 67 ] In our view, the Court of Appeal correctly intervened. We agree that the seven‑year sentence imposed at first instance was demonstrably unfit. It is clear the sentencing judge misapprehended the gravity of the offence. After reviewing a selection of cases, the sentencing judge concluded the range in Alberta was five to seven years, while the range was five to nine and one‑half years if other Canadian jurisdictions were taken into account. We agree with the Court of Appeal that the Alberta cases referred to by the sentencing judge were significantly factually distinct from this case.

[ 68 ] A more accurate range based on a review of reported case law nationally would be in the region of 8 to 15 years. For example, eight‑year sentences were imposed in Smith (2019) (1834 pills, as part of an 11‑year sentence), as well as R. v. Leach , 2019 BCCA 451 (11,727 pills, as part of a 16‑year sentence); R. v. Sinclair , 2016 ONCA 683 ; R. v. Solano‑Santana , 2018 ONSC 3345 (5000 pills); R. v. White , 2020 NSCA 33 , 387 C.C.C. (3d) 106 (2086 pills); and R. v. Borris , 2017 NBQB 253 (4200 pills). Other sentences imposed include: an 8‑year and two‑month sentence in R. v. Sidhu , C.J. Ontario, No. 17‑821, June 16, 2017, aff’d 2019 ONCA 880 , in which the offender trafficked 89 g of fentanyl and other drugs soon after being released on parole; a 10‑year sentence in R. v. Petrowski , 2020 MBCA 78 , 393 C.C.C. (3d) 102, for trafficking 51 g fentanyl where the offender used a co‑accused to insulate himself from detection; 11 years for trafficking 204.5 g of a fentanyl blend in R. v. Vezina , 2017 ONCJ 775 ; 13 years for trafficking 232 g fentanyl and large quantities of other drugs as part of a sophisticated drug trafficking operation in R. v. Mai , [2017] O.J. No. 7248 (QL) (Ont. S.C.J.) ; and 15 years for a profit‑motivated offender who was the directing mind of “a large‑scale drug trafficking operation involving an enormous amount of fentanyl” in R. v. Fuller , 2019 ONCJ 643 (the offender possessed about 3 kg of fentanyl in the course of the conspiracy).

[ 69 ] The sentencing judge’s error with respect to the range ultimately impacted his assessment of parity. It is clear that seven years is a demonstrably unfit sentence given the gravity of this offence and sentences imposed in other cases. Indeed, cases from Alberta show a range of five to seven years for offenders who are engaged in “commercial trafficking [of fentanyl] on more than a minimal scale”, in much smaller quantities and with less sophistication than Mr. Felix ( R. v. M.M.A. , 2018 ABQB 250 , at para. 21 ; R. v. Adams , 2018 ABPC 82 ). The sentence imposed on Mr. Felix by the sentencing judge was therefore a “substantial and marked departure from the sentences customarily imposed for similar offenders committing similar crimes” ( M. (C.A.) , para. 92 ).

[ 70 ] While not raised by the parties or the court below, this appeal provides an opportunity to emphasize that, when assessing the gravity of the offence, it is open to both the sentencing judge and the Court of Appeal to take into account the offender’s willingness to exploit at‑risk populations and communities. In this regard, choices which demonstrate a reckless disregard for human life increase not only the gravity of the offence but the moral culpability of the offender and may amount to an aggravating factor in sentencing.

[ 71 ] While all people and places merit protection, sentencing judges may, as they consider appropriate, give special consideration to the disproportionate harm caused to particularly vulnerable groups and/or vulnerable and remote locations, where escaping traffickers is more difficult and resources for combating addiction are more sparse. Here, for example, Mr. Felix was trafficking fentanyl destined for resale in the remote communities comprising the territory of Nunavut. As an outsider, he chose to traffic drugs to those vulnerable communities for easy money. It would have been open to the courts below to consider this as a significantly aggravating factor. Indeed, the Supreme Court of the Northwest Territories, which would “have front‑line experience and understand the needs of the community where the crime was committed” (Rowe J.’s reasons, at para. 121), has specifically denounced this sort of predatory conduct:

It has been said repeatedly, but bears repeating again, trafficking in cocaine has had a devastating effect on the people in Yellowknife and elsewhere in the Northwest Territories. . . .

Those that traffic in cocaine contribute directly to this. They prey on the most vulnerable members of the community for profit. And there are those who come to this jurisdiction simply to traffic in drugs because it is lucrative. There is easy money to be made off the addiction of others. [Emphasis added].

( R. v. Dube , 2017 NWTSC 77 , at pp. 12-13 )

Similarly, in Ontario, trafficking fentanyl to vulnerable northern communities has been found to be an aggravating factor ( Solano‑Santana , at para. 28 ). Accordingly, the objective harm caused by outsiders who engage in wholesale fentanyl trafficking to vulnerable communities may amount to an aggravating circumstance that carries with it the expectation that a sentence will be aimed at holding the offender accountable and communicating the wrongfulness of the behaviour.

[ 72 ] We do agree, however, that Mr. Felix has strong rehabilitative prospects. He has engaged in exclusively prosocial pursuits following his arrest, has demonstrated a commitment to change, and has the clear support of many friends and family.

[ 73 ] Accordingly, the sentence of 10 years imposed by the Court of Appeal should be upheld, given the submissions at trial, a review of the case law and the aggravating and mitigating circumstances. In upholding the decision of the Court of Appeal, we emphasize that the commission of wholesale trafficking offences in fentanyl may very well be expected to attract more significant sentences as the harm to the end user and the devastating consequences to communities plagued by addiction is not contested.

C. Mr. Parranto

[ 74 ] Mr. Parranto entered guilty pleas on two counts of possession of fentanyl for the purposes of trafficking in a wholesale commercial operation ( CDSA , s. 5(2)); two counts of illegal possession of a loaded handgun for use in the trafficking operation ( Criminal Code , s. 95); one count of possession of a handgun, knowing he was prohibited by court order ( Criminal Code , s. 117.01(1)); and breach of a recognizance ( Criminal Code , s. 145(3)). These charges arose from events that occurred on two separate dates: March 24, 2016, and October 22, 2016. What follows is a summary of the relevant facts derived from an Agreed Statement of Facts.

[ 75 ] At the time of his March arrest, Mr. Parranto was under a lifetime firearm prohibition and was bound by a recognizance order prohibiting him from possessing controlled substances and firearms. In March police recovered 27.8 g of fentanyl powder with an approximate street value of $5560 and $55,575 in cash. In October police recovered 485.12 g of fentanyl powder (capable of producing 500,000 individual doses) with an approximate street value of $97,064, along with $20,690 in cash.

[ 76 ] The sentencing judge reviewed all the information before him and calculated a notional global sentence of 20 years, with sentences of seven and eight years allocated to the two fentanyl trafficking counts respectively. The judge reduced the sentence by one‑third for Mr. Parranto’s guilty plea and by a further 1.2 years for “other mitigating circumstances” (sentencing reasons (Parranto), at para. 93 ). He then reduced it by one year based on the totality principle. The resulting period of incarceration was 11 years less time served. The Crown appealed. The Court of Appeal concluded that the sentencing judge made several errors in principle and that the sentence was demonstrably unfit. As a result, it substituted a global sentence of 14 years’ incarceration, less time served.

[ 77 ] We are of the view that the 11‑year global sentence imposed at first instance was demonstrably unfit and Court of Appeal did not abrogate the standard of review in intervening. There is no reason for this Court to disturb the sentence of 14 years imposed by the Court of Appeal.

[ 78 ] Much like in the case of Mr. Felix, the trial judge erred in his selection of comparator cases and in finding that the relevant range was five to seven years’ imprisonment. As described above, the national range for this offence is approximately 8 to 15 years. The sentencing judge referred to the seven‑year sentence in R. v. Aujla , 2016 ABPC 272 , the only reported decision in Alberta referring to “wholesale fentanyl trafficking.” However, Aujla is a poor comparator, as it dealt with smaller quantities of drugs (454 fentanyl pills), a single arrest, and a first‑time offender with no record and good rehabilitative prospects.

[ 79 ] Mr. Parranto was in possession of significant amounts of fentanyl in addition to large quantities of other drugs, guns [1] and body armour. [2] He had a lengthy and related criminal record, and following his release from detention in July 2016 for the March offences, he was able to and did re‑establish his presence as a wholesale trafficker in approximately 12 weeks.

[ 80 ] Against the gravity of the offence and these aggravating factors, this case also calls for consideration of Gladue principles. Despite being Métis, the record indicates that Mr. Parranto waived his right to a Gladue report and did not file a presentence report. Even where a Gladue report is waived, however, courts must “take judicial notice of the systemic and background factors affecting Aboriginal people in Canadian society”, including “such matters as the history of colonialism, displacement, and residential schools” ( Ipeelee , at para. 60 ; Gladue , at para. 83 ). Moreover, counsel made submissions indicating that Mr. Parranto had a disjointed childhood surrounded by drugs, alcohol and abuse. He began using drugs in the 1990s and has struggled with addiction to heroin. It was incumbent on the sentencing judge and the Court of Appeal to consider these circumstances in the context of the “broad systemic and background factors affecting Aboriginal people generally” ( Ipeelee , at paras. 59‑60 ). The offender is not required to show a “causal link between background factors and the commission of the current offence”, and Gladue principles must be applied in every case regardless of the seriousness of the offence ( Ipeelee , at paras. 81 and 87 ). In our view, Mr. Parranto’s background circumstances can be said to have played a part in bringing him before the court. Against this must be weighed the reality that Mr. Parranto committed the second set of offences less than three months after being released on bail for the first set of offences. This suggests that restorative justice principles such as rehabilitation are less salient in this case compared to other objectives including protection of the public.

[ 81 ] Based on the gravity of the offence, Gladue factors and the aggravating and mitigating circumstances, we agree with the Court of Appeal that a global sentence of 14 years is appropriate. While we would not make any finding as to how the 14 years should be apportioned between the various counts, we note that it would have been appropriate to impose the 9‑ and 12‑year sentences sought by the Crown at first instance with respect to the two fentanyl offences, with those sentences to run concurrently. A 12‑year sentence for the second fentanyl trafficking count would send an appropriately strong message that wholesale fentanyl trafficking is a serious offence consistent with a high level of moral culpability. As we concluded with respect to Mr. Felix, it will not be uncommon to see lengthy penitentiary terms imposed for this offence.

[ 82 ] Finally, contrary to Mr. Parranto’s submissions, the Court of Appeal did not intervene on the basis that the sentencing judge failed to apply the starting point of nine years that did not exist at the time Mr. Parranto was sentenced. As discussed above, the Court of Appeal correctly intervened because the sentence imposed was demonstrably unfit. Irrespective of the correctness of the intervention, we agree with the appellants that the Court of Appeal’s comments with respect to starting‑points and the starting‑point approach are incorrect in law (C.A. reasons (Parranto), at paras. 29 and 68; Friesen , at para. 37 ; Lacasse , at para. 60 ). These comments, however, were not the basis of the Court of Appeal’s intervention.

V. Conclusion

[ 83 ] We would dismiss both appeals and affirm the orders of the Court of Appeal of Alberta. In doing so, we confirm the legitimacy of starting points on this revised basis as a permissible form of appellate guidance, within the framework provided by this Court that emphasizes deference to sentencing judges in the delicate task that Parliament has charged them with ( Lacasse ; Friesen ). Just as the law recognizes the soundness of considering local conditions in crafting a fit sentence, there need not be a singular norm in achieving the goals of sentencing. Irrespective of the sentencing modality chosen, provincial appellate courts are best positioned to give the guidance necessary to achieve consistency of both reasoning and approach.

The reasons of Moldaver and Côté JJ. were delivered by

Moldaver J. —

I. Introduction

[ 84 ] I would dismiss the appeals from sentence and uphold the sentences of 10 years and 14 years imposed by the Court of Appeal. The sentences imposed by the sentencing judges in both cases were demonstrably unfit. They fall markedly below the range of sentences that are warranted in cases like this, involving the directing minds of largescale fentanyl trafficking operations. In such cases, more severe sentences than those imposed by the Court of Appeal would have been justified; however, in the circumstances, the Court of Appeal cannot be faulted for failing to impose higher sentences than those sought by the Crown at the sentencing hearings.

[ 85 ] With respect to the role of starting points in sentencing, I agree with my colleague, Rowe J.

[ 86 ] I find it necessary, however, to write separately to raise what I believe to be an issue of overriding concern in these cases. Specifically, I wish to focus on the gravity of largescale trafficking in fentanyl for personal gain and the need to impose severe penalties, ranging from mid‑level double digit penitentiary terms up to and including life imprisonment, for those who do so. In R. v. Friesen , 2020 SCC 9 , this Court held that appellate courts can and should depart from prior sentencing precedents when those precedents no longer reflect “society’s current understanding and awareness of the gravity of a particular offence and blameworthiness of particular offenders” (para. 35; see also paras. 108 and 110). In my view, society’s understanding of the gravity of largescale fentanyl trafficking has increased such that an upward departure is mandated.

II. Analysis

A. The Dangers Posed by Trafficking in Hard Drugs

[ 87 ] The dangers posed by trafficking in hard drugs, such as heroin and cocaine, have long been recognized in Canada. Over the past few decades, however, society’s awareness of the true gravity of trafficking in such drugs has grown to the point that we are reminded, on a daily basis, of the death, destruction, and havoc it causes in communities across Canada.

[ 88 ] Trafficking in such substances causes both direct and indirect harms to society. Directly, the distribution and abuse of hard drugs leads to addiction, debilitating adverse health effects, and, all too frequently, death by overdose. As Lamer J. (as he then was) astutely observed, where addiction and death occur — as they so often do — those who oversee the distribution of these drugs are personally “responsible for the gradual but inexorable degeneration of many of their fellow human beings” ( R. v. Smith , [1987] 1 S.C.R. 1045, at p. 1053).

