COURT OF APPEAL FOR ONTARIO
CITATION: R. v. Aravena, 2015 ONCA 250
DATE: 20150416
DOCKET: C53435, C53310, C57090, C53786, C57574
Doherty, Watt and Pardu JJ.A.
BETWEEN
Her Majesty the Queen
Respondent
and
Marcelo Aravena, Brett Gardiner, Wayne Earl Kellestine, Frank John Mather, Dwight Mushey
Appellants
Counsel:
Delmar Doucette and Daniel Santoro, for Marcelo Aravena
Christopher D. Hicks and Aliki Yorgiadis, for Brett Gardiner
Michael Dineen, for Wayne Earl Kellestine
Dirk Derstine and Janani Shanmuganathan, for Frank John Mather
James L. D. Miglin, for Dwight Mushey
John Corelli, Alexander Alvaro and Holly Loubert, for the respondent
Heard: September 15, 16, 17, 18, 2014
On appeal from the convictions entered on October 29, 2009 by Justice Thomas A. Heeney of the Superior Court of Justice, sitting with a jury, and the sentences imposed on October 30, 2009.
Doherty and Pardu JJ.A.:
Background
[1] During the evening and early morning of April 7-8, 2006, eight members of the Toronto Bandidos motorcycle gang were shot and killed on a farm property owned by Wayne Kellestine. Michael Sandham, Frank Mather, Marcelo Aravena, Brett Gardiner, Dwight Mushey, and Wayne Kellestine were all charged with eight counts of first degree murder. After a lengthy trial, the jury convicted Sandham, Mushey and Kellestine of eight counts of first degree murder. The Crown had alleged that these three had actually perpetrated the murders. The jury convicted Frank Mather and Marcelo Aravena of manslaughter on the first homicide and seven counts of first degree murder on the other homicides. The jury convicted Brett Gardiner of manslaughter on the first two homicides and first degree murder on the other six counts. It was the Crown’s position that Mather, Aravena and Gardiner had aided and abetted in the murders.
[2] Kellestine, Mushey, Mather, Aravena and Gardiner appeal their convictions. Sandham initially appealed, but has abandoned his appeal.
[3] It was the Crown’s position that the massacre at Kellestine’s farm was the culmination of internal strife within the Bandidos motorcycle gang organization. The American Bandidos were unhappy with the Toronto chapter. They believed that the Toronto chapter was not complying with directions from American headquarters. Kellestine was a dissident member of the Toronto chapter. The American Bandidos ordered Kellestine to “pull the patches” (expulsion from the organization akin to excommunication) from the rest of the members of the Toronto chapter.
[4] The Winnipeg group of Bandidos was also unhappy with the Toronto chapter. The Toronto chapter had refused the Winnipeg group full status. Acting on instructions from the United States, Sandham, Aravena, Gardiner, Mushey and M.H.[^1] travelled from Winnipeg to Kellestine’s farm in southwest Ontario to assist in the “pulling of the patches.” Mather, a friend of Kellestine, was staying with Kellestine at his farm.
[5] On the Crown’s theory, Kellestine persuaded eight members of the Toronto Bandidos to come to his farm under the pretence of a meeting to discuss Bandidos business. While waiting for the arrival of the Toronto Bandidos, the “farm team”, consisting of the Winnipeg Bandidos, Kellestine and Mather, prepared for the arrival of the Toronto Bandidos by assembling and cleaning firearms, and putting on gloves. The Manitoba licence plates were removed from a car parked in the yard, so that the Toronto Bandidos would not suspect that the Winnipeg group was at the farm. Aravena was responsible for opening the gate to let each of the four cars containing the Toronto Bandidos into Kellestine’s fenced yard. Mather, Aravena and Gardiner were to stay in the farmhouse. Kellestine established radio communication between the farmhouse and the barn, and placed Mather and Gardiner in charge of monitoring a police scanner located in the house.
[6] It was understood that “pulling the patches” of the Toronto Bandidos was likely to be a violent affair. The Toronto members referred to themselves as “the No Surrender Crew”, meaning they would never voluntarily give up their club patches. During the preparations before the arrival of the Toronto Bandidos, Kellestine, on more than one occasion, told the others that, “[i]f we kill one we kill them all.”
[7] Sandham hid in the loft of the barn awaiting the arrival of the Toronto Bandidos. Mushey and M.H. hid outside of the barn. Kellestine escorted the eight Toronto Bandidos into the barn. Gunfire erupted. Sandham, from his position in the loft and wearing a bullet proof vest, shot and killed Raposo, the first victim (count one). Kellestine also opened fire, injuring several of the other Toronto Bandidos. He ordered them to the ground. Mushey and M.H., armed with firearms, rushed into the barn when they heard the shots. Mather and Aravena, armed with a shotgun and a baseball bat, respectively, also rushed to the barn to help when they heard the shots.
[8] Kellestine, with the help of the others who stood guard, confined the Toronto Bandidos in the barn, confiscating their weapons, keys, cell phones and identification. Kellestine pretended there were “guys on the roof” ready and willing to shoot should the Toronto Bandidos put up any opposition. Aravena helped Kellestine maintain this ruse.
[9] In the ensuing hours, the remaining Toronto Bandidos were escorted out of the barn one by one, placed in a car and shot, execution style. Kellestine, Sandham, and Mushey were the shooters.
[10] All of the appellants participated in extensive efforts to clean up the farm, destroy evidence and transport the bodies and vehicles to other locations.
[11] It was the position of the Crown at trial that all of the murders were first degree murders either because they were planned and deliberate, or alternatively, with the exception of the first homicide, because they were committed in the course of an unlawful confinement of the victims.
Grounds of Appeal
[12] Counsel for the various appellants diligently advanced several grounds of appeal. The court required submissions from the Crown on five of the grounds of appeal argued on behalf of the appellants. These reasons will address only those five grounds of appeal. They are:
• Did the trial judge err in law in holding that the common law defence of duress was not available to persons charged as parties to a murder?
• Did the trial judge err in holding that there was an evidentiary basis upon which Mather and Aravena could be convicted as aiders or abetters under the constructive first degree murder provision?
• Did the trial judge err in holding that there was an evidentiary basis upon which the jury could find that Gardiner aided and abetted in the first degree murders alleged in counts three to eight?
• Did the trial judge err in his instructions to the jury as to the potential evidentiary effect of Aravena’s post-offence conduct?
• Did the trial judge err in holding that M.H. was a confidential informant in respect of his dealings with the Winnipeg police and could not be questioned about those activities?
Issue #1: Duress
THE DURESS RELATED ISSUES AT TRIAL
[13] Counsel for the Crown and the accused requested a pre-trial ruling on the availability of duress as a defence to murder. The trial judge ruled that duress was not available as a defence to murder regardless of whether an accused was alleged to be a perpetrator or a party to the murder: R. v. Sandham (2009), 70 C.R. (6th) 203 (Ont. S.C.). In a second ruling, the trial judge held that duress was available as a defence to the included offence of manslaughter.
[14] After the evidence was completed and during pre-charge submissions, counsel for Aravena and Mather argued that the jury should be instructed on the defence of duress as it applied to the included offence of manslaughter on at least some of the murder charges. Counsel for Gardiner did not seek a duress instruction as it applied to manslaughter.
[15] The trial judge held that because there was no evidence that Mather, who did not testify, was threatened by anyone, there could be no “air of reality” to a duress defence for him on any of the charges. With respect to Aravena, who did testify, the trial judge held that there was some evidence of an implied threat directed at Aravena by Kellestine. He further held, however, that Aravena had voluntarily placed himself in a situation in which he knew he could be compelled by Kellestine to commit criminal acts. The trial judge held that in putting himself in that position, Aravena had lost any potential claim to a duress defence on the included offence of manslaughter on any of the murder charges.
[16] In his necessarily lengthy jury instructions, the trial judge made brief reference to the threat allegedly made against Aravena by Kellestine. The trial judge told the jury:
Duress is not available as a matter of law to a charge of murder, and for other legal reasons I need not get into, it is entirely irrelevant to this case. In a nutshell, it is not open to anyone to say to an innocent victim “you will die so that I can live”.
THE DURESS ARGUMENTS ON APPEAL
[17] None of the appellants argue that the trial judge erred in holding that on the evidence adduced at trial, there was no air of reality to a duress defence for any appellant on the included offence of manslaughter on any of the charges. However, Aravena, Mather and Gardiner submit that the trial judge erred in law in his pre-trial ruling that the common law defence of duress was not available on the murder charges.
[18] The Crown did not allege that the appellants, Aravena, Mather and Gardiner, had actually committed any of the murders. The Crown contended that each were aiders and abettors in the murders. These appellants submit that the common law defence of duress is available in Canada to persons charged as parties to a murder, and that the trial judge, in holding to the contrary, misread binding authority from the Supreme Court of Canada and ignored binding authority from this court.
[19] Aravena, Mather and Gardiner further argue that the pre-trial ruling precluding reliance on the defence of duress prejudiced them at trial. Mather and Gardiner contend that the trial could well have taken a different evidentiary course had the trial judge correctly ruled on the pre-trial motion that duress was a potential defence to persons charged with aiding and abetting a murder. Aravena’s argument in support of his prejudice claim is different. Aravena submits that by willingly placing himself under Kellestine’s control in respect of certain criminal activities (unlawful confinement), he may have foreclosed the defence of duress as it related to manslaughter, but the defence remained available on the murder charges.
[20] In response, the Crown submits that the trial judge properly held that the common law defence of duress is not available on a charge of murder regardless of the nature of the alleged participation in the murder. The Crown further submits that even if the defence was available as a matter of law to alleged aiders and abetters to a murder, the trial judge’s ruling caused no prejudice to Gardiner, Mather or Aravena. The Crown notes that all three had a full opportunity to develop a duress defence as it related to manslaughter, if so inclined.
[21] Gardiner made no attempt to develop the defence and did not suggest at trial that there was any evidence to support a duress defence. Mather led no evidence of any threats that could possibly give the defence an air of reality. The Crown contends that Gardiner and Mather cannot overcome the absence of any air of reality to the duress defence in the trial record by inviting this court to speculate as to what might have happened had the trial judge ruled that the defence was available on the murder charges. Crown counsel contends that it is fanciful to suggest that Mather or Gardiner, having chosen to lead no evidence of a duress defence in response to manslaughter, would have advanced a duress defence on the murder charges. Finally, Crown counsel observes that there is still nothing in either the trial record or the appeal record to suggest any air of reality to a duress claim advanced by Gardiner or Mather.
[22] With respect to Aravena, the Crown’s no substantial wrong argument goes as follows. First, the trial judge was wrong to hold that the evidence gave an air of reality to Aravena’s claim of implied threats of a sufficient nature to invoke a duress defence. The Crown contends that the evidence simply does not support that conclusion. Second, the Crown submits that the trial judge’s reasons for holding that Aravena’s association with Kellestine rendered the duress defence unavailable on the manslaughter charges are equally applicable to the murder charges.
[23] In sum, the Crown submits the accused suffered no substantial wrong or miscarriage of justice even if the duress defence was available as a matter of law on the murder charges. The Crown submits that, on the evidence, there was no air of reality to that defence.
THE DEFENCE OF DURESS
[24] Ever since R. v. Paquette, 1976 CanLII 24 (SCC), [1977] 2 S.C.R. 189, there have been two different duress defences in Canada. The statutory duress defence set out in s. 17 of the Criminal Code applies to persons who have “actually committed” an offence and whose liability rests on s. 21(1)(a) of the Criminal Code. We will refer to these individuals as perpetrators. The common law defence of duress, preserved by s. 8(3) of the Criminal Code, applies to persons whose liability flows from s. 21(1)(b) or (c) (aider or abettor) or s. 21(2) (common design). We will refer to these individuals as parties.[^2]
[25] The anomaly, if not confusion, created by Paquette has been significantly mitigated by the recent duress jurisprudence. The statutory language of the duress defence in s. 17 has undergone significant constitutional surgery: see R. v. Ruzic, 2001 SCC 24, [2001] 1 S.C.R. 687. The common law defence and what remains of the statutory defence have also been reworked to more effectively reflect their mutual underlying rationale: see R. v. Hibbert, 1995 CanLII 110 (SCC), [1995] 2 S.C.R. 973; R. v. Ryan, 2013 SCC 3, [2013] 1 S.C.R. 14, at para. 44. As the law presently stands, the essential elements of the two duress defences are “largely the same”: Ryan, at para. 81.
[26] The elements of both the statutory and the common law defences of duress are helpfully summarized in Ryan, at para. 81.
• There must be an explicit or implicit threat of present or future death or bodily harm. The threat can be directed at the accused or a third party.
• The accused must reasonably believe that the threat will be carried out.
• There is no safe avenue of escape. This element is evaluated on a modified objective standard.
• A close temporal connection between the threat and the harm threatened.
• Proportionality between the harm threatened and the harm inflicted by the accused. The harm caused by the accused must be equal to or no greater than the harm threatened. This is also evaluated on a modified objective standard.