[ 89 ] Trafficking also leads indirectly to a host of other ills, including an increase in all manner of crime, committed by those seeking to finance their addiction, as well as by organized crime syndicates ( Pushpanathan v. Canada (Minister of Citizenship and Immigration) , [1998] 1 S.C.R. 982, at paras. 85‑87 , per Cory J., dissenting, but not on this point; R. v. Kang‑Brown , 2008 SCC 18 , [2008] 1 S.C.R. 456, at para. 184 , per Deschamps J., dissenting, but not on this point). Given that much of this criminal activity is violent, trafficking has come to be understood as an offence of violence, even beyond the ruinous consequences it has for those who abuse drugs and in the process, destroy themselves and others. Indeed, as Doherty J.A. has explained, violence is such a predictable consequence of the illicit drug trade that it cannot be dissociated from it:

Cocaine sale and use is closely and strongly associated with violent crime. Cocaine importation begets a multiplicity of violent acts. Viewed in isolation from the conduct which inevitably follows the importation of cocaine, the act itself is not a violent one in the strict sense. It cannot, however, be disassociated from its inevitable consequences. [Emphasis added.]

( R. v. Hamilton (2004) , 72 O.R. (3d) 1 (C.A), at para. 104 )

See also, R. v. Pearson , [1992] 3 S.C.R. 665, at pp. 694‑95, where the Court relied on the Groupe de travail sur la lutte contre la drogue, Rapport du groupe de travail sur la lutte contre la drogue (1990), at pp. 18-19, which noted that it is a mistake to view drug trafficking under the control of organized crime as less serious than more openly violent crimes.

[ 90 ] A further and perhaps even more devastating consequence of the hard drug trade is its impact on families and the intergenerational trauma it causes:

Trafficking in drugs, and in particular hard drugs such as cocaine, is a crime whose victims can be found far beyond the individuals who become addicted to the drugs. Families can be torn apart by either the loss of the individual to the addiction itself or to the violence that all too often accompanies the drug trade. . . .

Children suffer immense harm from the effects of addiction in their home, whether this addiction be from pre‑natal impact or from physical and/or emotional violence in the homes that they should be safe in. The future of these children and their families is damaged and all of society pays the price.

( R. v. Profeit , 2009 YKTC 39 , at paras. 25‑26 )

See also M. Barnard, Drug Addiction and Families (2007), at pp. 14‑17, explaining that children whose parents abuse drugs are at an increased risk of physical abuse, emotional abuse, and neglect.

[ 91 ] Finally, the trafficking of hard drugs leads to “significant if not staggering” costs to society in terms of health care and law enforcement expenses, as well as lost productivity ( Pushpanathan , at para. 89 ; see also R. v. Lloyd , 2016 SCC 13 , [2016] 1 S.C.R. 130, at para. 82 ). In 2017, for example, it is estimated that abuse of opioids and cocaine in Canada resulted in total costs of $9.6 billion (Canadian Substance Use Costs and Harms Scientific Working Group, Canadian Substance Use Costs and Harms (2015‑2017) (2020), at p. 1).

[ 92 ] Trafficking in hard drugs is thus a “crime with such grievous consequences that it tears at the very fabric of society” ( Pushpanathan , at para. 79 ). Significant penitentiary sentences are regularly imposed for individuals who traffic in large quantities of such drugs, which, as we shall see, are far less deadly than fentanyl (see, e.g., R. v. Bains , 2015 ONCA 677 , 127 O.R. (3d) 545 (9 years for possession of one kilogram of heroin for the purpose of trafficking); R. v. Athwal , 2017 ONCA 222 (12 years for conspiracy to possess heroin for the purpose of trafficking); R. v. Chukwu , 2016 SKCA 6 , 472 Sask. R. 241 (10 years for the possession of 0.59 kilograms of heroin for the purpose of trafficking); R. v. Dritsas , 2015 MBCA 19 , 315 Man. R. (2d) 205 (9 years for possession of cocaine for the purposes of trafficking for a “high‑level cocaine trafficker involved in kilo‑level deals” (para. 9)).

B. The Dangers Posed by Largescale Fentanyl Trafficking

[ 93 ] As grave a threat as drugs such as heroin and cocaine pose, that threat pales in comparison to the one posed by fentanyl and its analogues. Indeed, over the past decade, fentanyl has altered the landscape of the substance abuse crisis in Canada, revealing itself as public enemy number one.

[ 94 ] Synthetically produced and readily available on the illicit market, fentanyl is an extremely dangerous and powerful painkiller and sedative. As with other opioids, such as heroin and morphine, it is a highly addictive substance, which, when taken outside of controlled medical environments, puts its users at risk of serious harm, including brain damage, organ damage, coma, and death. Fentanyl’s potential for harm is, however, significantly greater than other opioids. It is, for example, estimated to be 80‑100 times more potent than morphine and 25‑50 times more potent than pharmaceutical grade heroin. Given its strength, a lethal dose will often be less than two milligrams, an amount as small as a single grain of salt ( R. v. Smith , 2016 BCSC 2148 , 363 C.R.R. (2d) 365, at para. 24 ). The risk of overdose and death from fentanyl is thus extremely high, particularly for naïve users or where it is taken in combination with other substances, such as alcohol or other opioids. The risk of overdose is also one that can be difficult to guard against, as traffickers often surreptitiously mix small amounts of fentanyl with other substances to create a cheaper product with the same effects, thereby drastically increasing their profitability (H. Hrymak, “A Bad Deal: British Columbia’s Emphasis on Deterrence and Increasing Prison Sentences for Street‑Level Fentanyl Traffickers” (2018), 41 Man. L.J. 149, at p. 153). This deceptive practice leaves users vulnerable and unaware, especially as fentanyl is physically indistinguishable from other hard drugs, such as heroin, oxycodone, and cocaine (C. C. Ruby, Sentencing (10th ed. 2020); Smith , at para. 24; R. v. Joumaa , 2018 ONSC 317 , at para. 12).

[ 95 ] Analogues or derivatives of fentanyl further exacerbate the risks, as these substances can be far more potent than even fentanyl itself, with some estimated to be as much as 100 times more potent than fentanyl. One such analogue, carfentanil, is so toxic that it “has no safe or beneficial human use, even within the medical community in highly controlled environments” (A. Sabbadini and A. Boni, Sentencing Drug Offenders (loose-leaf), at s. 2:1600.10).

[ 96 ] Beyond its mere potential to cause harm, however, fentanyl has had — and continues to have — a real and deadly impact on the lives of Canadians. Indeed, trafficking in fentanyl is so deadly that various courts have described it as a national crisis, reflective of an increased understanding of the gravity of the harm it causes (see, e.g., R. v. Smith , 2017 BCCA 112 , at para. 50 ; R. v. Vezina , 2017 ONCJ 775 , at para. 58 ; R. v. Aujla , 2016 ABPC 272 , at para. 1 ). This heightened understanding is supported by the available statistical evidence. The expert evidence on the record before us establishes, for instance, that fentanyl‑related deaths in Alberta increased by 4,858 percent between 2011 and 2017, rising from 12 deaths in 2011 to 583 deaths in 2017. More broadly, federal statistics on opioid‑related deaths show that, between January 2016 and March 2021, approximately 23,000 Canadians lost their lives due to accidental apparent opioid‑related deaths, with fentanyl involved in 71 percent of these deaths (Special Advisory Committee on the Epidemic of Opioid Overdoses, Opioid and Stimulant-related Harms in Canada (September 2021) (online). The epidemic also shows no signs of abating, with over 6,000 accidental deaths occurring in 2020 alone, 82 percent of which involved fentanyl (Government of Canada, Federal actions on opioids to date (June 2021) (online)). These figures throw into stark relief the dark and inescapable reality that “[e]very day in our communities, fentanyl abuse claims the lives of Canadians” ( R. v. Loor , 2017 ONCA 696 , at para. 33 ).

[ 97 ] The scale of fentanyl’s devastating impact becomes even more apparent when one considers that, between 2016 and 2020, there were approximately 3,400 homicides across Canada, a number far below the number of fentanyl‑related deaths (Statistics Canada, Table 35‑10‑0069‑01 — Number of homicide victims, by method used to commit the homicide , July 27, 2021 (online)). This disparity makes clear that, in a very real way, those individuals responsible for the largescale distribution of fentanyl within our communities are a source of far greater harm than even those responsible for the most violent of crimes.

[ 98 ] The time has thus come for our perception of the gravity of largescale trafficking in fentanyl to accord with the gravity of the crisis it has caused. Largescale trafficking in fentanyl is not a crime marked merely by the distribution and sale of an illicit substance; rather, it is a crime marked by greed and the pursuit of profit at the expense of violence, death, and the perpetuation of a public health crisis previously unseen in Canadian society. In many ways, “[t]rafficking in fentanyl is almost the equivalent of putting multiple bullets in the chambers of a revolver and playing Russian roulette. It is the most efficient killer of drug users on the market today” ( R. v. Frazer , 2017 ABPC 116 , 58 Alta. L.R. (6th) 185, at para. 11 ). Put simply, it is a crime that can be expected to not only destroy lives, but to undermine the very foundations of our society.

[ 99 ] My comments that follow do not apply to sentences for street-level trafficking, or where traffickers are motivated by a need to support their own addiction. Rather, the focus of this guidance is on the directing minds of largescale fentanyl trafficking operations.

[ 100 ] In my view, heavy penitentiary sentences will be appropriate where offenders have trafficked in large quantities of fentanyl and assumed leadership roles in the trafficking operation. Indeed, in the context of largescale fentanyl trafficking operations, substantial sentences should be neither unusual nor reserved for exceptional circumstances. As this Court has previously explained, maximum sentences should not be reserved for the “abstract case of the worst crime committed in the worst circumstances”, but rather should be imposed whenever the circumstances warrant it ( R. v. L.M. , 2008 SCC 31 , [2008] 2 S.C.R. 163, at para. 22 ; see also Friesen , at para. 114 ).

III. Conclusion

[ 101 ] Ultimately, largescale trafficking in fentanyl is a crime that preys disproportionally on the misery of others — the marginalized and those whose lives are marked by hopelessness and despair. It is a crime motivated by greed and by a callous disregard for the untold grief and suffering it leaves in its wake. Above all, it is a crime that kills — often and indiscriminately. It follows, in my view, that what matters most is that those individuals who choose to prey on the vulnerable and profit from the misery of the Canadian public for personal gain are sentenced in accordance with the severity of the harms they have caused. Fentanyl trafficking, and largescale trafficking in particular, are a source of unspeakable harm. Accordingly, while the range of sentences currently imposed for the directing minds of largescale fentanyl operations straddles the upper single digits and lower double digits, sentencing judges should feel justified, where circumstances warrant, in applying a higher range, consisting of mid-level double digit sentences and, in particularly aggravating circumstances, potential sentences of life imprisonment.

The following are the reasons delivered by

Rowe J. —

TABLE OF CONTENTS

Paragraph

I. Introduction

102

II. Analysis

107

A. Principles of Sentencing and Appellate Review

108

(1) Sentencing: Broad Discretion to Arrive at Proportionate and Individualized Sentences

108

(2) Appellate Review: A Deferrential Approach

116

(a) Error Correction

117

(b) Developing the Law and Providing Guidance to Sentencing Judges

123

(i) Sentencing Ranges

124

(ii) Starting Points

127

B. The Rationale Underlying the Starting-Point Approach

133

(1) Discretion and Individualization

134

(2) Proportionality

144

(3) Role of Appellate Courts

146

(4) Conclusion

150

C. Starting Points in Practice

151

(1) Setting the Starting Point and the Role of Appellate Courts

153

(a) Setting Starting Points Extends Beyond the Role of Appellate Courts

153

(b) Ossification and Starting Points

160

(2) Starting-Point Sentencing Is Inconsistent With the Principles of Sentencing

165

(a) Starting Points Impede Individualized Sentencing and Constrain Discretion

166

(i) Starting Points Privilege Consideration of the Gravity of the Offence Over the Offender’s Moral Blameworthiness

167

(ii) Pre‑Weighing Factors by the Appellate Court

173

(iii) Starting Points Produce an Artificial Sentencing Methodology by Overemphasizing the Judicially Created Category

176

(iv) The “Clustering” Effect of Starting Points

179

(b) Starting Points Overemphasize Denunciation and Deterrence

184

(c) The Starting-Point Approach Is Contrary to Parliament’s Direction for Sentencing Aboriginal Offenders

186

(d) The Starting-Point Approach Does Not Provide Sufficient Guidance to Sentencing Judges

191

(3) Rigid Appellate Intervention Is Inevitable

194

III. Disposition

204

I. Introduction

[ 102 ] These appeals provide an opportunity for this Court to resolve the “issue of importance” it identified in R. v. Friesen , 2020 SCC 9 , at para. 41 : are “starting points . . . a permissible form of appellate guidance”? I would answer this question in the negative. The starting‑point approach pioneered by the Court of Appeal of Alberta is, in theory and in practice, contrary to Parliament’s sentencing regime and this Court’s jurisprudence. The starting-point approach undermines the discretion of sentencing judges and departs from the standard of deference required by appellate courts. As a result, it thwarts the imposition of proportionate and individualized sentences.

[ 103 ] That the starting-point methodology constrains sentencing judges’ discretion and operates as an effective instrument of control for the Court of Appeal should come as no surprise, as that is what it was designed to do. Nor can we be uncertain as to what the Court of Appeal intended. It has been forthright and plain spoken, in R. v. Arcand , 2010 ABCA 363 , 40 Alta. L.R. (5th) 199, and in its subsequent decisions.

[ 104 ] Where I differ from the Court of Appeal is on these foundational points: the rigid control they exercise over sentencing and the constraints on individualized sentencing are, in my view, neither legitimate, nor necessary. This is neither accidental nor is it hidden. To the contrary, it is deliberate and it is open.

[ 105 ] My colleagues say that they are putting forth “a revised understanding” of starting points (Brown and Martin JJ.’s reasons, at para. 3). I am skeptical, however, as to the impact this so‑called revised approach will have. This Court has provided guidance on numerous prior occasions, yet the Court of Appeal’s approach has remained unchanged ( R. v. McDonnell , [1997] 1 S.C.R. 948; R. v. Lacasse , 2015 SCC 64 , [2015] 3 S.C.R. 1089; Friesen ). As a result, in my view, to offer yet more helpful suggestions to ameliorate the problems inherent in starting-point methodology is naïve. The problems are baked into the methodology and they are rooted in the purposes for which it was developed and for which it has been maintained. There is only one effective response: to say that starting-point methodology can no longer be used. Being definitive in so saying is long overdue.

[ 106 ] Lastly, on the merits and on the additional guidance he provides, I agree with Justice Moldaver and would adopt his reasons. I would dismiss the appeals.

II. Analysis

[ 107 ] My analysis proceeds in three steps. First, I review the principles of sentencing and appellate review. Second, I show how the rationale underlying the starting-point approach is inconsistent with these principles. Finally, I turn to the practical effects of starting points and their incompatibility with these same principles.