• The accused is not a party to a conspiracy or association whereby the accused is subject to compulsion and actually knew that threats and coercion to commit an offence were a possible result of this criminal activity conspiracy or association.
[27] If the defence is available in law, and the evidence gives an air of reality to the defence, an accused is entitled to an acquittal unless the Crown disproves one or more of the essential elements of the defence on a reasonable doubt standard: D. Stuart, Canadian Criminal Law, 6th ed. (Toronto: Carswell, 2011), at p. 493; see also R. v. Perka, 1984 CanLII 23 (SCC), [1984] 2 S.C.R. 232, at pp. 257-58.
THE DEFENCE OF DURESS AND MURDER
Is the Law Settled?
[28] Section 17 expressly provides that the statutory defence of duress “does not apply” to the offences listed in the section. Murder is one of those offences.[^3] Subject to a successful constitutional challenge[^4], the perpetrator of a murder cannot avoid culpability even if he or she committed murder in the face of a true “kill or be killed” option.
[29] The trial judge, after an analysis of Paquette and an extensive review of common law authorities from other jurisdictions, determined that the common law defence of duress was not available to persons charged with murder regardless of whether they were charged as perpetrators or parties. It was, therefore, unnecessary for him to consider the constitutionality of the murder exception in s. 17: Sandham, at para. 124. As the Crown alleged that all of the appellants who raise duress on appeal were parties to the murders, the constitutionality of the murder exemption in s. 17 is not directly in issue on this appeal.
[30] An examination of the common law defence of duress in Canada begins with Paquette. Paquette was concerned with liability for murder under s. 21(2) of the Criminal Code. A man was shot and killed during a robbery. The Crown contended that the accused had joined in a common design to commit robbery and that he knew or ought to have known that the others might intentionally cause bodily harm while carrying out the robbery. The accused alleged that he only assisted the others in the robbery because his life was threatened.
[31] Martland J., for the entire court, relied on Lynch v. Director of Public Prosecutions for Northern Ireland, [1975] A.C. 653 (H.L. Eng.). In Lynch, a slim majority had accepted that the common law defence of duress was available to a person charged as a party to a murder. After extensive citation from Lynch, Martland J. embraced the majority position, at pp. 196-97:
I am in agreement with the conclusion reached by the majority that it was open to Lynch, in the circumstances of that case, to rely on the defence of duress, which had not been put to the jury. If the defence of duress can be available to a person who has aided and abetted in the commission of murder, then clearly it should be available to a person who is sought to be made a party to the offence by virtue of s. 21(2).
[32] Although Paquette turned on liability through s. 21(2) of the Criminal Code and not liability as an aider or abettor under s. 21(1)(b) or (c), several Canadian appellate courts, including this court, have treated Paquette as holding that the common law defence of duress is available to persons charged as aiders and abettors to a murder: R. v. Ruzic (1998), 1998 CanLII 5265 (ON CA), 128 C.C.C. (3d) 97 (Ont. C.A.), at para. 49, aff’d 2001 SCC 24, [2001] 1 S.C.R. 687; R. v. Curran (1977), 1977 ALTASCAD 284, 38 C.C.C. (2d) 151 (Alta. C.A.), at pp. 154-55, leave to appeal refused, [1978] 1 S.C.R. xi; R. v. Hartford and Frigon (1979), 1979 CanLII 520 (BC CA), 51 C.C.C. (2d) 462 (B.C. C.A.), at p. 467; R. v. Pelletier (1986), 1986 CanLII 1179 (BC CA), 29 C.C.C. (3d) 533 (B.C. C.A.), at p. 542; R. v. Amarales, 2008 ONCA 692, 237 C.C.C. (3d) 148, at paras. 87-88; R. v. Earhart, 2010 ONCA 874, 272 C.C.C. (3d) 475, leave to appeal refused, [2011] S.C.C.A. No. 397.[^5]
[33] Laskin J.A., writing for the court in Ruzic, summarized the holding in Paquette, at para. 49:
But the court held that s. 17 applies only to persons who commit offences themselves, that is, as principals. It does not apply to persons who commit offences as parties. Persons charged as parties can rely on the common law defence which, in Canada, does not exclude any offence from its reach. [Emphasis added.]
[34] As the statutory defence applied in Ruzic, the comments of Laskin J.A. as to the reach of the common law defence are obiter. The Supreme Court of Canada in Ruzic made no reference to our colleague’s obiter and expressly refrained from any comment on the offences exempted from the statutory defence of duress: Ruzic, at para. 19.
[35] In Hibbert, the accused was charged as an aider to an attempted murder and ultimately convicted of aggravated assault. He relied on the common law defence of duress and the trial judge left the defence with the jury. While the primary importance of Hibbert lies in its explanation of the difference between the mens rea required of a party to an offence and the defence of duress, Lamer C.J., for the court, appeared to confirm the holding in Paquette that the common law defence of duress is available to a person charged as an aider and abettor in a murder. He said, at para. 38:
Finally, I am satisfied that the interpretation of the mens rea for liability under s. 21(1)(b) that I am proposing will not result in unjust convictions in cases involving coercion by threats of death or bodily harm, since in these cases the common law defence of duress will remain available to the accused.
[36] After overruling Paquette as it relates to the mens rea required for party liability, Lamer C.J. said, at para. 44:
I hasten to point out, however, that overturning this holding in Paquette does not affect the validity of that case’s first aspect, namely, that the common law defence of duress continues to apply in cases involving party liability under s. 21 of the Code. Furthermore, it can be noted in passing that, on the facts of Paquette, the accused’s acquittal could well have been supported on the basis of the excuse provided by the common law defence of duress rather than on the notion that his intention to assist in the commission of the robbery was “negated” by duress. [Emphasis added.]
[37] The trial judge, at para. 67, read the above passage in this way:
It is important to note that Lamer C.J. was referring to robbery, not murder, in suggesting that the common law defence of duress would be available.
[38] We do not agree with that interpretation. The reference by the Chief Justice to “the accused’s acquittal” in the passage from Hibbert must refer to Paquette’s acquittal on the murder charge. Paquette was only charged with murder. He was not charged with robbery.