A. Principles of Sentencing and Appellate Review

(1) Sentencing: Broad Discretion to Arrive at Proportionate and Individualized Sentences

[ 108 ] Until 1996, the Criminal Code , R.S.C. 1985, c. C‑46, did not provide guidance as to the principles or objectives of sentencing (A. Manson et al., Sentencing and Penal Policy in Canada: Cases, Materials, and Commentary (3rd ed. 2016), at pp. 37-38). In common with other common law jurisdictions, Canadian courts adopted an approach to sentencing emphasizing discretion, proportionality and individualization (G. Brown, Criminal Sentencing as a Practical Wisdom (2017), at p. 25). This approach was notably articulated by the Court of Appeal for Ontario in R. v. Willaert , [1953] O.R. 282 , in an oft‑cited passage:

I am respectfully of opinion that the true function of criminal law in regard to punishment is in a wise blending of the deterrent and reformative , with retribution not entirely disregarded, and with a constant appreciation that the matter concerns not merely the Court and the offender, but also the public and society as a going concern. Punishment is, therefore, an art — a very difficult art — essentially practical, and directly related to the existing needs of society. . . . It is therefore impossible to lay down hard and fast and permanent rules. [Emphasis added; p. 286.]

[ 109 ] Although Parliament later gave additional direction to sentencing judges, the “wise blending” approach remains good law: judges must use their discretion to weigh different penal aims in light of all circumstances to arrive at sentences that are “fit” for the offence and the offender (Manson et al., at p. 41a).

[ 110 ] In its 1996 sentencing reform, Parliament codified the objectives and principles of sentencing in ss. 718 to 718.2 of the Criminal Code . Section 718 now provides that the “fundamental purpose of sentencing is to protect society and to contribute, along with crime prevention initiatives, to respect for the law and the maintenance of a just, peaceful and safe society by imposing just sanctions”. This purpose is achieved having regard to six objectives: denunciation, general and specific deterrence, separation of offenders, rehabilitation, reparation, and the promotion of a sense of responsibility in the offender and acknowledgement of the harm caused to the victim and to the community ( R. v. Nasogaluak , 2010 SCC 6 , [2010] 1 S.C.R. 206 , at para. 39 ). The Controlled Drugs and Substances Act , S.C. 1996, c. 19, contains similar provisions on sentencing ( ss. 10(1) and 10(2) ).

[ 111 ] As mandated by s. 718.1 of the Criminal Code , in all cases, “whatever weight a judge may wish to accord to the objectives listed above, the resulting sentence must respect the fundamental principle of proportionality” ( Nasogaluak , at para. 40 (emphasis in original)). Proportionality is the “ sine qua non of a just sanction” ( R. v. Ipeelee , 2012 SCC 13 , [2012] 1 S.C.R. 433, at para. 37 ). This principle provides that “ sentences must be proportionate to the gravity of the offence and the degree of responsibility of the offender” ( Friesen , at para. 30 ). There are two converging perspectives on proportionality: first, it serves a restraining function as it requires that a sentence not exceed what is just and appropriate, and second, it seeks to ensure that the sentence “properly reflects and condemns [the role of offenders] in the offence and the harm they caused” ( Nasogaluak , at para. 42 ). At the end of the day, a just sanction is one that reflects both of these “perspectives on proportionality and does not elevate one at the expense of the other ” ( Ipeelee , at para. 37 ).

[ 112 ] For its part, s. 718.2 provides a non‑exhaustive list of secondary principles that must guide the sentencing process. These principles include “the consideration of aggravating and mitigating circumstances, the principles of parity and totality, and the instruction to consider ‘all available sanctions other than imprisonment that are reasonable in the circumstances’, with particular attention paid to the circumstances of aboriginal offenders” ( Nasogaluak , at para. 40 , quoting s. 718.2 of the Criminal Code ). In particular, parity requires that “similar offenders who commit similar offences in similar circumstances . . . receive similar sentences” ( Friesen , at para. 31 ). It is principally because of parity that appellate courts sometimes adopt sentencing ranges or starting points to serve as a guide for sentencing judges ( Lacasse , at paras. 56‑57 ). In Friesen , this Court explained that parity is an expression of proportionality: “A consistent application of proportionality will lead to parity. Conversely, an approach that assigns the same sentence to unlike cases can achieve neither parity nor proportionality . . .” (para. 32).

[ 113 ] In order to produce proportionate sentences, sentencing must be a “highly individualized exercise” ( Lacasse , at para. 58 ; see also R. v. Boudreault , 2018 SCC 58 , [2018] 3 S.C.R. 599, at para. 58 ). Sentencing judges must decide a profoundly contextual issue: “For this offence, committed by this offender, harming this victim, in this community, what is the appropriate sanction under the Criminal Code ?” ( R. v. Gladue , [1999] 1 S.C.R. 688, at para. 80 (emphasis in original)). They must determine which objectives of sentencing merit greater weight and evaluate the importance of mitigating or aggravating factors, to best reflect the circumstances of each case ( Nasogaluak , at para. 43 ; R. v. Lyons , [1987] 2 S.C.R. 309; R. v. M. (C.A.) , [1996] 1 S.C.R. 500 , at para. 92 ; R. v. Hamilton (2004) , 72 O.R. (3d) 1 (C.A.)).

[ 114 ] Individualization flows from proportionality: a sentence that is not tailored to the specific circumstances of both the offender and the offence will not be proportional to the gravity of the offence and the degree of responsibility of the offender ( R. v. Proulx , 2000 SCC 5 , [2000] 1 S.C.R. 61 , at para. 82 ). Simply stated, [ translation ] “[a] proportional sentence is thus an individualized sentence (J. Desrosiers and H. Parent, “Principes”, in JurisClasseur Québec — Collection droit pénal — Droit pénal général (loose-leaf), by M.-P. Robert and S. Roy, eds., fasc. 20, at para. 17).

[ 115 ] Parliament vested sentencing judges with “a broad discretion” to craft individualized and proportionate sentences ( Nasogaluak , at para. 43 ; see also M. (C.A.) , at para. 90 , referring to what is now s. 718.3(1) of the Criminal Code ). “Far from being an exact science or an inflexible predetermined procedure, sentencing is primarily a matter for the trial judge’s competence and expertise. The trial judge enjoys considerable discretion because of the individualized nature of the process” ( R. v. L.M. , 2008 SCC 31 , [2008] 2 S.C.R. 163, at para. 17 ). It is possible that, in a given case, more than one particular sentence would be appropriate and reasonable (M. Vauclair and T. Desjardins, in collaboration with P. Lachance, Traité général de preuve et de procédure pénales (28th ed. 2021), at No. 47.2). “ Proportionality will be achieved by means of a ‘complicated calculus’ whose elements the trier of fact understands better than anyone” ( L.M. , at para. 22 ). Thus, flexibility is essential to meet the needs of individual justice. In short, discretion is the means to achieve proportionality in sentencing.

(2) Appellate Review: A Deferential Approach

[ 116 ] Appellate courts play a dual role in sentence appeals : they are courts of error correction and they have a role “in developing the law and providing guidance” ( Friesen , at paras. 34‑35 ; Lacasse , at paras. 36‑37 ).

(a) Error Correction

[ 117 ] As a corollary to the wide discretion conferred to sentencing judges in the determination of a “just and appropriate” sentence, this Court has adopted a deferential approach to appellate review of sentencing decisions. An appellate court can only vary a sentence if (1) the sentence is demonstrably unfit or (2) the sentencing judge made an error in principle that had an impact on the sentence ( Friesen , at para. 26 ).

[ 118 ] A sentence is “demonstrably unfit” if it constitutes an “unreasonable departure” from the fundamental principle of proportionality, which is a “very high threshold” ( Lacasse , at paras. 52‑53 ). It will reach this level where it “is in substantial and marked departure from the sentences customarily imposed for similar offenders committing similar crimes” ( M. (C.A.) , at para. 92 ; see also Lacasse , at para. 67 ; R. v. Suter , 2018 SCC 34 , [2018] 2 S.C.R. 496 , at paras. 23‑24 ). As for errors in principle, they include “an error of law, a failure to consider a relevant factor, or erroneous consideration of an aggravating or mitigating factor” ( Friesen , at para. 26 ). An appellate court can only intervene if the error in principle had an impact on the sentence ( ibid. , citing Lacasse , at para. 44 ).

[ 119 ] Accordingly, an appellate court cannot intervene simply because it would have weighed relevant factors and objectives differently ( Lacasse , at paras. 49‑51 ). The weighing of factors can form an error in principle “[o]nly if by emphasizing one factor or by not giving enough weight to another, the trial judge exercises his or her discretion unreasonably” ( Friesen , at para. 26 (text in brackets in original), quoting R. v. McKnight (1999) , 135 C.C.C. (3d) 41 (Ont. C.A.), at para. 35 ). Similarly, an appellate court cannot vary a sentence simply because it would have put the sentence in a different range or category ( Friesen , at para. 37 ).

[ 120 ] Deference arises from Parliament’s choice to grant discretion to sentencing judges to individualize sentences (para. 38). The appellate standard of review “complement[s] and reinforce[s]” the view that proportionate sentences are best achieved by the application of judicial discretion (P. Healy, “Sentencing from There to Here and from Then to Now” (2013), 17 Can. Crim. L.R. 291, at p. 295). Appellate courts’ limited powers to vary sentences “ underscores the importance of individualized decision‑making in determining a fit disposition” ( ibid. ).

[ 121 ] In addition, this Court in Friesen emphasized three functional reasons justifying the appellate standard of review: (1) sentencing judges see and hear all the evidence in person; (2) they usually have front‑line experience and understand the needs of the community where the crime was committed; and (3) appellate courts should generally defer to sentencing judges’ decisions “to avoid delay and the misuse of judicial resources” (para. 25; see also Lacasse , at paras. 11 and 48 ; R. v. Shropshire , [1995] 4 S.C.R. 227, at para. 46 ).

[ 122 ] If the appellate court finds an error in principle that had an impact on the sentence, or that the sentence is demonstrably unfit, it can intervene and “apply the principles of sentencing afresh to the facts” ( Friesen , at para. 27 ). It must, however, defer to the judge’s findings of fact and identification of aggravating and mitigating factors if they are not affected by an error in principle ( para. 28).

(b) Developing the Law and Providing Guidance to Sentencing Judges

[ 123 ] Second, appellate courts “serve an important function” in providing guidance to sentencing judges and “minimizing the disparity of sentences . . . for similar offenders and similar offences committed throughout Canada” ( M. (C.A.) , at para. 92 , see also Vauclair and Desjardins, No. 47.37, at p. 1168 ). While appellate courts often use sentencing ranges, other courts, particularly Alberta’s, use starting points as a form of guidance. I will review both forms of appellate guidance in general terms before turning to the issue of whether starting points are an appropriate form of guidance in sentencing.

(i) Sentencing Ranges

[ 124 ] Sentencing ranges are “summaries of the minimum and maximum sentences imposed in the past, which serve in any given case as guides for the application of all the relevant principles and objectives” ( Lacasse , at para. 57 ). Ranges can also develop the law in the ordinary way of appellate courts, responding to emerging needs that are not addressed in the existing jurisprudence ( Friesen , at para. 35 ).

[ 125 ] Sentencing ranges can assist sentencing judges when considering all the relevant circumstances of the offence and the offender to arrive at a fit sentence. They are not “averages”, “hard and fast rules”, or “binding” ( Lacasse , at para. 57 ; Friesen , at para. 37 ). The usual range of sentences imposed for an offence may be a useful tool at the disposal of sentencing judges as it “gives some sense [to] proportionality and parity” (A. Manson, The Law of Sentencing (2001), at p. 65). But sentencing judges are not obliged to have regard to ranges or to begin the sentencing exercise at the range. Their fundamental duty is to impose a sentence that is “proportionate to the gravity of the offence and the degree of responsibility of the offender” ( s. 718.1 of the Criminal Code ). Conversely, sentencing within a range is not determinative of fitness:

[ translation ] ... it should be borne in mind that ranges are only indications. Care must be taken not to apply them categorically. Although they constitute a useful reference tool for a judge, they do not exempt him or her from analyzing the particular circumstances of each case, as the focus of the sentencing process continues to be on the individual . This is why offenders sometimes receive different sentences for the same offence.

[Emphasis added.]

( Calderon v. R. , 2015 QCCA 1573 , at para. 30 )

[ 126 ] Since sentencing ranges are only guidelines, they do not, and cannot, increase the authority of appellate courts to vary a sentence. Ensuring that sentences are similar cannot short‑circuit the standard of review and “be given priority over the . . . deference [owed] to the [sentencing] judges’ exercise of discretion” ( L.M. , at para. 35 ; see also Ferland v. R. , 2009 QCCA 1168 , [2009] R.J.Q. 1675, at para. 22 ).

(ii) Starting Points

[ 127 ] The starting-point approach involves a three‑step process. First, the court of appeal must describe the category created (e.g., “commercial cocaine trafficking”, “major sexual assault”) . Second, it sets a starting point for that category which represents the sentence that would apply to a typical offender of good character with no criminal record in a typical case ( McDonnell , at para. 59 , per McLachlin J., dissenting, citing R. v. Sandercock (1985) , 1985 ABCA 218 , 22 C.C.C. (3d) 79 (Alta. C.A.) ; see also R. v. Felix , 2019 ABCA 458 , 98 Alta. L.R. (6th) 136 (“ Felix (ABCA)”), at para. 45 ). Third, the sentencing judge refines the sentence to the specific facts of the individual case and offender ( Arcand , at para. 104 ).

[ 128 ] Starting points are, by their nature, a prescriptive form of appellate guidance, in that they provide a sequence for sentencing judges to follow when determining a fit sentence. As the expression “starting point” makes clear, sentencing judges must identify the relevant category and corresponding starting point to commence the sentencing exercise. They must then adjust the sentence depending on the facts of the case and the characteristics of the offender, in light of the factors built into the starting point ( McDonnell , at para. 60 ; Arcand , at para. 105 ).

[ 129 ] Although this Court has not disavowed the starting-point approach, it has been clear that they cannot alter the broad discretion granted to sentencing judges ( McDonnell , at paras. 32‑33 and 43 ; Friesen , at para. 37 ). It also specifically disapproved of the Court of Appeal of Alberta’s treatment of starting points as binding and its failure to give proper effect to the standard of review ( Friesen , at paras. 37 and 40‑41 ).

[ 130 ] As this Court noted in Friesen (at para. 41 ), practitioners, academics, and judges — including judges at the Court of Appeal of Alberta — have nevertheless expressed concerns about the starting-point approach and whether it is consistent with the broad discretion of sentencing judges to impose individualized sentences (see, e.g., A. Manson, “ McDonnell and the Methodology of Sentencing” (1997), 6 C.R. (5th) 277 ; J. Rudin, “Eyes Wide Shut: The Alberta Court of Appeal’s Decision in R. v. Arcand and Aboriginal Offenders” (2011), 48 Alta. L. Rev. 987; Arcand , at para. 352 ; R. v. Lee , 2012 ABCA 17 , 58 Alta. L.R. (5th) 30, at paras. 55 , 61 and 76; R. v. Gashikanyi , 2017 ABCA 194 , 53 Alta L.R. (6th) 11, at paras. 19 , 22 and 77‑78; R. v. D.S.C. , 2018 ABCA 335 , [2019] 3 W.W.R. 259, at para. 40 ). This case gives us the opportunity to assess the merits of these concerns.

[ 131 ] At the outset, I note that although “local characteristics in a given region may explain certain differences in the sentences imposed on offenders by the courts” ( Lacasse , at para. 89 ), the Criminal Code sets out uniform principles and objectives for sentencing in the country. In my view, this uniformity should ideally extend to sentencing methodology (C. C. Ruby, Sentencing (10th ed. 2020), at § 2.12). For instance, the methodology for choosing whether sentences should be consecutive or concurrent and for assessing joint submissions does not and should not vary across provinces and territories. In the same manner, the use of sentencing ranges or starting points should also be generally consistent in this country.

[ 132 ] The starting‑point approach is a distinct form of appellate guidance —indeed, a distinct theory of sentencing — with its own rationale and methodology. Below, I review the starting‑point approach set out in Arcand , noting how the rationale and functions of this approach depart from the ordinary operation of Canadian sentencing law articulated by the Criminal Code and this Court’s jurisprudence. I begin with a discussion of issues in the underlying rationale for starting points, and then turn to methodological problems with the starting‑point approach in application.

B. The Rationale Underlying the Starting-Point Approach

[ 133 ] The foundational rationale of starting points is the idea that appellate courts are institutionally responsible for creating and enforcing a uniform approach to sentencing to prevent the injustice that would inevitably flow from variation in sentencing. On this view, variations are an impediment to the rule of law goal of maintaining public confidence in the criminal justice system. Starting points aim to achieve a greater degree of parity and thus reduce and confine discretion of sentencing judges. As I will explain, this rationale is incompatible with the principles of sentencing and this Court’s jurisprudence.

(1) Discretion and Individualization

[ 134 ] First, the starting-point approach seeks to reduce arbitrariness, disparity and idiosyncratic decision‑making in sentencing in order to maintain public confidence in the administration of justice ( Arcand , at paras. 24, 70 and 102 ).

[ 135 ] According to Arcand , appellate courts bear a “special institutional responsibility” to maintain public trust in sentencing (para. 7). Absent such appellate leadership, unjust sanctions and associated erosion of public confidence in the justice system inevitably flow from the lack of a uniform approach (paras. 8 and 119). The objective is thus to achieve parity and constrain sentencing judges’ discretion.

[ 136 ] Conversely, the Court of Appeal of Alberta reasons that sentencing ranges do not sufficiently constrain sentencing judges’ discretion and they lead to disparity. Starting points provide more guidance than ranges to sentencing judges who are left to “[t]hro[w] a mental dart at the range” ( Arcand , at para. 122 ). According to the Court of Appeal of Alberta, sentencing ranges are often the product of “limited and parochial” research, where counsel will only try to find cases supporting their point of view (para. 123). Ranges also provide an insufficient basis for appellate intervention because “the mere fact a sentence falls within a range, or not significantly outside it, tells an appeal court little, if anything, about whether the actual sentence is fit” (para. 124).

[ 137 ] The flaw in this rationale is apparent. Variability resulting from individualization is an essential feature of just sentencing, not a problem. Establishing a sentence for a “typical offender” is counterproductive: “. . . the search for a single appropriate sentence for a similar offender and a similar crime will frequently be a fruitless exercise of academic abstraction” ( M. (C.A.) , at para. 92 ).

[ 138 ] Giving effect to Parliament’s choice to confer broad discretion on sentencing judges to impose individualized sentences will inevitably produce variation in sentences. As Cory and Iacobucci JJ. said in Gladue , “ [d]isparity of sentences for similar crimes is a nat ural consequence of th[e] individualized focus [of sentencing] ” (para. 76, see also Proulx , at para. 86 ; Healy, at pp. 294‑95). Eliminating disparity is “practically impossible” given the many and varied factors arising in a given case that make it difficult to compare with “seemingly similar cases” (T. Quigley, “Has the Role of Judges in Sentencing Changed . . . or Should it?” (2000), 5 Can. Crim. L.R. 317, at p. 324). I agree with LeBel J.A.’s (as he then was) assessment of individualization and its relationship with parity:

[ translation ] Sentence individualization remains a fundamental principle in the Canadian sentencing system. It certainly provokes much criticism, sometimes relating to disparities between sentences, such criticism at times being levelled with almost no knowledge of the specifics of each case. This principle is so important that the imposition of a sentence that is abstract and standardized and takes no account of individual factors can constitute an error of law.

( R. v. Lafrance (1993) , 59 Q.A.C. 213, at para. 33 )

[ 139 ] Although the Court of Appeal of Alberta frequently relies on disparity in sentencing for a given offence to justify setting a starting point, such “disparity” is often (1) merely asserted (see, e.g., Arcand , at para. 102 ) or (2) the natural result of individualized sentencing (see, e.g., Gashikanyi , at paras. 29‑31 , for Berger J.A.’s rebuke of the claim of disparity in R. v. Hajar , 2016 ABCA 222 , 39 Alta. L.R. (6th) 209).

[ 140 ] Focussing on variability as the problem is not only inconsistent with this Court’s jurisprudence, it also “helps to create or reinforce other problems: jail becomes the norm, starting point or ranges of sentence become hardened into fixed sentences, and factors leading to systemic discrimination are either ignored or inadequately dealt with” (Quigley, at p. 324 (footnote omitted)).

[ 141 ] Moreover, sentencing ranges are not a “rudimentary sentencing guideline . . . of limited value” ( Arcand , at para. 122 ). The fear of arbitrary sentencing and undue disparity is belied by the Criminal Code itself, which codifies proportionality and parity. The Code explicitly directs judges to consider the facts of other cases to identify patterns when sentencing. Ranges provide a reference to the sentences imposed in similar cases while preserving the individualized and discretionary nature of the sentencing process. As I explained, the mere fact that a sentence falls within a range does not necessarily makes it fit or render it immune from appellate review.

[ 142 ] The Court of Appeal of Alberta has taken upon itself the role as the guardian of a system of sentencing that, without its tight supervision, would fall into disorder thereby undermining the confidence of the public. However, this is a dubious claim as it is difficult to see how the public could have an informed opinion about sentencing decisions in their jurisdictions given the limited data on such issues (A. N. Doob, “The Unfinished Work of the Canadian Sentencing Commission” (2011), 53 C.J.C.C.J. 279, at p. 281). More importantly, there is no evidence of a crisis of public confidence in the sentencing system in provinces that do not use starting points. Despite its repeated claims about the risks that variability in sentencing has bred distrust within the public, the Court of Appeal of Alberta in Arcand failed to “provide any basis for the assumptions regarding how the public feels about sentencing” (Rudin, at p. 997). In fact, research shows that “the public is supportive of judicial discretion and individualized sentencing” (K. N. Varma and V. Marinos, “Three Decades of Public Attitudes Research on Crime and Punishment in Canada” (2013), 55 C.J.C.C.J. 549, at pp. 555‑56 ). This is a far cry from the claim that, under the sentencing range approach, “the public may suspect that some courts too simply sift the collection until they find a case that produces the result the judge is looking for” ( Arcand , at para. 123 ).

[ 143 ] Thus, the premise of the starting‑point approach — that variability is a problem that needs to be fixed — means that the entire approach is built on a flawed foundation. Individualization is critical to Canadian sentencing; it is an imperative, not a problem.

(2) Proportionality

[ 144 ] Second, the starting-point approach is grounded in an erroneous view of the fundamental principle of proportionality. The Court of Appeal of Alberta positions the starting‑point approach itself as the foundation of proportionality: “Inherent in the proportionality principle are (1) the relative ranking of offences as well as ranking of categories within offences based on varying degrees of seriousness; and (2) the setting of starting points for those offences or categories” ( Arcand , at fn. 73, see also para. 103). For the Court of Appeal of Alberta, unlike starting points, “a sentencing range is not an essential component of the proportionality principle. A range develops as an outcome of proportionality; it is not necessarily a determinant of it” (para. 125 (emphasis in original)). But proportionality is not achieved by such a ranking of offences and categories of offences. Rather, it is achieved in each case through individualized sentencing that takes into account the “ the specific circumstances of both the offender and the offence so that the ‘punishment fits the crime’” ( Proulx , at para. 82 ).

[ 145 ] Sentencing judges also give effect to proportionality through the common law process, by examining sentences imposed in similar cases. However, the Court of Appeal’s approach is dismissive of the value of precedent in achieving proportionality, which draws on the experience, philosophy and perspective of all judges over time . As it explained, the use of precedent for sentences “is not usually a useful exercise. . . . Instead, the value of sentencing precedent lies in its principles and (where given) its starting points” ( R. v. M. (B.S.) , 2011 ABCA 105 , 44 Alta. L.R. (5th) 240 , at para. 7 ). Essentially, the Court of Appeal has rejected the common law method in sentencing, which, as explained in Friesen , achieves both parity and proportionality over time through individualized sentencing in each case:

In practice, parity gives meaning to proportionality. A proportionate sentence for a given offender and offence cannot be deduced from first principles; instead, judges calibrate the demands of proportionality by reference to the sentences imposed in other cases. Sentencing precedents reflect the range of factual situations in the world and the plurality of judicial perspectives. Precedents embody the collective experience and wisdom of the judiciary. They are the practical expression of both parity and proportionality. [para. 33]

This is a key distinguishing feature between starting points and sentencing ranges.

(3) Role of Appellate Courts

[ 146 ] Finally, according to the Court of Appeal of Alberta, appellate courts have significant advantages over sentencing judges, which justifies their “leadership role” in sentencing ( Arcand , at para. 87 ). Appellate courts are “institutionally better able to properly assure similarity of treatment in sentencing” (para. 82). They can “appreciate overall sentencing practices, patterns and problems in their jurisdiction” and have “an enlarged reservoir of experience, philosophy and perspective to be synthesized into acceptable, and accepted, starting points” (para. 153). In fact, by codifying parity as a principle of sentencing in 1996, “Parliament necessarily reinforced the powers of courts of appeal” (para. 82).

[ 147 ] As the foregoing makes clear, the starting-point approach is premised on a misconception of the role of appellate courts in sentencing. The “reinforced” role of appellate courts is incompatible with this Court’s sentencing jurisprudence, which emphasizes discretion and adopted a correspondingly deferential standard of appellate review (see, e.g., R. v. Stone , [1999] 2 S.C.R. 290, at para. 230 ). Sentencing judges, rather than appellate courts, are in a privileged position and possess “unique qualifications of experience and judgment” to determine what is a just and appropriate sentence ( M. (C.A.) , at para. 91 ).

[ 148 ] Moreover, appellate courts can demonstrate “leadership” and provide guidance without setting starting points. As mentioned, appellate courts can “depart from prior precedents and sentencing ranges in order to impose a proportionate sentence” ( Friesen , at para. 108 ; see also Manson (2001), at p. 66), while respecting the sentencing judges’ primary responsibility for sentencing.

[ 149 ] In sum, because the starting‑point approach sees individualization of approach as a threat to the rule of law, it requires appellate courts to create and enforce a uniform approach to sentencing. It reverses the logic of deference to sentencing judges and frames appellate courts, rather than sentencing courts, as being primarily institutionally responsible and capable for sentencing.

(4) Conclusion

[ 150 ] The starting-point approach is flawed, as it rests on unsound legal foundations. The foundational problems of the starting-point approach I described go to the root of it. It views the wide discretion of sentencing judges as a problem, sentencing ranges as rudimentary and insufficient constraints, and appellate courts as endowed with an interventionist role. It is premised on a fundamentally different understanding of the roles of appellate and sentencing courts. Thus, starting points are a mechanism to shift effective decision‑making authority from individual sentencing judges and concentrate that authority in the Court of Appeal as a body that tightly supervises sentencing in the province. Given these foundational problems, I would disavow the approach as a whole.

C. Starting Points in Practice

[ 151 ] I turn now from doctrinal concerns with starting points to the practical problems that arise from their use. The latter is a manifestation in practice of what the former describes at the level of principle. In short, applying unsound doctrine gives rise to problematic consequences. To tinker with those consequences while leaving intact the unsound doctrine will yield little of significance. It is with this perspective that I address the application of the starting-point approach.

[ 152 ] In my view, the starting-point approach produces practical issues at each of its stages: (1) how starting points are set; (2) how they are applied by sentencing judges; and (3) how they are reviewed by appellate courts. I examine each in turn.

(1) Setting the Starting Point and the Role of Appellate Courts

(a) Setting Starting Points Extends Beyond the Role of Appellate Courts

[ 153 ] The first step in the starting‑point approach involves the appellate court categorizing offences and setting a number for each category. This Court has made it clear that such judicially‑created categories cannot be treated as binding, because it would usurp Parliament’s role ( McDonnell , at para. 33 ).

[ 154 ] Judicial categorization of offences is not, per se , an issue. A degree of categorization is necessary for broad offences with wide penalty ranges. This Court has emphasized the importance of clarity in categorization for all appellate guidance in sentencing ( Friesen , at para. 39 ; Stone , at para. 245 ).

[ 155 ] However, even if starting points are not treated as binding, setting starting points is a policy‑intensive process, which the legislature or a statutory body such as a sentencing commission would be better suited to conduct. This is made clear by the Court of Appeal of Alberta’s approach to setting starting points, which resembles a legislative process. For example, in setting a starting point for major sexual interference, the court sought “evidence, records, literature and arguments” in relation to 13 complex issues ( R. v. Bjornson , 2012 ABCA 230 , 536 A.R. 1, at para. 8 ; Hajar , at para. 54 ). The Court of Appeal in Bjornson and Hajar can be seen as exercising a parallel function to Parliament, defining subcategories of offences and new rules of sentencing for each subcategory.