[39] Hibbert’s apparent acceptance of duress as a defence to a charge of attempted murder fortifies our view that Hibbert confirms Paquette’s holding that the common law defence of duress is available to those charged with aiding or abetting a murder. The difference between the murderer and the attempted murderer lies only in achievement of the intended goal. No meaningful distinction can be drawn between the person who, under threat, aids in a murder and the person who, under the same threat with the same intent, engages in the same conduct yet, for reasons that may have nothing to do with his effort or his intent, does not end up assisting in a killing: see R. v. Gotts, [1992] 2 A.C. 412 (H.L. Eng.), at pp. 425-26; David Ormerod, Smith and Hogan’s Criminal Law, 13th ed. (Oxford: Oxford University Press, 2011), at pp. 361-62.
[40] On the authorities as they stood at the time of the trial judge’s reasons, we think he erred in holding that he was not bound by those authorities to recognize that the common law defence of duress was available to persons charged as parties to a murder: see R. v. P.C., 2012 ONSC 5362, 99 C.R. (6th) 116. However, subsequent to the trial judge’s decision, the Supreme Court delivered its reasons in Ryan. After outlining the elements of the statutory and common law defences of duress, and to a large degree reconciling the two defences, the court turned briefly to the remaining differences between the statutory defence and the common law defence. The court, at para. 83, identified two differences, the second of which speaks directly to the issue at hand:
The second [difference] is that the statutory version of the defence has a lengthy list of exclusions whereas it is unclear in the Canadian common law of duress whether any offences are excluded. This results in the rather incoherent situation that principals who commit one of the enumerated offences cannot rely on the defence of duress while parties to those same offences, however, can.
We accordingly leave to another day the questions of the status of the statutory exclusions and what, if any, exclusions apply at common law. [Emphasis added.]
[41] Ryan continues the modification and reconciliation of the common law and statutory duress defences. In doing so, Ryan leaves open the question of whether any offences should be excluded entirely from the reach of those defences. In the face of that direction, it falls to this court to answer the question left open in Ryan, through reference to basic criminal law principles, thejuridical rationale underlying the duress defence, the elements of the duress defence as outlined in the controlling jurisprudence, the fundamental principles enshrined in the Charter, and the common law authorities from other jurisdictions.
What Should the Common Law Be?
[42] Our analysis proceeds from three premises. First, the duress defence to be considered is the defence as it has been described and defined in Hibbert, Ruzic and Ryan. As explained in Ruzic, at paras. 58-59, the competing individual interests of the accused and the victim and the very real public order concerns engaged when a duress defence is advanced require that the defence be subject to “reasonable, but strict standards”. The analysis assumes that the defence proffered meets those “strict standards”; see also Perka, at pp. 250-51.
[43] Second, the analysis assumes that the person advancing the defence has the full mens rea required of an aider and abettor to murder. That is, the person knows that the perpetrator intends to kill the victim and he acts intending to assist the perpetrator in the killing: see R. v. Kirkness, 1990 CanLII 57 (SCC), [1990] 3 S.C.R. 74, at p. 88; R. v. Briscoe, 2010 SCC 13, [2010] 1 S.C.R. 411, at paras. 16-18. In short, regardless of the nature of the assistance provided, the party advancing the defence fully intends to assist in the murder of an innocent third party.
[44] Third, on this analysis, there are only two potential solutions. Either duress is a full defence, resulting in an acquittal and subject to the normal requirement that if the evidence gives rise to the defence the Crown must disprove the defence, or duress is not a defence to murder.
[45] Various law reform bodies and academic commentators have considered other choices, including recognizing duress as a partial defence that would reduce murder to manslaughter (as in the case of provocation), and reversing the burden of proof as it relates to the defence: see, e.g., The Law Commission, Murder, Manslaughter and Infanticide (No. 304) (London, UK: The Law Commission, 2006), at pp. 111, 116. Recognizing duress as a partial defence to murder is a legislative remedy that goes beyond the scope of judicial interpretation of the common law. No other jurisdiction recognizes the common law defence of duress as a partial defence. As to reversing the burden of proof, there are obvious constitutional implications to that approach. The now extensive duress and necessity jurisprudence from the Supreme Court of Canada does not suggest any reversal of the burden of proof. This court cannot take the bold step of reversing the burden absent submissions offering persuasive grounds for doing so: see R. v. Stone, 1999 CanLII 688 (SCC), [1999] 2 S.C.R. 290, at paras. 179-80, per Bastarache J., and at para. 48, per Binnie J., dissenting.
[46] The analysis begins with a bedrock principle of the criminal law. Criminal liability requires voluntary conduct. Voluntariness reflects individual choice and with choice comes responsibility for one’s actions. Society can properly hold individuals to account if they choose to act. Where there is no choice, punishment of the actor is incompatible with individual autonomy and arguably serves no utilitarian value. Not surprisingly, voluntariness has been described as a principle of fundamental justice protected by s. 7 of the Charter: see Ruzic, at para. 43; R. v. Daviault, 1994 CanLII 61 (SCC), [1994] 3 S.C.R. 63, at pp. 91-92; R. v. Luedecke, 2008 ONCA 716, 236 C.C.C. (3d) 317, at para. 53.
[47] Voluntariness as a criminal law term is most often used in its narrow physical sense to refer to a willed and volitional act. The condemned man who walks to his death acts voluntarily in the narrow physical sense of the word. He wills his legs forward to meet his hangman rather than require another to carry him to his end. This narrow physical concept of voluntariness does not assist persons who act under duress. Those persons, like the condemned man, engage in willed volitional conduct. They choose to engage in criminal conduct rather than suffer the harm threatened: Ruzic, at para. 44.
[48] Voluntariness has been extended beyond physical voluntariness to explain the rationale underlying the defences of necessity and duress. In Perka, at pp. 248-49, Dickson J. described necessity as a common law defence that could excuse what was otherwise criminal conduct. He referred to the rationale for excusing criminal conduct as resting:
[O]n a realistic assessment of human weakness, recognizing that a liberal and humane criminal law cannot hold people to the strict obedience of laws in emergency situations where normal human instincts, whether of self-preservation or of altruism, overwhelmingly impel disobedience.
[49] Relying on G. Fletcher, Rethinking Criminal Law (Boston: Little, Brown, 1978), Dickson J., at p. 249, linked the operation of the necessity defence as an excuse to the voluntariness requirement:
Literally this voluntariness requirement simply refers to the need that the prohibited physical acts must have been under the conscious control of the actor. Without such control, there is, for purposes of the criminal law, no act. The excuse of necessity does not go to voluntariness in this sense. The lost alpinist who on the point of freezing to death breaks open an isolated mountain cabin is not literally behaving in an involuntary fashion. He has control over his actions to the extent of being physically capable of abstaining from the act. Realistically, however, his act is not a “voluntary” one. His “choice” to break the law is no true choice at all; it is remorselessly compelled by normal human instincts. This sort of involuntariness is often described as “moral or normative involuntariness”.
[50] In Hibbert, at para. 54, the court embraced the Perka analysis as applicable to the common law defence of duress, holding that conduct that fell within the scope of the common law defence should be viewed as normatively involuntary and, therefore, not subject to criminal liability. Ruzic built on the rationale developed in Perka and Hibbert, distinguishing, at paras. 32-41, between moral involuntariness and moral blameworthiness. A person’s conduct could be morally involuntary in the sense that she had no realistic choice, but that does not necessarily make her conduct morally blameless. The duress defence excuses, it does not justify.
[51] In Ruzic, at para. 47, LeBel J., for the court, accepted that the concept of moral or normative involuntariness, like physical involuntariness, was a principle of fundamental justice protected by s. 7:
Although moral involuntariness does not negate the actus reus or mens rea of an offence, it is a principle which, similarly to physical involuntariness, deserves protection under s. 7 of the Charter. It is a principle of fundamental justice that only voluntary conduct – behaviour that is the product of a free will and controlled body, unhindered by external constraints – should attract the penalty and stigma of criminal liability. Depriving a person of liberty and branding her with the stigma of criminal liability would infringe the principles of fundamental justice if the accused did not have any realistic choice.
[52] As explained in Ruzic, at paras. 34-47, the concept of moral involuntariness as a principle of fundamental justice rests on an acceptance of individual autonomy and choice as essential preconditions to the imposition of criminal liability. If, on a reasonable assessment of the circumstances, it must be said that an accused had no realistic choice but to act as he did, fundamental justice requires that the accused not be branded a criminal for so acting. Society may regret or even deplore the accused’s failure to “rise to the occasion”, but it cannot, in a criminal justice system predicated on individual autonomy, justly criminalize and punish conduct absent a realistic choice: see Perka, at pp. 249-50. Ryan, at para. 40, confirmed the status of moral involuntariness as a principle of fundamental justice protected by s. 7 of the Charter.
[53] Moral involuntariness does not depend exclusively on an individual’s perception that she had no realistic choice but to act as she did. Moral involuntariness is measured, in part, “on the basis of society’s expectation of appropriate and normal resistance to pressure”: Perka, at p. 259. Professor Yeo offers this helpful description of “moral involuntariness”:
The term “moral involuntariness” alludes to the nature of the concept as well as to how and why it operates to exculpate a defendant of the crime charged. Beginning with “involuntariness”, this embodies a “fundamental organizing principle of our criminal law”, which is “[t]he treatment of criminal offenders as rational, autonomous and choosing agents”. In the context of a plea of necessity, a defendant is pleading that the circumstances placed him or her in a predicament that forced him or her to make a choice when, in different circumstances, the choice would not otherwise have been made. Turning to the adjective “moral” appearing in the term, this stipulates that social policy and values form an integral part of this concept. The inquiry raised by “moral involuntariness” is therefore whether, taking into account all relevant social policy considerations, the circumstances that impacted on the defendant’s ability to choose freely a course of action, rendered him or her not criminally responsible.[^6]
[54] The “social policy considerations” infused into the concept of moral voluntariness recognize that excuses which provide a defence for what is otherwise criminal behaviour cannot be entirely subjective. Excuses must take into account a variety of societal concerns, including the need to maintain social order and to protect persons who are the innocent victims of those who act under threats from others: Ruzic, at para. 58.
[55] Societal policy concerns germane to the assessment of moral involuntariness manifest themselves in the elements of the duress defence as described in Ruzic and Ryan. The no safe avenue of escape criterion and the requirement of a close temporal connection between the threat and the harm threatened serve to restrict the defence to circumstances in which the absence of a realistic choice manifests itself in objectively discernable circumstances. Referring to these two requirements, the court in Ryan observed, at para. 52:
The addition of [these] common law requirements to replace the now defunct immediacy and presence elements of s. 17 thus act to temper the once purely subjective belief as to the threat. Furthermore, they bring the statutory provision in line with the principle of moral involuntariness. Considering that society’s opinion of the accused’s actions is an important aspect of the principle, it would be contrary to the very idea of moral involuntariness to simply accept the accused’s subjective belief without requiring that certain external factors be present.
[56] Social policy concerns also pervade the proportionality requirement of the duress defence: see Yeo, at pp. 26-27. In Ryan, at para. 54, the court described proportionality as “a crucial component of the defence of duress” which “derives directly from the principle of moral involuntariness”; see also Paciocco, at p. 286.
[57] The proportionality requirement ultimately separates those society is prepared to excuse for yielding to threats from those society decides should not have succumbed to the pressure. Proportionality in the context of the duress defence is assessed on a “modified objective” standard: Ryan, at para. 72.
[58] Proportionality is the product of two different but related evaluations. First, the harm threatened must be equal to or greater than the harm inflicted in response to the threat. Second, the accused’s choice to inflict harm must “accord with what society expects from a reasonable person similarly situated in that particular circumstance”: Ryan, at para. 73.
[59] In many circumstances, the first evaluation will also answer the second, in that the imposition of greater harm than the harm threatened will inevitably fail to meet society’s expectations. In some circumstances, however, the second evaluation may be determinative. Take for example a man who inflicts bodily harm on a young child to avoid similar bodily harm to himself. The harm inflicted may be no greater than the harm avoided, but a jury may decide that a reasonable person similarly situated would have resisted the threat and refused to inflict the harm on the child. If the jury so concluded, the duress defence would fail.
[60] The criminal association exclusion further defines and limits the concept of moral involuntariness as applied to duress. A person who by association with a criminal organization or criminal endeavour voluntarily assumes the risk of being compelled by threats to engage in criminal conduct cannot rely on that compulsion to excuse his criminal conduct. Ryan explains at para. 77:
This is consistent with the principle of moral involuntariness. An accused that, because of his or her criminal involvement, knew coercion or threats were a possibility cannot claim there was no safe avenue of escape, nor can he or she truly be found to have committed the resulting offence in a morally involuntary manner.
[61] Moral involuntariness as a principle of fundamental justice is a reflection of the central importance of individual autonomy and choice in the imposition of criminal liability. The statutory and common law defences of duress, as developed in Hibbert, Ruzic and Ryan, take moral voluntariness as both the rationale for the duress defence and as the justification for the strict limits on the availability of the defence. In doing so, the court has adhered to its oft repeated admonition that the common law must be developed and applied in accordance with the values and principles enshrined in the Charter: see, e.g., R. v. Rodgers, 2006 SCC 15, [2006] 1 S.C.R. 554, at para. 18; R. v. Golden, 2001 SCC 83, [2001] 3 S.C.R. 679, at para. 86.