[ 156 ] Starting points can then operate like judicially‑created criminal offences. For example, the three‑year starting point for commercial trafficking in cocaine applies to “more than a minimal” trafficking ( R. v. Maskill (1981), 1981 ABCA 50 , 29 A.R. 107 (C.A.) , at para. 20 ). Whether the three‑year starting-point applies is determined by reference to various “indicia” of commerciality, which can be “developed through the case law” ( R. v. Melnyk , 2014 ABCA 313 , 580 A.R. 389 , at para. 6 ). In effect, the Court of Appeal has created an offence for the purpose of sentencing. But creating new offences is the exclusive preserve of Parliament, not the courts ( s. 9 (a) of the Criminal Code ; McDonnell , at paras. 33-34 ; Friesen , at para. 37 ). The principle of separation of powers demands that courts respect the limits of their role in the constitutional order and “show proper deference for the legitimate sphere of activity” of the other branches of the state ( New Brunswick Broadcasting Co. v. Nova Scotia (Speaker of the House of Assembly) , [1993] 1 S.C.R. 319, at p. 389), which have distinct roles and institutional capacities ( Ontario v. Criminal Lawyers’ Association of Ontario , 2013 SCC 43 , [2013] 3 S.C.R. 3 , at para. 29 ; R. v. Chouhan , 2021 SCC 26 , [2021] 2 S.C.R. 136, at paras. 130‑31 , per Rowe J.).

[ 157 ] Provincial appellate courts are not suited to make policy on sentencing, since they lack the resources to gather information to do so and are structured to respond to individual cases (Canadian Sentencing Commission, Sentencing Reform: A Canadian Approach (1987), at p. xxiv; see also Manson (1997), at pp. 291-92; M. R. Bloos and W. N. Renke, “Stopping Starting Points: R. v. McDonnell ” (1997), 35 Alta. L. Rev. 795, at pp. 806-7). Appellate courts can certainly provide guidance and develop the law in light of policy considerations, but they are ill‑equipped to make general sentencing policy such as establishing starting points . They do not have the institutional capacity to gather and process the type of information the Court of Appeal of Alberta sought in Bjornson and Hajar . Sentencing ranges do not raise similar concerns. Even when appellate courts set the law in a new direction, they do not engage in the policy‑intensive exercise as the Court of Appeal of Alberta did in cases like Hajar . In Friesen , for instance, this Court simply said that an upward departure from prior precedents may be required (paras. 107‑14). It did not “adop[t] a form of starting point”, contrary to the suggestion of the Alberta Court of Appeal in R. v. Hotchen , 2021 ABCA 119 , 22 Alta. L.R. (7th) 64, at para. 12 .

[ 158 ] In other jurisdictions with starting‑point sentencing, setting starting points is generally undertaken by sentencing commissions, which are better equipped than appellate courts to deal with these interlocking policy issues (see, e.g., A. Ashworth and J. V. Roberts, “The Origins and Nature of the Sentencing Guidelines in England and Wales”, in A. Ashworth and J. V. Roberts, eds., Sentencing Guidelines: Exploring the English Model (2013), 1, at p. 3 ). In fact, the Canadian Sentencing Commission rejected the use of numerical sentencing grids found in some states in the United States (pp. 296‑300; J. V. Roberts, “Structuring Sentencing in Canada, England and Wales: A Tale of Two Jurisdictions” (2012), 23 Crim. L.F. 319, at p. 323 ). More importantly, Parliament rejected models of guideline sentencing established or overseen by provincial appellate courts. Rather, it chose broad discretion guided by a common set of purposes and principles.

[ 159 ] I would add that among other problems, starting points raise procedural fairness concerns. The offender is in a difficult position because he or she may not have the resources to bring before the court all the evidence, social science literature, statistics and arguments to guide the court in setting a starting point. Offenders are concerned with their own case, not future hypothetical offenders or general principles of sentencing. In addition, the starting point set then applies to future offenders, who do not have the opportunity to challenge the Crown’s evidence.

(b) Ossification and Starting Points

[ 160 ] Sentencing requires flexibility “to ensure a result that is fit for the offender and for the administration of criminal justice” (Manson et al., at p. 38). Flexibility means that what is considered to be a “fit sentence” can evolve. As MacKay J.A. said in Willaert , “[a] punishment appropriate to-day might have been quite unacceptable two hundred years ago and probably would be absurd two hundred years hence. It is therefore impossible to lay down hard-and-fast and permanent rules” (p. 286). The starting-point approach does not provide adequate room for such flexibility.

[ 161 ] In Alberta, for instance, starting points can only be varied by the Court of Appeal in accordance with its reconsideration procedure, which requires leave ( Arcand , at paras. 107 and 195‑200 ). The premise of this requirement is that the variation of a starting point changes the law by overturning a binding precedent (para. 187). This issue of rigidity is compounded by the fact that starting points are often established in response to the prevalence of an offence in a community, with a view of requiring the imposition of higher sentences (see, e.g., R. v. Johnas (1982), 1982 ABCA 331 , 41 A.R. 183 (C.A), at para. 5 ; R. v. Matwiy (1996), 1996 ABCA 63 , 178 A.R. 356 (C.A.), at para. 33 ). For instance, in the instant case, the Court of Appeal considered the public health crisis in Alberta in the creation of the starting point ( Felix (ABCA), at para. 40). But, even if the circumstances seen as justifying a high starting point may subside, the starting point remains in place until it is overruled.

[ 162 ] The inability of starting points to evolve is borne out in practice. I am unaware of a starting point that has been reconsidered by the Court of Appeal of Alberta. Thus far, in effect, once a starting point has been established, it is set in stone.

[ 163 ] By contrast, variation of sentencing ranges occurs gradually through the common law method:

“Ranges” are not embedded in stone. Given their nature as guidelines only, I do not view them as being fixed in law, as is the case with binding legal principles. They may be altered deliberately, after careful consideration, by the courts. Or, they may be altered practically, as a consequence of a series of decisions made by the courts which have that effect. If a range moves by virtue of the application of individual cases over time, it is not necessary to overrule an earlier range that may once have been in vogue; it is only necessary to recognize that the courts have adapted and the guidelines have changed. [Emphasis added.]

( R. v. Wright (2006) , 83 O.R. (3d) 427, at para. 22 ; see also H. Parent and J. Desrosiers, Traité de droit criminel , t. III, La peine (3rd ed. 2020), at pp. 699‑700.)

[ 164 ] In my view, the application of the reconsideration procedure to starting points is another sign that the starting‑point approach is fundamentally incompatible with Canadian sentencing law. Like many other aspects of the starting‑point approach, it is premised on the notion that starting points are and must be binding. The rigidity of starting points is not only due to their aggressive enforcement by the Court of Appeal. Rather, it is ingrained into its operation, including the reconsideration procedure.

(2) Starting‑Point Sentencing Is Inconsistent With the Principles of Sentencing

[ 165 ] Once starting points are set, the next step is for sentencing judges to apply them. In my view, this is another area in which the starting-point approach is inconsistent with the principles of sentencing prescribed by Parliament and by this Court. Even if the Court of Appeal were to give proper effect to the standard of review and stopped treating starting points as binding, something that the Court of Appeal of Alberta has resisted, it does not resolve the inherent issues raised by sentencing judges applying the starting-point approach. First, sentencing judges applying the starting‑point approach have less discretion to fully consider all relevant circumstances and are thus are less likely to arrive at individualized and proportionate sentences. Second, starting points overemphasize deterrence and denunciation. Third, they are incompatible with Parliament’s direction for sentencing Aboriginal offenders. Finally, they fail to provide guidance to sentencing judges in certain important matters.

(a) Starting Points Impede Individualized Sentencing and Constrain Discretion

[ 166 ] Sentencing judges must individualize sentencing “both in method and outcome” ( Friesen , at para. 38 ). But the starting-point approach does not enable sentencing judges to sufficiently individualize sentencing as its very purpose is to restrain judicial discretion.

(i) Starting Points Privilege Consideration of the Gravity of the Offence Over the Offender’s Moral Blameworthiness

[ 167 ] Starting points are defined solely in relation to the gravity of the offence. The moral blameworthiness and personal characteristics of the offender are only a secondary consideration, when adjusting the sentence away from the starting point. This is a methodological problem because both the gravity of the offence and the offender’s moral blameworthiness must be considered in an integrated manner to achieve proportionate sentences. Parliament defines proportionality in terms of both of these factors ( s. 718.1 of the Criminal Code ).

[ 168 ] While the starting-point approach allows for deviation to take into account the individual circumstances of the accused, this is an inadequate response. Sentencing judges ought not to prioritize one element of proportionality as a matter of law, even if the process as a whole ultimately accounts for both elements of proportionality. In Proulx , Lamer C.J. explained that offence‑based presumptions in sentencing are inconsistent with the sentencing methodology prescribed by Parliament because they lead to insufficient consideration of the moral blameworthiness of the offender:

My difficulty with the suggestion that the proportionality principle presumptively excludes certain offences from the conditional sentencing regime is that such an approach focuses inordinately on the gravity of the offence and insufficiently on the moral blameworthiness of the offender. This fundamentally misconstrues the nature of the principle. Proportionality requires that full consideration be given to both factors. [Emphasis in original; para. 83]

[ 169 ] I agree with Bennett J.A. of the Court of Appeal for British Columbia, dissenting, but not on this point:

The real difficulty with “starting point sentences” is that they tend to ignore, or at least lessen the importance of the circumstances of the offender and focus on the circumstances of the offence. This may result in sentences that are not proportional, as they do not sufficiently consider the moral blameworthiness of the offender.

( R. v. Agin , 2018 BCCA 133 , 361 C.C.C. (3d) 258, at para. 97 )

[ 170 ] Sentencing judges using a presumptive sentence crafted solely in relation to the gravity of the offence do not follow a truly individualized process:

It is submitted that though our system permits greater divergence in sentence it retains the undoubted virtue of placing the particular offence and the particular offender first in priority. This helps to keep sentencing human and minimize any tendency to devolve into a mechanical enterprise. It would be wrong, in our sentencing system, to make any single factor more important than the principle that sentence be appropriate to the particular offence and the individual offender. Sensitivity and flexibility in sentencing requires that the approach to be taken should flow from the facts of the case and not from any single rule, however useful or certain that rule may be. [Emphasis added.]

(C. C. Ruby, Sentencing (2nd ed. 1980), at pp. 423-24)

[ 171 ] Unlike starting points, ranges are not presumptions. Sentencing judges do not have to begin their analysis at the range and adjust the sentence to reflect the individualized circumstances of the accused. Ranges simply demonstrate the typical breadth of sentences for an offence. In considering the gravity of the offence and the moral blameworthiness of the offender together, they allow judges to give full effect to both of these elements.

[ 172 ] I find it telling that the Court of Appeal of Alberta considered that using starting points for young offenders was inappropriate because of the significant measure of individualization in sentencing a young person ( R. v. W. (C.W.) (1986), 1986 ABCA 47 , 43 Alta. L.R. (2d) 208, at pp. 212-13). This Court has now clearly stated that “sentencing is first and foremost an individualized exercise” for adult offenders as well ( Boudreault , at para. 58 ). If starting points are incompatible with individualization for young offenders, they should be for adult offenders as well.

(ii) Pre‑Weighing Factors by the Appellate Court

[ 173 ] By “building in” some factors to the starting point, the appellate court effectively prescribes the weight to be given to these factors by sentencing judges, displacing the sentencing judges’ discretion to determine their weight. A portion of the analysis is already conducted. Pre‑weighing factors directly encroaches on sentencing judges’ discretion: “The weighing of relevant factors, the balancing process is what the exercise of discretion is all about” ( McKnight , at para. 35 , cited with approval in Nasogaluak , at para. 46 ). Even where it is not done explicitly, as my colleagues propose (at para. 46), mitigating factors will always be implicitly “built in” to starting points, as the very notion of a starting point requires a baseline or archetypal offence and offender.

[ 174 ] Take “good character”, a factor built in starting points ( Arcand , at para. 132 ; Felix (ABCA), at para. 45). Absent a finding that the offender’s character was “even better than might be supposed”, a sentencing judge may not give it different or additional weight ( Arcand , at para. 135 ). Sentencing judges, not appellate courts, should determine what weight to give to this factor in light of the individual circumstances of the offender. Building this factor into the starting point interferes with sentencing judges’ ability and duty to consider all relevant circumstances in sentencing ( Stone , at para. 244 ). Moreover, building in “good character” assumes that it applies in the same way to all offenders and overlooks that good character has many degrees (Bloos and Renke, at p. 803). Such a methodology is incompatible with individualized sentencing.

[ 175 ] I acknowledge that some sentencing ranges have also incorporated mitigating factors such as prior good character into the sentencing range (see, e.g., R. v. H. (C.N.) (2002) , 62 O.R. (3d) 564 (C.A.), at para. 32 ; R. v. Voong , 2015 BCCA 285 , 374 B.C.A.C. 166; R. v. Cunningham (1996) , 104 C.C.C. (3d) 542 (Ont. CA), at p. 546). They do not represent the norm. In my view, building in such factors in ranges is also questionable having regard to individualization. But, in any case, building in factors in a range does not impede individualization in the same way as starting points. Sentencing judges retain the discretion to evaluate the weight of the good character of the accused (for example) within or even outside the range. Judges using the starting-point approach do not.

(iii) Starting Points Produce an Artificial Sentencing Methodology by Overemphasizing the Judicially Created Category

[ 176 ] Under the starting‑point approach, categorization is pivotal. For example, while “not always easy to draw” ( Felix (ABCA), at para. 51), the distinction between commercial and wholesale trafficking in cocaine is the difference between a three and four and one-half‑year starting point ( R. v. Rahime , 2001 ABCA 203 , 95 Alta. L.R. (3d) 237, at para. 18 ; R. v. Ma , 2003 ABCA 220 , 23 Alta. L.R. (4th) 14). If the same proportion applies to fentanyl, this elusive distinction will determine whether a six or nine‑year starting point is applicable. This improperly shifts the main focus from whether a sentence is just and appropriate for an offender to which judicially‑created category applies. In addition, the starting-point approach requires parties and judges to identify which factors can justify deviation from the starting point having regard to the factors that are already built in the hypothetical baseline offence. As M. R. Bloos and M. C. Plaxton say, starting points “subject the sentencing process to an unnatural compartmentalization of factors that either over‑complicates the process, or forces the sentencing judge to give certain factors more or less weight than is appropriate” (“Starting-Point Sentencing and the Application of Laberge In Unlawful Act Manslaughter Cases” (2003), 6 C.R. (6th) 346, at p. 352) .

[ 177 ] Sentencing judges should not be burdened with such artificial and convoluted debates, rules and categories. They should instead consider all relevant factors in an integrated manner and compare them to comparable cases to determine the just and appropriate sentence for the offender before the court (Bloos and Plaxton, at p. 352).

[ 178 ] Sentencing ranges do not raise similar concerns. Ranges for related categories typically overlap, while the jump from one starting point to the next is jagged. Ranges instruct sentencing judges to view the gravity of the offence and the responsibility of the offender as a matter of degree, on a continuum. As a result, as the intervener the Legal Aid Society of Alberta notes, ranges “rarely see judges bogged down in disagreements about whether an offence comes within a defined subcategory, or whether a fact aggravates or mitigates compared to some hypothetical baseline offence” (I.F., at para. 3). It is a methodology more conducive to individualized sentencing.

(iv) The “Clustering” Effect of Starting Points

[ 179 ] Commentators have noted that the starting-point approach has the “natural effect of bunching sentences around a median rather than spreading them across a range to suit individualized circumstances” (Manson (1997), at p. 282). This is the “clustering” effect, which is “antithetical to individualization: the more clustering we observe, the less likely it is that sentence outcomes reflect the unique nature of each case or are proportionate” (I. D. Marder and J. Pina‑Sánchez, “Nudge the judge? Theorizing the interaction between heuristics, sentencing guidelines and sentence clustering” (2020), 20 C.C.J. 399, at p. 401). By contrast, sentences spread along ranges can better reflect individualized circumstances.

[ 180 ] The parties contest whether “clustering” can be empirically established. However, I note that “clustering” effects in sentencing have been studied by several commentators (see, e.g., Marder and Pina‑Sánchez; M. W. Bennett, “Confronting Cognitive ‘Anchoring Effect’ and ‘Blind Spot’ Biases in Federal Sentencing: A Modest Solution for Reforming a Fundamental Flaw” (2014), 104 J. Crim. L. & Criminology 489, at pp. 523‑29 (finding that even advisory sentencing grids acted as a “hulking anchor for most judges”); D. M. Isaacs, “Baseline Framing in Sentencing” (2011), 121 Yale L.J. 426, at p. 426 (finding that “sentences disproportionately cluste[r] around the typical sentence in a typical crime baseline regime” (emphasis deleted)); C. Guthrie, J. J. Rachlinski and A. J. Wistrich, “Inside the Judicial Mind” (2001), 86 Cornell L. Rev. 777, at pp. 787‑94 (on the “anchoring” effect of numerical values more generally)). The anchoring effect is a well‑known cognitive bias: final judgments are disproportionately biased toward the starting point of a decision‑maker’s reasoning (A. Tversky and D. Kahneman, “Judgment under Uncertainty: Heuristics and Biases” (1974), 185 Science 1124, at p. 1128). While sentencing ranges can also produce clustering effects, this tends to be in jurisdictions where ranges are established by sentencing grids (e.g., various U.S. states) or by sentencing commissions (e.g., the United Kingdom) (Marder and Pina‑Sánchez; Bennett).

[ 181 ] The Crown points to sentences that deviate significantly from the starting point as evidence of the absence of clustering effect. In many of the cases it cites, however, the Court of Appeal intervened and varied restrained sentences that were crafted because of the individual circumstances of offenders and offences for not properly applying a starting point ( R. v. Corbiere , 2017 ABCA 164 , 53 Alta. L.R. (6th) 1; R. v. Giroux , 2018 ABCA 56 , 68 Alta. L.R. (6th) 21; R. v. L’Hirondelle , 2018 ABCA 33 ; R. v. Melnyk , 2014 ABCA 344 , 584 A.R. 238). Consequently, these cases do not stand for the proposition that the starting-point approach is compatible with individualized sentencing and do not refute the clustering effect.

[ 182 ] In addition, “clustering” also follows logically. For instance, it will be more difficult to impose a 90‑day sentence for an offender convicted of trafficking in cocaine or crack cocaine with strong rehabilitative potential and significant mitigating factors if the starting point is 3 years than if there is a 6 months to 4 years range of sentence (comp. R. v. Godfrey , 2018 ABCA 369 , 77 Alta. L.R. (6th) 213, and R. v. Zawahra , 2016 QCCA 871 ).

[ 183 ] While there is a paucity of empirical evidence, what there is reinforces the views that starting points are problematic in practice, as well as in principle.

(b) Starting Points Overemphasize Denunciation and Deterrence

[ 184 ] In theory, the Court of Appeal could set starting points to account for the various principles of sentencing. But this would overlook how starting points operate in practice. Starting points are often established to emphasize deterrence and denunciation and, in turn, to ensure more retributive punishment (Manson (1997), at p. 279, citing R. v. Sprague (1974) , 19 C.C.C. (2d) 513 (Alta. C.A.); Johnas ). This runs contrary to the objectives of the 1996 sentencing reform: “. . . reducing the use of prison as a sanction [and] expanding the use of restorative justice principles in sentencing” ( Gladue , at para. 48 ; s. 718.2 (e) of the Criminal Code ). As one author notes:

Restorative justice principles do not appear in any of the Alberta starting point decisions. Invariably, they describe offences as calling for sentences which emphasize deterrence and denunciation. While some offences and offenders deserve harsh sanctions, to create a judicial category of an offence, and to then remove from the sentencing matrix all reference to restorative justice principles is not aligned with the philosophy expressed by Parliament in the enactment of Part XXIII of the Criminal Code , particularly ss. 718.2 (d) and (e) . [Emphasis added.]

(P. L. Moreau, “Trouble for Starting Points?” (2021), 68 C.R. (7th) 129, at p. 135)

[ 185 ] Starting points make it more difficult for judges to give adequate weight to restorative justice principles because they are designed to be easy to move up and hard to move down. As mentioned, starting points generally build in the absence of a criminal record, prior good character, and the principle of restraint ( Arcand , at paras. 132‑36, 293 and 333 ). Moreover, “[t]he connection of the crime to addiction is not significantly mitigating” and “[c]ompliance with bail conditions [is] at best neutral” ( Godfrey , at para. 16 ). Of course, starting points are not minimum sentences ( Arcand , at para. 131 ). Still, they explicitly or implicitly foreclose reliance on multiple mitigating factors (Bloos and Renke, at p. 803; Manson (2001), at p. 68). In short, judges applying a sentencing tool that overemphasizes deterrence and denunciation and that forecloses reliance on multiple mitigating factors risk overlooking lower, appropriate sentences.

(c) The Starting-Point Approach Is Contrary to Parliament’s Direction for Sentencing Aboriginal Offenders

[ 186 ] Section 718.2(e) directs sentencing judges to pay “particular attention to the circumstances of Aboriginal offenders”. In Gladue , this Court explained that “the effect of s. 718.2 ( e ) is to alter the method of analysis which sentencing judges must use in determining a fit sentence for aboriginal offenders” (para. 93(5); Ipeelee , at para. 59 ). Gladue requires sentencing judges “to undertake the sentencing of aboriginal offenders individually, but also differently”, taking into account the systemic and background factors that bear on the culpability of the offender and the types of sanctions which might be appropriate because of the offender’s Aboriginal heritage or connection (para. 93(6); Ipeelee , at paras. 72‑73 and 75 ).

[ 187 ] Sentencing Aboriginal offenders is a highly individualized process that acknowledges that “the circumstances of aboriginal people are unique” ( Gladue , at para. 93 (6)). Therefore, methodologically, it would be an error for sentencing judges to determine an appropriate sentence for Aboriginal offenders by reference to a “typical” non‑Aboriginal offender:

If the offender is Aboriginal, then courts must consider all of the circumstances of that offender, including the unique circumstances described in Gladue . There is no sense comparing the sentence that a particular Aboriginal offender would receive to the sentence that some hypothetical non‑Aboriginal offender would receive, because there is only one offender standing before the court. [Emphasis added.]

( Ipeelee , at para. 86 )

[ 188 ] “[C]omparing the sentence that a particular Aboriginal offender would receive to the sentence that some hypothetical non‑Aboriginal offender would receive” is precisely the starting-point methodology, applied to Aboriginal offenders. The starting‑point approach reduces the holistic Gladue analysis to a matter of minor subtraction from the starting point (in some instances) (see, e.g., Corbiere and Arcand ). This is an inherent problem with the starting-point approach. The failure to give proper effect to the Gladue methodology constitutes an error justifying appellate intervention ( Ipeelee , at para. 87 ).

[ 189 ] Unsurprisingly given this improper sentencing methodology, the starting‑point approach appears to deflect Alberta courts from the different process of sentencing set out in Gladue . For instance, in Arcand , the Court of Appeal of Alberta was almost silent on the significance of Mr. Arcand being an Aboriginal person and to the impact of Gladue (Rudin, at p. 1007; for other cases that show how starting points work to frustrate Gladue considerations, see L’Hirondelle ; Corbiere ; Giroux ; R. v. Wilson , 2009 ABCA 257 , 9 Alta. L.R. (5th) 283; R. v. Huskins , 2018 ABPC 227 ; R. v. Soosay , 2017 ABQB 478 ). Professor T. Quigley also noted that appellate judges “armed with the starting point approach and the fetish against disparity” were less open to conditional sentences and sentencing circle dispositions (“Are We Doing Anything about the Disproportionate Jailing of Aboriginal People?” (1999), 42 Crim. L.Q. 129, at p. 144).

[ 190 ] This is not a coincidence. Rather, it is a consequence of the starting point approach, which is driven by the “typical offender” and only adjusts for individual offenders and which tends to increase the rates of incarceration (Manson (1997), at pp. 280-83). The current approach is contrary to Parliament’s direction for a highly individualized process that takes fully into account the unique circumstances of Aboriginal offenders.

(d) The Starting-Point Approach Does Not Provide Sufficient Guidance to Sentencing Judges

[ 191 ] In addition to constraining discretion, starting points are paradoxically not as effective as sentencing ranges to provide adequate guidance to sentencing judges in important respects. First, starting points do not inform the normal range of variation. Ranges are more illustrative because they emerge from a series of precedents and show the typical breadth of possible sentences. By contrast, as even a proponent of starting point sentencing accepts, starting points provide little guidance about “the degree or quantum of effect generated by aggravating or mitigating factors” (P. Moreau, “In Defence of Starting Point Sentencing” (2016), 63 Crim. L.Q. 345, at p. 355).

[ 192 ] Second, unlike ranges, starting points cannot “reflect the range of factual situations in the world and the plurality of judicial perspectives” and “embody the collective experience and wisdom of the judiciary” ( Friesen , at para. 33 ). The absence of illustrating precedents can make starting points more mechanical in their application. I agree with Professor Ruby’s critique of the Court of Appeal of Alberta’s approach to precedent:

The Alberta Court of Appeal — alone amongst all the provincial courts of appeal — has developed a doctrine where only cases, even from their own court, that the Court itself has designated as sentencing precedents can be used in argument by counsel. . . .

However, this insistence that some precedents from the Court of Appeal are simply not to be used runs counter to the statutory mandate in section 718.2 (b) of the Criminal Code , which requires that a court take into consideration the principle that “a sentence should be similar to sentences imposed on similar offenders for similar offences committed in similar circumstances”. The Code does not say that a similarity with some sentences, but not all relevant ones, is sufficient. Indeed, the rule seems to be a violation of that principle. It is an instruction suitable for children: “Do as we say, not as we do”. [Emphasis added.]

((2020), at §§ 4.22 and 4.23)

[ 193 ] Thus, starting points limit sentencing judges’ ability to draw from a broad pool of experience which, in turn, runs counter to the principle of parity.

(3) Rigid Appellate Intervention Is Inevitable

[ 194 ] The last relevant feature of the starting-point approach is its aggressive enforcement by appellate courts. Sentencing judges who do not commence sentencing from a starting point or who deviate significantly from the starting point can expect to be overturned ( R. v. Ilesic , 2000 ABCA 254 , 89 Alta. L.R. (3d) 299, at para. 6 ; Arcand , at paras. 116‑17 ; Ruby (2020), at § 4.12). In my view, reiterating the settled standard of appellate review is insufficient to address this. This Court has consistently reminded provincial appellate courts not to vary sentences unless they are demonstrably unfit or tainted by a material error in principle ( Lacasse , at para. 11 ; Friesen , at para. 26 ). Yet the Court of Appeal of Alberta continues to rely on starting points to circumvent this standard. Indeed, by faithfully adhering to starting-point methodology, it is difficult for them to do otherwise.

[ 195 ] In McDonnell , this Court disapproved of the interventionist approach of the Court of Appeal of Alberta in relation to starting points. It stated that “it can never be an error in principle in itself to fail to place a particular offence within a judicially created category . . . for the purposes of sentencing” because deference should be shown to sentencing judges’ decisions (para. 32). In Proulx , it reiterated that starting points can only be set out as “guides” to lower courts (para. 86).

[ 196 ] Despite this strong language, “the Alberta Court of Appeal refuses to view starting points as mere guidance” (Ruby (2020), at § 23.15). The Court of Appeal has made it clear that since starting points are designed to minimize disparity in sentencing, sentencing judges are not free to ignore them or to pay lip service to them ( Arcand , at para. 118 ; R. v. Innes , 2012 ABCA 283 , at para. 5 ). It has reminded sentencing judges that although they can deviate from the starting point, “caution is in order” because “[f]acts relied on to deviate from the starting point should be relevant to sentence and reasonably justify deviation ” ( Arcand , at para. 106 ). In other words, it has treated starting points as binding. As O’Brien and Hunt JJ.A. noted in their separate reasons in Arcand , the majority did not “adhere completely to Supreme Court authority” and effectively elevated starting points to a “rule of law” (paras. 302(a) and 352).