[62] In the same way that the principle of moral involuntariness has played a central role in shaping the elements of the common law defence of duress, it must figure prominently in any assessment of whether any offences, and in particular murder, should be excluded from the reach of that defence. The Crown’s claim that the common law defence of duress does not reach persons charged as parties to murder can be accepted only if the Crown can demonstrate either that the exclusion of murder is compatible with the principle of moral involuntariness or that the exclusion is a justifiable limitation on that principle.
[63] Whether excluding murder from the duress defence is consistent with the principle of moral involuntariness turns on the rationale underlying the defence and the interaction between that rationale and the proportionality requirement. Duress operates as an excuse and not a justification. While a person told to “kill or be killed” could perhaps never justify killing the innocent third party as the lesser of two evils, it is much more difficult to assert that the two harms are not of “comparable gravity”: R. v. Latimer, 2001 SCC 1, [2001] 1 S.C.R. 3, at para. 31.
[64] Because duress operates as an excuse, harms of comparable gravity satisfy at least the first arm of the proportionality inquiry. The principle of moral involuntariness recognizes that the criminal law “is designed for the common man, not for a community of saints or heroes”: Ruzic, at para. 40. The criminal law cannot demand acts of heroism, but must instead set standards of conduct “which ordinary men and women are expected to observe if they are to avoid criminal responsibility”: R. v. Howe, [1987] A.C. 417 (H.L. Eng.), at p. 430; see also Lynch, at p. 670.
[65] Moral involuntariness by its very nature demands a fact-intensive inquiry. Neither prong of the proportionality requirement will always favour sacrificing one’s own life over assisting in the murder of another. Choosing to aid in the murder of another will not always amount to choosing an evil greater than the evil threatened. For example, a person may be presented with a choice between taking the life of an innocent third party and the killing of her own child. The putative victims are equally innocent. Surely, the harms flowing from either choice are “of comparable gravity”.
[66] Nor, on any realistic view, can a decision to assist in a murder to avoid a harm threatened always fall below the standard of what “society expects from a reasonable person similarly situated in [the] particular circumstance”: Ryan, at para. 73. Consider, for example, a person who had no connection to the Bandidos or to the meeting at Kellestine’s farm, but who happened to attend at the farm for some innocent purpose that night. Assume Kellestine’s group took him captive and held him in the barn while Kellestine removed and murdered two of the victims. If that person was then ordered under threat of death by Kellestine to assist in the removal and murder of the next victim, would society expect the ordinary (not the heroic or exceptional) person to refuse Kellestine’s order and give up his own life? Could it be said that the person had “a realistic choice”? We think not.
[67] In taking the view that the duress defence could excuse a choice to assist in a murder, we of course do not suggest that the defence would be readily made out. Indeed, the duress defence is strictly limited and is not easily accessible in any situation. Moreover, the greater the harm caused by yielding to the threat, the more difficult it will be for an accused to make out the defence, particularly on the proportionality element. If an accused chooses to assist in a murder, it may well be that nothing short of a threat of immediate death to that person or some other person could ever satisfy the proportionality requirement.[^7]
[68] The trial judge accepted that a per se exclusion of murder from duress would result in the conviction and punishment of morally involuntary conduct. He said, at para. 85:
However, it is clear that in both the defence of necessity and the defence of duress, there is the additional requirement of proportionality. Even where an act is morally involuntary, the law will not excuse a person from committing an act that is disproportionate to the harm with which he is threatened.
[69] The trial judge continued, at para. 110:
[D]enying a duress defence for murder could result in an accused being convicted for a crime that is “morally involuntary”, but that results from the fact that killing an innocent person fails the test of proportionality, which is a separate part of the defence. It is grounded in the victim’s right to life that is guaranteed in s. 7 of the Charter. To the extent that the rights of the victim are given priority to those of the accused, this results from the balancing of the interests of the accused as against the protection of society.
[70] The trial judge treated the proportionality requirement in the duress defence as operating outside of the principle of moral involuntariness. On his analysis, the defence captured only conduct that is both morally involuntary and proportional. Proportionality required a balancing of the competing rights of the victim and the accused. In the case of murder, that balancing always favoured the victim.
[71] The trial judge did not have the benefit of the reasons in Ryan. In those reasons, the court made it clear that the proportionality requirement in the duress defence “derives directly from the principle of moral involuntariness”: Ryan, at para. 54. The proportionality requirement serves, along with other elements of the defence, to identify conduct that is morally involuntary and therefore incapable of supporting a criminal conviction.
[72] As the trial judge acknowledged, on his approach, a separate proportionality analysis that balanced the rights of an accused against the victim’s right to life would permit the criminalization of morally involuntary conduct. Presumably on this approach, a victim’s right to life could equally permit the criminalization of physically involuntary conduct.
[73] In our view, while the victim’s right to life is a crucial factor in assessing whether an accused’s conduct is proportional and therefore morally involuntary, the victim’s right to life cannot render conduct that is otherwise involuntary punishable under the criminal law. A proportionality requirement which looks to whether an accused had any realistic choice is an essential component of the moral involuntariness inquiry.
[74] In holding that duress is not available as a defence to persons charged with murder, the trial judge relied on a long line of English authorities reaching back over 200 years. With the solitary exception of the majority judgment in Lynch, English authorities have consistently excluded murder from the common law defence of duress: see, e.g., Blackstone, Commentaries on the Laws of England, vol. 4 (Oxford: Clarendon Press, 1769), at p. 30; G. Williams, Criminal Law: The General Part, 2d ed. (London: Stevens & Sons Limited, 1961), at pp. 759-61; R. Card, Card, Cross and Jones: Criminal Law, 17th ed. (Oxford: Oxford University Press, 2006), at pp. 789-91; R. v. Howe (1986), [1987] A.C. 417 (H.L. Eng.); Abbott v. The Queen, [1977] A.C. 755 (P.C.); R. v. Hasan, [2005] UKHL 22, [2005] 2 A.C. 467.
[75] The English authorities provide two main reasons for denying the duress defence to persons charged with murder. First, these authorities argue that permitting duress as a defence to murder strengthens the hand of terrorists and other criminal organizations. Allowing duress as a defence to murder encourages criminals to use threatened intermediaries as a means of conducting their criminal activity and encourages those putative intermediaries to yield to the threats rather than resist.