[ 197 ] Even after Lacasse , where this Court reaffirmed that appellate courts may not intervene because they would have put the sentence in another category (at para. 51), the Court of Appeal of Alberta has continued say that failure to place an offence within the correct category constitutes a reviewable error (see, e.g., R. v. Reddekopp , 2018 ABCA 399 , 79 Alta. L.R. (6th) 215, at para. 5 ; Godfrey , at para. 6 ) and to refer to its jurisprudence establishing starting points as “binding” ( Godfrey , at para. 15 ; R. v. J.A.S. , 2019 ABCA 376 , at paras. 12‑13 ; Hajar , at para. 160-61 ; D.S.C. , at para. 40 ). In Parranto , the Court of Appeal said that departure from the starting-point approach constitutes an error and that sentencing judges cannot treat starting points as “merely hortatory or suggestive” ( 2019 ABCA 457 , 98 Alta. L.R. (6th) 114, at paras. 28‑29 ). This does not respect the authority of McDonnell , Lacasse and other decisions from this Court.

[ 198 ] Sentencing judges have stated that the Court of Appeal of Alberta’s approach restrains their discretion and their ability to craft individualized sentences or to depart from the starting-point approach despite the Supreme Court’s clear pronouncements (see, e.g., R. v. Moriarty , 2016 ABPC 25 , 34 Alta. L.R. (6th) 110, at para. 145 ; R. v. Boriskewich , 2017 ABPC 202 , 62 Alta. L.R. (6th) 194, at paras. 51‑52 ).

[ 199 ] The relationship between the standard of review and starting points was reiterated, again, in Friesen : “. . . starting points cannot be binding in either theory or practice, and appellate courts cannot interpret or apply the standard of review to enforce them, contrary to R. v. Arcand . . . ” (para. 37 ).

[ 200 ] Yet, at least in some cases, the Court of Appeal of Alberta continues to treat starting points as binding and as constraining sentencing judges’ discretion to consider all relevant factors. For instance, in R. v. Wakefield , 2020 ABCA 352 , it found it was an error in principle to proceed from the wrong starting point and varied the sentence (para. 26 , see also R. v. Roberts , 2020 ABCA 434 , 17 Alta. L.R. (7th) 255, at paras. 26 and 50 ; R. v. Morton , 2021 ABCA 29 , at para. 25 ). This Court’s calls for deference and individualized sentencing have not been heeded.

[ 201 ] The continued rigid enforcement of starting points by the Court of Appeal of Alberta makes clear the fundamental incompatibility between the starting-point approach and a deferential standard of review. The Court of Appeal of Alberta has identified several avenues to circumvent the standard of review because of starting points: it intervenes if the sentencing judge failed to advert to the prescribed starting point ( R. v. Tran , 2010 ABCA 317 , 490 A.R. 229, at para. 16 ; Giroux , at para. 12 ); if the sentence fails to give sufficient weight to deterrence and denunciation, as recorded in the starting point ( Godfrey , at para. 17 ); if the sentencing judge considered a factor already built in the starting point, which leads to double counting ( R. v. G.B. , 2013 ABCA 93 , 544 A.R. 127, at para. 5 ; R. v. Brodt , 2016 ABCA 373 , 46 Alta. L.R. (6th) 213, at para. 7 ); and if the sentence departs significantly from the starting point ( Corbiere , at paras. 25‑27 ). Departures from starting points facilitate appellate intervention in comparison to sentencing ranges:

. . . it is much easier to upset a sentence of, say, 15 months’ imprisonment when the starting point is 3 years than it is when the range is 12 months to 3 years. The departure from the low end of the range may be seen on appellate review as acceptable while the departure from the starting point is more readily adjudged to be reversible error.

( Gashikanyi , at para. 35 )

[ 202 ] In effect, the starting-point approach multiplies potential “errors” and, in turn, increases the scope for the court of appeal to vary sentences.

[ 203 ] The time is past due to deal decisively with the methodological problems inherent in starting points. Those flaws are structural. They cannot be cured by repeating, yet again, exhortations relating to the standard of review. This should come as no surprise since the purpose of the starting-point methodology is appellate control and this purpose is embedded in how it is designed to operate. Accordingly, the undermining of proper appellate review of sentencing is not a misapplication of starting-point methodology. Rather, it is the fulfilment of its purpose. Thus, the only effective response, in my view, is to say that the starting-point methodology should no longer be used.

III. Disposition

[ 204 ] I would disavow the starting-point approach to sentencing. I agree with Justice Moldaver and adopt his reasons for the disposition of the appeals and the additional guidance he provides.

The reasons of Abella and Karakatsanis JJ. were delivered by

Karakatsanis J. —

[ 205 ] I have read the reasons of my colleagues Brown and Martin JJ. I agree with their analysis, from paras. 9 to 54, that starting points are a permissible form of appellate sentencing guidance, provided that starting points are not used to curtail the highly deferential sentencing standard of appellate review. However, I do not agree with their disposition to dismiss these two appeals. Despite the emphasis on the starting-point approach in the parties’ submissions, the merits of these appeals ultimately stand or fall on the well-settled principles of appellate deference in sentencing. Sentencing is an art, not a science, and appellate courts must show great deference to the trial judge’s exercise of discretion in crafting a fit, individualized sentence. Unless the trial judge made an error in principle that impacted the sentence or the sentence was demonstrably unfit, appellate intervention is not justified: R. v. Lacasse , 2015 SCC 64 , [2015] 3 S.C.R. 1089, at paras. 39-44 and 52-55 .

[ 206 ] In my view, neither trial judge made an error in principle, nor was either sentence demonstrably unfit. In both appeals, the Court of Appeal did not act with restraint and deference but rather assumed a scrutineering, interventionist posture. The trial judges were faulted for their reasonable exercises of discretion, their factual findings were disregarded, and their demonstrated appreciations of the gravity of the offences before them were ignored. Both appeals should therefore be allowed with the original sentences restored.

[ 207 ] My colleagues set out the facts of each appeal in detail and I do not propose to re-state those facts. Suffice to say the appellants were both very prolific and sophisticated fentanyl traffickers. Felix received a global sentence of 7 years ( 2019 ABQB 183 ), while Parranto received a global sentence of 11 years ( 2018 ABQB 863 ).

[ 208 ] As I will explain, their respective sentencing reasons show that both trial judges appreciated the very grave nature of the offences committed by the appellants. If the sentences ultimately imposed were more lenient than the gravity of the offences would suggest, this was not because the trial judges took lenient views of wholesale fentanyl trafficking. Rather, it was because each trial judge reasonably exercised their discretion to place great weight on mitigating factors and rehabilitative sentencing principles. It was not open to the Court of Appeal — and it is not open to this Court — to reweigh those factors or to second-guess those principles.

I. Decisions Below

A. R. v. Felix, 2019 ABQB 183 (Burrows J.)

[ 209 ] The trial judge in Felix’s case spoke in clear terms of the dangers of fentanyl and the seriousness of Felix’s offences:

Mr. Felix’s crimes involve two very dangerous illicit drugs — fentanyl and cocaine. Both are highly addictive. Their use can cause serious physical and mental injury and even death to the user. A fentanyl user takes a particularly serious risk. Even a relatively modest dose can be an overdose and can easily cause death.

The illicit use of these dangerous drugs impacts not only on the user but also on the rest of society in a many ways. The already strained resources of the health care system must respond to the physical and mental injury sustained by users. Drug users often finance their addictions through crime. In addition, violent crimes are often committed by those active in the drug trade in order to protect sales territories or to retaliate when drug debts are not paid. The victims of these crimes are often innocent bystanders.

Drug trafficking crimes are fundamentally different than other crimes. Both the criminal and the immediate victim, the user, are willing participants in the crime. No one involved resists the commission of the offence. This feature magnifies the already very significant seriousness of drug offences. [paras. 37-39 ]

[ 210 ] The trial judge clearly recognized that fentanyl trafficking was a very serious offence.

[ 211 ] Similarly, the trial judge did not minimize Felix’s high degree of culpability. He found that Felix’s moral blameworthiness was “very significantly aggravating” and “toward the high end” of the moral blameworthiness range: paras. 41‑42.

[ 212 ] However, the trial judge considered a number of strong mitigating factors. He found that Felix’s rehabilitation prospects were “extremely promising”: para. 48. He emphasized that Felix had no criminal record and a positive background, including operating a legitimate company that worked to rebuild and restore Fort McMurray in the aftermath of the 2016 wildfire: paras. 31 and 44. The trial judge also noted Felix’s “very solid and widespread support among his friends, family members, work colleagues, and neighbors” (para. 44), and referred to an “impressive collection” of 17 positive reference letters (para. 35). Finally, he noted the reduced mitigating effect of Felix’s guilty plea that, while late, did avoid a lengthy trial: para. 43.

[ 213 ] The trial judge then turned to the sentencing range that could be deduced from Alberta’s fentanyl trafficking jurisprudence. He identified a five- to seven-year range but noted that, given Felix’s high moral blameworthiness, his sentence “should be placed at least at the top” of that range: paras. 79-80. The trial judge also considered placing Felix within a higher range of five to nine and a half years but did not do so as this “would not give appropriate weight to the mitigating factors . . . or to the fact that the other two persons accused with Mr. Felix received sentences of 5 years”: para. 81.

[ 214 ] The trial judge concluded his reasons with a discussion of sentencing objectives. He emphasized the need to “strongly denounce” Felix’s conduct with a sentence that would “constitute a strong deterrence to others”: para. 82. However, he balanced the emphasis on denunciation and deterrence with the mitigating factors he had identified and Felix’s rehabilitative prospects, including his commitment to “socially positive pursuits”, and imposed a seven-year sentence: paras. 83-84.

B. R. v. Felix, 2019 ABCA 458 , 98 Alta L.R. (6th) 136 (Antonio J.A., Paperny, Watson, Slatter and Crighton JJ.A. concurring)

[ 215 ] The Court of Appeal overturned Felix’s seven-year sentence. In the Court of Appeal’s view, the trial judge chose “incorrect” comparator cases in deducing the five- to seven-year range: para. 72. The Court of Appeal identified three errors stemming from the trial judge’s “direct handling of the precedents he considered” (paras. 72-73):

• failing to adequately distinguish commercial from wholesale trafficking when selecting comparators;

• failing to account for the accused’s role in the organization, in particular by treating sentences imposed on his “runners” as direct comparators; and

• double-counting the mitigation of the guilty plea, by using it to discount the sentence relative to precedents that had already been discounted by guilty pleas.

[ 216 ] The Court of Appeal further held that “[t]hese errors resulted in a sentence that [was] demonstrably unfit”: para. 73.

[ 217 ] The Court of Appeal then sentenced Felix afresh in light of the nine-year starting point it had set. It held that Felix’s role at the top of his organization was a “weighty aggravating factor”: para. 75. In stark contrast to the trial judge’s findings, the Court of Appeal found that there was “little to offer in mitigation”: para. 76. The Court of Appeal found reference letters that the trial judge called an “impressive collection” (para. 35) to be of “little mitigating value”, containing “inaccurate assumptions about Mr. Felix’s character and contribution to the community” (para. 77). In sum, the Court of Appeal found that a fit sentence would have been 13 years but reduced this to 10 years; the sentence the Crown sought at trial: paras. 79, 82.

C. R. v. Parranto, 2018 ABQB 863 (Ouellette J.)

[ 218 ] Like the trial judge in Felix’s case, the trial judge in Parranto’s case was clearly aware of the dangers of fentanyl, noting its addictive nature, potential to cause death, and the rise of fentanyl-related deaths in recent years: paras. 49-52 and 66-67 . He found that the primary focus needed to be denunciation and deterrence: para. 66. As the trial judge stated:

The fact that at least one person dies every day from a fentanyl-related death requires condemnation of the people who are in the business of selling fentanyl. [para. 66]

[ 219 ] Once again, the trial judge clearly viewed wholesale fentanyl trafficking to be a very grave offence.

[ 220 ] In terms of Parranto’s moral culpability, the trial judge found him to be a “middleman in wholesale fentanyl trafficking” because, while Parranto was not on the “low end” of moral culpability, he was less morally culpable than those producing or importing fentanyl: para. 69.

[ 221 ] Like in Felix’s case, the trial judge in Parranto’s case deduced a range of five to seven years for wholesale fentanyl trafficking: para. 65.

[ 222 ] The trial judge then identified a number of aggravating and mitigating circumstances. He found it aggravating that Parranto had a criminal record and committed his second set of offences while on release for the first set: paras. 77-79. He found one “significant” mitigating factor in the form of Parranto’s early guilty pleas and the remorse they showed: paras. 80-81. He also found that Parranto’s Métis heritage was a mitigating factor and that Parranto’s addiction was relevant to sentence, even if it was not a “true mitigating circumstance”: paras. 82-83.

[ 223 ] The trial judge then turned to the task of crafting a fit sentence for Parranto. The trial judge’s approach was clear and methodical with respect to the weight he afforded to each relevant factor. He first determined what would be fit sentences for each set of offences charged without regard to mitigating circumstances and the totality principle. For the first set of offences, the trial judge found that a fit sentence would be nine years: seven years for the fentanyl trafficking, two years consecutive for the firearms charges, with concurrent sentences for the breaches of the court order and recognizance: paras. 88-89. For the second set of offences, the trial judge found that a fit sentence would be 11 years: 8 years for fentanyl trafficking, 3 years consecutive for the firearms charges, and again, concurrent sentences for the breaches of the court order and recognizance: para. 90.

[ 224 ] Therefore, the notional global sentence, before considering mitigation and totality, was a 20-year prison sentence.

[ 225 ] The trial judge then turned to the mitigating circumstances and totality. He took up the Crown’s initial suggestion to reduce the sentence by one-third in light of the remorse shown by Parranto’s guilty pleas, leaving a sentence of 13.2 years: para. 92. He reduced this to 12 years given Parranto’s Métis heritage and addiction: paras. 93-94. Finally, the sentence was reduced by another year on the basis of totality, leaving an 11-year global sentence: paras. 95-96.

D. R. v. Parranto, 2019 ABCA 457 (Watson J.A., Paperny, Slatter, Crighton and Antonio JJ.A. concurring)

[ 226 ] The Court of Appeal overturned Parranto’s sentence, finding that the trial judge made a “series of interrelated errors”, with the “principal error aris[ing] in his handling of the starting point approach”: para. 24. The Court of Appeal further held that the sentence was demonstrably unfit as it did not “serve the objectives of denunciation, deterrence . . . or protection of the public”: para. 25.

[ 227 ] The Court of Appeal identified four errors in principle in the trial judge’s reasons. First, the trial judge’s finding that knowledge of potential harm should not be imputed to Parranto was dismissed as a “matter of conjecture”: paras. 46-47. Second, the trial judge’s use of Parranto’s addiction to reduce the sentence was critiqued for resting on a “questionable” evidential basis: paras. 51-52. Third, the Court of Appeal held that there was “no clear foundation” for the trial judge to use Parranto’s Métis heritage as a mitigating factor: para. 53. Fourth and finally, the Court of Appeal found a number of errors in the trial judge’s totality analysis. Totality, the Court of Appeal said, is a “concept” rather than a “principle”, and as such “does not outweigh sentencing principles”: para. 54. According to the Court of Appeal, the trial judge erred by using totality “multiple times”, resulting in “free rides” for certain offences: para. 56.

[ 228 ] In the end, the Court of Appeal increased Parranto’s global sentence to the 14 years initially sought by the Crown at trial: paras. 68-69.

II. Analysis

[ 229 ] The principles governing appellate sentencing review are well-settled. Appellate courts can only intervene if the trial judge has erred in principle in a way that impacted the sentence or if the sentence was demonstrably unfit: Lacasse , at paras. 43-44 and 52-55 .

[ 230 ] Errors in principle include errors of law, the failure to consider a relevant factor, the erroneous consideration of aggravating or mitigating factors, and unreasonable weighing or balancing of different factors: R. v. Friesen , 2020 SCC 9 , at para 26 . A trial judge has not erred in principle simply because the appellate court would have weighed the relevant factors differently: Lacasse , at para. 49 . Appellate intervention on the basis of the weight or emphasis given to different factors will only be justified if the trial judge exercised their discretion unreasonably: R. v. Nasogaluak , 2010 SCC 6 , [2010] 1 S.C.R. 206, at para. 46 , citing R. v. McKnight (1999) , 135 C.C.C. (3d) 41 (Ont. C.A.), per Laskin J.A. Similarly, where the Criminal Code , R.S.C. 1985, c. C-46, does not dictate otherwise, the sentencing judge has discretion over which sentencing objectives in s. 718 to prioritize, such as denunciation, deterrence, and rehabilitation, and how much weight to afford to the secondary sentencing principles in s. 718.2 , such as parity and restraint: Lacasse , at paras. 54-55 ; Nasogaluak , at para. 43 .

[ 231 ] It is also not an error in principle to use the “wrong” sentencing range. The choice of a sentencing range is a matter that falls within the trial judge’s discretion and an appellate court cannot intervene just because it would have used a different range: Lacasse , at para. 51 ; R. v. Lloyd , 2016 SCC 13 , [2016] 1 S.C.R. 130, at para. 52 .

[ 232 ] Even if an error in principle is found, this does not necessarily permit appellate intervention. Deference must still be shown to the original sentence unless the error in principle impacted the sentence imposed: Lacasse , at paras. 43-44 .

[ 233 ] In the absence of any errors in principle that impacted the sentence, an appellate court can only intervene if the sentence imposed is demonstrably unfit. A sentence is demonstrably unfit where it constitutes an unreasonable departure from the fundamental principle that a sentence must be proportionate to the gravity of the offence and the degree of responsibility of the offender: Lacasse , at paras. 53-55 ; Criminal Code , s. 718.1.

[ 234 ] A sentence is not demonstrably unfit simply because it falls outside of a particular sentencing range. A proportionate sentence must reconcile the principles of individualization and parity: the trial judge must calibrate a sentence that is proportionate for this offence by this offender, while also being consistent with sentences for similar offences in similar circumstances: Lacasse , at para. 53 . However, parity is a secondary sentencing principle, subordinate to proportionality ( Lacasse , at para. 54 ), and cannot “be given priority over the principle of deference to the trial judge’s exercise of discretion”: R. v. L.M. , 2008 SCC 31 , [2008] 2 S.C.R. 163, at para. 35 . As LeBel J. explained in L.M. , “[t]he principle of parity does not preclude disparity where warranted by the circumstances : para. 36 (emphasis in original), citing F. Dadour, De la détermination de la peine: principes et applications (2007), at p. 18.

[ 235 ] However, sentences that significantly depart from an established sentencing range or starting point may signal potential fitness concerns. Sentencing ranges and starting points are tools to provide busy trial judges with a jurisprudential view of the gravity of the offence. Sentencing ranges provide trial judges with “historical portraits” of how past judges have applied the principles and objectives of sentencing in similar circumstances: Lacasse , at para. 57 . Similarly, a properly applied starting point represents appellate guidance regarding the gravity of a given type of offence: Brown and Martin JJ.’s reasons, at paras. 20 and 44. When a sentence significantly departs from a range or starting point, there is an understandable concern that the trial judge may have exercised their discretion unreasonably in some way.

[ 236 ] A significant deviation does not necessarily reflect a demonstrably unfit sentence. As this Court stated in Lacasse , deviation from a range cannot justify appellate intervention “unless the sentence that is imposed departs significantly and for no reason from the contemplated sentences. Absent an error in principle, an appellate court may not vary a sentence unless the sentence is demonstrably unfit”: para. 67. Even though the starting point reflects appellate guidance regarding the gravity of the offence, where there is a significant deviation, appellate review must be conducted in accordance with the ordinary sentencing standard of review. If, for example, the deviation results from an unreasonably lenient view of the gravity of the offence, or from unreasonably minimizing the offender’s moral blameworthiness, then the sentence will be an unreasonable departure from the proportionality principle and will be demonstrably unfit. If, on the other hand, the deviation stems from heavy but reasonable emphasis on mitigating factors and rehabilitative principles, appellate intervention is not justified.

[ 237 ] In other words, whether a sentence is demonstrably unfit is a qualitative rather than a quantitative assessment. What matters is whether the trial judge imposed a proportionate sentence by reasonably appreciating the gravity of the offence and the degree of responsibility of the offender in the specific circumstances of the case: Lacasse , at paras. 58 and 67 ; Nasogaluak , at para. 44 ; L.M. , at para. 36 . If the trial judge has done so then appellate intervention is not justified, regardless of how far the sentence deviates from a range or starting point: Lacasse , at paras. 53 and 67 ; Brown and Martin JJ.’s reasons, at paras. 29-30 and 38.

III. Application

[ 238 ] In my view, the Court of Appeal was not justified in intervening in either case. None of the purported errors in principle are borne out on a fair reading of the trial judge’s reasons and neither sentence was demonstrably unfit. I begin with Felix’s appeal and then turn to Parranto’s appeal.

A. Felix

[ 239 ] In my view, the trial judge did not err in principle. The Court of Appeal identified three errors in principle in the trial judge’s analysis but all three stem from the trial judge’s purportedly “incorrect” identification of the five- to seven-year range: paras. 72-73. Similarly, my colleagues say the trial judge should have used the “more accurate” 8- to 15-year range: Brown and Martin JJ.’s reasons, at para. 68. However, the range chosen was within the trial judge’s discretion and the Court of Appeal could only intervene if the sentence was demonstrably unfit: Lacasse , at para. 51 ; Lloyd , at para. 52 .

[ 240 ] Nor was the seven-year sentence demonstrably unfit. As I have explained, the trial judge demonstrated his appreciation of the severity of Felix’s offences, commenting on the dangers of fentanyl and the seriousness of the drug trafficking: paras. 37-39. He also recognized Felix’s “significantly aggravating” degree of moral blameworthiness and culpability (paras. 40-42) and the need to “strongly denounce” and deter this conduct (para. 82). In my view, it cannot be said that the trial judge took such a lenient view of wholesale fentanyl trafficking or minimized Felix’s culpability to such a degree that the sentence was an unreasonable departure from the proportionality principle: see Lacasse , at para. 53 . Rather, the reason the sentence imposed was seven years — and not longer — was because of the trial judge’s weighty but reasonable emphasis on the mitigating factors and rehabilitative principles: paras. 48 and 83. That other judges may have weighed those factors differently does not justify appellate intervention.

[ 241 ] My colleagues suggest that “[i]t is clear the sentencing judge misapprehended the gravity of the offence”: Brown and Martin JJ.’s reasons, at para. 67. However, such a conclusion has no foundation in the trial judge’s reasons. Rather, it seems to stem solely from the trial judge’s failure to use the “more accurate” range that my colleagues would have used. Just as the choice of a range itself cannot be an error in principle and does not render a sentence demonstrably unfit, it cannot be said that the choice of the “wrong” range demonstrates a misapprehension of the gravity of the offence. To find otherwise, in my view, is directly contrary to this Court’s continuous line of authority in L.M. , Nasogaluak , Lacasse , and Friesen .

[ 242 ] The Court of Appeal did not take issue with the trial judge’s appreciation of the gravity of Felix’s offences. Where the opinion of the Court of Appeal differed from that of the trial judge was with regard to the mitigating factors. The trial judge found multiple strong mitigating factors, including Felix’s “extremely promising” prospects of rehabilitation: para. 48. Conversely, the Court of Appeal found that there was “little to offer in mitigation”: para. 76. In the absence of errors in principle though, it was not open to the Court of Appeal to reweigh these mitigating factors. The trial judge’s sentence, even if it significantly departed from the “correct” range or starting point, represented a reasonable appreciation of the gravity of the offence and Felix’s degree of responsibility.

[ 243 ] In sum, while the original sentence certainly could have been longer, the trial judge made no errors in principle that impacted the sentence and it was not demonstrably unfit. I would therefore allow the appeal and restore the original sentence.

B. Parranto

[ 244 ] As with Felix, the Court of Appeal and my colleagues take issue with the trial judge’s selection of a five- to seven-year range in Parranto’s case: C.A. reasons, at para. 30; Brown and Martin JJ.’s reasons, at para. 77. Again, the choice of a range was within the trial judge’s discretion. This choice is not itself an error in principle, nor does it indicate a misapprehension of the gravity of the offences.

[ 245 ] In my view, the sentence was not demonstrably unfit. The trial judge appreciated the gravity of Parranto’s offences and his moral blameworthiness. The trial judge’s initial notional sentence for the fentanyl offences, before considering mitigating factors and totality, was 15 years. This lengthy notional sentence represents an appropriately grave view of grave offences.

[ 246 ] Indeed, neither the Court of Appeal nor my colleagues appear to take issue with the trial judge’s notional sentence. Rather, their concerns arise with what it was reduced to after the trial judge considered the mitigating factors and totality.

[ 247 ] The Court of Appeal took issue with the trial judge’s use of three mitigating factors: (i) Parranto’s lack of knowledge of the harms of fentanyl; (ii) Parranto’s addiction; and (iii) Parranto’s Métis heritage (paras. 46-47 and 51-53). However, all three critiques impermissibly intrude upon the trial judge’s factual findings.

[ 248 ] First, the trial judge found that “it would be wrong to impute knowledge of potential harm to Mr. Parranto”: para. 68. The Court of Appeal disagreed, dismissing this finding as a “matter of conjecture”: para. 47. In the absence of palpable and overriding error though, the Court of Appeal was required to defer to the trial judge’s factual findings and was not entitled to disagree.

[ 249 ] Furthermore, the Court of Appeal found that the trial judge erred in using Parranto’s addiction as a mitigating factor because of its “questionable” evidentiary basis: para. 52. Again, this was not open to the Court of Appeal. The trial judge made a factual finding that Parranto suffered from addiction: paras. 48, 83 and 93. The Court of Appeal was obliged to respect that finding absent palpable and overriding error.

[ 250 ] Similarly, the Court of Appeal faulted the trial judge for using Parranto’s Métis heritage as a mitigating factor: para. 53. This seemed to be rooted in the Court of Appeal’s concern that the defence did not request a Gladue report at trial: para. 5 . However, the Crown conceded that Parranto’s Métis heritage was a mitigating factor (trial reasons, at para. 82) — a fair concession given this Court’s direction to take judicial notice of the systemic factors that affect Indigenous persons: R. v. Ipeelee , 2012 SCC 13 , [2012] 1 S.C.R. 433, at paras. 59‑60 . The trial judge evidently found there was a sufficient factual basis to make this finding even in the absence of a Gladue report. Once again, it was not open to the Court of Appeal to disagree.

[ 251 ] Finally, the Court of Appeal identified an error in principle in the trial judge’s totality analysis. However, the Court of Appeal began from the erroneous premise that totality is a “concept” that “does not outweigh sentencing principles”: para. 54. On the contrary, totality is indeed a sentencing principle — it is a function of the fundamental principle of proportionality that is engaged when consecutive sentences are imposed: R. v. M. (C.A.) , [1996] 1 S.C.R. 500, at para. 42 ; Friesen , at paras. 157-58 . Different judges may have approached totality differently in this case but that does not mean the trial judge erred in his analysis, nor that the sentence was demonstrably unfit.

[ 252 ] Again, Parranto certainly could have received a longer sentence but that is not the test. The trial judge did not err in principle and the sentence was not demonstrably unfit. The original sentence should be restored.

IV. Conclusion

[ 253 ] For these reasons, I would allow both appeals and restore the original sentences.

Appeals dismissed, Abella and Karakatsanis JJ. dissenting.

Solicitors for the appellant Cameron O’Lynn Parranto : Moreau & Company, Edmonton.

Solicitors for the appellant Patrick Douglas Felix : Advocate Law, Red Deer, Alta.

Solicitor for the respondent: Public Prosecution Service of Canada, Halifax.

Solicitor for the intervener the Attorney General of Manitoba: Attorney General of Manitoba, Winnipeg.

Solicitor for the intervener the Attorney General of Alberta: Justice and Solicitor General, Appeals, Education & Prosecution Policy Branch, Edmonton.

Solicitors for the intervener the Criminal Trial Lawyers’ Association: Pringle Chivers Sparks Teskey, Vancouver.

Solicitors for the intervener the Canadian Civil Liberties Association: McKay Ferg, Calgary.

Solicitor for the intervener the Aboriginal Legal Services: Aboriginal Legal Services, Toronto.

Solicitor for the intervener the Legal Aid Society of Alberta: Legal Aid Society of Alberta, Edmonton.

Solicitors for the intervener Association québécoise des avocats et avocates de la défense: Caissy et Marceau-Bouchard, avocats, Amqui, Que.

[1] Criminal Code , s. 95 and CDSA s. 5(3) ( a )(i)(C)

[2] Contrary to Body Armour Control Act , S.A. 2010, c. B-4.8