[76] This is essentially a deterrence-based policy argument. Whatever the merits of the policy argument, it cannot withstand the head-on collision with the principle of moral involuntariness, now entrenched in s. 7 of the Charter. Even if punishing morally involuntary conduct would deter criminal organizations from coercing others into criminal acts and deter threatened persons from engaging in criminal acts, punishment of those whose conduct is morally involuntary will still infringe on the individual’s right to liberty in a manner that is contrary to the principles of fundamental justice. Unless that infringement can be justified under s. 1 of the Charter, criminal policy goals, no matter how legitimate, cannot be pursued at the expense of the constitutional protections afforded by s. 7 of the Charter.
[77] Furthermore, even if the principles of fundamental justice did not operate as a brake on an unfettered policy of deterrence, the validity of the deterrence argument is dubious. The argument assumes that persons subject to the sorts of threats that would engage the duress defence would take the time to consider their options in light of the manner in which they might be treated by the criminal law were they to be arrested and prosecuted at some point in the future. Surely, a person faced with a true “kill or be killed” option would have little regard for how the criminal law might react to the choice she makes. Any concern she might have about future punishment would be trumped by a more immediate concern about her own survival.
[78] We are also unable to see how excluding murder from the defence of duress would deter terrorists and other criminal organizations from using coercion to conscript others to assist in crime. The terrorist or organized criminal is guilty of the crime regardless of whether the coerced party has a defence. We cannot see how the criminal would have any interest in what the law might eventually do to the coerced party.
[79] There is nothing in the Canadian experience to support the contention that accepting duress as a defence to persons charged as parties to murder would encourage criminal organizations to employ duress or would encourage individuals to yield to threats. Duress has been an accepted defence to persons charged as a party to murder in several Canadian provinces, including Ontario, for many years. We are unaware of any data or commentary suggesting that the availability of this defence has created problems in the enforcement or administration of the criminal law. Nor do we know of any such data in various civil jurisdictions in which duress is an accepted defence to murder or in those common law jurisdictions which have expanded duress to murder by statute.[^8]
[80] The second reason advanced in the English cases and relied on by the trial judge for excluding murder from the duress defence rests on the assertion that the innocent victim’s right to life is inherently more valuable than any right of the accused. As the trial judge put it, at para. 99:
In considering whether an accused will be excused for murdering a victim to save his own life, we are dealing with precisely what LeBel J. was referring to: the conflicting interests of the accused and the victim. Both want to live. Both have a right to life. Their interests are equivalent and in direct conflict. In such a situation, the common law has established a hierarchy, and has held that the interests of the victim must prevail.
[81] That position reflects a conception of duress as a justification: see, e.g., R. v. Dudley and Stephens (1884), 14 Q.B.D. 273, at pp. 286-88; Howe, at pp. 430-33. On that view, duress is not a defence to murder because killing the innocent victim is never justified: it cannot be said that taking one life to save another is the lesser of two evils, or that the harm avoided is greater than the harm caused. Justifying the killing of the innocent victim would require prioritizing the right to life of one over the right to life of another. That kind of evaluation is both unworkable and inconsistent with Charter values.
[82] Canadian criminal law does not, however, regard duress as a justification. As outlined above, duress is an excuse. The person excused from criminal liability is not said to have accomplished a greater good, but is rather said to have had no realistic choice but to act as she did. The harm caused/harm avoided inquiry in an excuse-based defence, looks not to the promotion of the greater good (the harm avoided is greater than the harm caused), but rather for a level of comparable harm (the harm caused is not out of proportion to the harm avoided).
[83] An individual told to “kill or be killed” cannot make a decision that will fully vindicate the right to life, especially if the choice is between the lives of two equally innocent third parties. Whatever the threatened person decides, an innocent life may well be lost. A per se rule which excludes the defence of duress in all murder cases does not give the highest priority to the sanctity of life, but rather, arbitrarily, gives the highest priority to one of the lives placed in jeopardy.
[84] The availability of the defence of duress cannot be settled by giving automatic priority to the right to life of the victim over that of an accused. Instead, the right to life of the victim must be factored into the proportionality assessment as part of the broader moral involuntariness inquiry.
Conclusion
[85] The common law defence of duress has been available to persons charged as parties to murder in Ontario for at least 35 years. We would hold that the defence as strictly defined in the recent Supreme Court jurisprudence remains available to persons charged as parties to murder. In our view, that holding is consistent with the nature of the duress defence, basic criminal law policies, and the operative principles of fundamental justice.
[86] The constitutionality of the murder exception to the duress defence in s. 17 of the Criminal Code is not before the court. However, it follows from this analysis that, subject to any argument the Crown might advance justifying the exception as it applies to perpetrators under s. 1 of the Charter, the exception must be found unconstitutional.
DID THE RULING THAT DURESS WAS NOT AVAILABLE AS A DEFENCE TO MURDER PREJUDICE THE APPELLANTS?
(i) Gardiner
[87] Counsel for Gardiner does not submit that there was an air of reality to a duress defence on the evidence as it stood at trial. He is correct in not advancing that argument. There was no evidence that Gardiner acted under duress as defined in Ruzic and Ryan. His trial counsel did not even request that the defence be left with the jury on the included offences of manslaughter.
[88] Counsel for Gardiner submits that had the trial judge correctly held that duress is available for murder, Gardiner may have adduced evidence capable of giving an air of reality to the defence. Counsel suggests that Gardiner may have testified or that trial counsel may have introduced information through other witnesses that would have supported the defence of duress. Counsel attempts to overcome the obviously speculative nature of this submission by pointing to evidence that Gardiner was afraid during the carnage at Kellestine’s farm and that he took orders from the perpetrators of the murders during the confinement and murders. Counsel suggests that this evidence provides the evidentiary kernel from which the duress defence may well have been developed had the trial judge not erroneously ruled that it was not available on the murder charges.
[89] We cannot accept this submission for several reasons. First, the evidence said to suggest to the possibility of a duress defence for Gardiner does no such thing. A fearful subordination to the orders of others is miles from the kind of conduct required to bring the duress defence into play: see R. v. Mena (1987), 1987 CanLII 2868 (ON CA), 34 C.C.C. (3d) 304 (Ont. C.A.), at p. 320. Gardiner’s subservience is not a substitute for the kind of threat that would be required to lend an air of reality to a duress defence in these circumstances.
[90] Second, Gardiner had every opportunity to develop or at least lay out evidence in support of his duress defence at trial if such evidence existed. Gardiner could have done so by way of a proffer of evidence as part of the pre-trial motion on the availability of the defence. Gardiner could also have developed the evidence to advance a duress defence on the manslaughter charges. Gardiner’s decision not to adduce any evidence relevant to duress despite the availability of the defence on the manslaughter charges suggests to us that either there was no such evidence, or that counsel made a tactical decision that Gardiner was better off not attempting to advance a duress defence. As was stressed on several occasions at the trial, the duress defence could have opened the evidentiary door to Crown evidence detailing the criminal activities of the Bandidos and the chain of authority that prospective members agreed to when they became affiliated with the gang. As Gardiner’s counsel at trial acknowledged, such evidence could cause significant harm to the defence.
[91] It is always possible that things might have gone differently at trial had a trial judge made different pre-trial rulings. In our view, however, it puts too little weight on legitimate finality interests in the criminal law to suggest that the mere possibility that things might have gone differently should vitiate a verdict. With respect, counsel for Gardiner’s submissions go no further than to speculate about what might have occurred. That speculation runs firmly against the grain of what actually occurred. We see no realistic possibility that Gardiner could have or would have put the duress defence in play on the murder charges had the trial judge ruled that the defence was available: see R. v. Khawaja, 2012 SCC 69, [2012] 3 S.C.R. 555, at paras. 90-91.
(ii) Mather
[92] The appellant Mather, like Gardiner, does not challenge the trial judge’s ruling that there was no air of reality to a duress defence to any of the manslaughter charges. Also, like Gardiner, counsel for Mather argues that but for the trial judge’s erroneous pre-trial ruling, he may have testified and “the strategy pursued at trial may have been different”.
[93] In support of this submission, counsel speculates as to what Mather may have said had he chosen to testify. That speculation is not helpful. Furthermore, counsel offers no reason why, if Mather could have offered evidence to support a duress defence, he did not do so on the manslaughter charges. In our view, the effect of the trial judge’s erroneous holding that duress was not available as a defence to murder must be measured by reference to the trial record. Taking that measurement, we conclude that Mather suffered no prejudice.
(iii) Aravena
[94] Aravena testified that he was present when Kellestine shot and killed John Muscedere (the second murder). He further testified that immediately after killing Muscedere, Kellestine threatened to kill Aravena and his family if Aravena “talked”.
[95] Aravena’s trial counsel (not Mr. Doucette) argued that Kellestine’s threat after Muscedere’s murder provided a basis upon which the trial judge should leave the defence of duress on the included offence of manslaughter in counts three through eight. Counsel did not suggest that the duress defence was available for manslaughter on the first two murder counts.
[96] The trial judge correctly held that Kellestine’s express threat after Muscedere’s murder was not the kind of threat that could support a duress defence. The threat was not made to coerce Aravena to commit a crime, but was made to secure his silence about a crime Kellestine had already committed.
[97] The trial judge, however, considered Kellestine’s express threat in conjunction with the rest of the evidence and decided that, on the totality of the evidence, there was an air of reality to Aravena’s claim that he had acted under an implied threat of death if he did not do as he was told.
[98] The trial judge ultimately held, however, that the duress defence was not available to Aravena because, on the unchallenged evidence, Aravena had:
[V]oluntarily put himself into the clutches of an organization where he will be subject to compulsion, and specifically into the clutches of his tormentor, Wayne Kellestine, the National Sergeant at Arms of that organization, who was issuing the orders and who made the threat, on Aravena’s evidence.
That renders the defence of duress unavailable on Aravena’s own evidence, and on the authority of Li that I just quoted, it deprives the defence of an air of reality.
[99] On appeal, counsel for Aravena accepts that the trial judge properly ruled that the criminal association exception precluded Aravena’s reliance on duress as a defence to manslaughter. He submits, however, that it does not follow that the exception also foreclosed the defence on the murder charges. Counsel relies on the trial judge’s observation drawing that very distinction:
Although it is not necessary for the purposes of the decision I have arrived at, it is nevertheless useful to remember that we are not talking about Aravena being ordered or compelled to kill anybody. If we were, that would be murder, and the duress defence would be unavailable for that reason.
Instead we are talking about more benign orders, to keep watch over the Toronto Bandidos until they can safely be transported back to Toronto. That is certainly well within the realm of orders that, on the evidence, a person attempting to be a good Bandido prospect would be expected to follow, and on the evidence, did follow.
[100] Counsel submits that the criminal association exception to the duress defence operates only to exclude the type of criminal conduct an accused knew she could be compelled to perform when he voluntarily joined the criminal organization. Counsel concedes that there was no air of reality to a claim that Aravena did not know that he could be required to unlawfully confine others as a member of Kellestine’s group. That knowledge excluded the duress defence on the manslaughter charges because those charges were based on the unlawful confinement of the victims. Counsel goes on to argue, however, that it was very much an open question whether Aravena knew that as a consequence of submitting to Kellestine’s authority he could be required to assist Kellestine in a murder. Counsel submits that the question of what Aravena knew should have been left with the jury as part of a duress instruction.
[101] The Crown responds that even if the duress defence was available as a matter of law, the defence was not available to Aravena on the evidence and therefore this court should apply the curative proviso in s. 686(1)(b)(iii). Counsel submits that on the entirety of the evidence, there was no air of reality to Aravena’s assertion that he acted under the threat of death or bodily harm. Crown counsel also argues that the trial judge’s holding that the criminal association exception foreclosed the defence on the manslaughter charges is equally applicable to the murder charges.
[102] The trial judge’s finding that there was an air of reality to an implied threat by Kellestine against Aravena was based on his careful review of the entirety of the evidence, especially Aravena’s testimony. We would not interfere with that finding. Deference to the trial judge’s finding does not, however, end the matter. The trial judge found an implied threat based in large part on Aravena’s evidence of Kellestine’s direct threat made against him and his family after the second murder. Without that direct threat, the evidence could not, in our view, support the trial judge’s finding of an implied threat to kill or cause bodily harm to Aravena if he did not cooperate.
[103] In our view, the trial judge’s finding of an implied threat sufficient to support a duress defence must be taken in reference to the murders after Kellestine’s express threat (counts three through eight). There was no air of reality to a claim that Aravena acted under a threat of death when he aided or abetted in the second murder. A consideration of whether the failure to leave the defence of duress prejudiced Aravena must therefore accept the jury’s verdict on count two and proceed on the basis that Aravena was a full and willing participant in the second murder.
[104] Any claim by Aravena that he acted under duress after the second murder and that he did not know he could be compelled to assist in murders when he associated himself with Kellestine’s criminal activities, loses all credibility in the face of the jury’s finding that Aravena willingly assisted Kellestine in the second murder. Even on the appellant’s approach to the criminal association exception, the duress defence is excluded by that exception on counts three to eight.
[105] Apart entirely from whether the trial judge’s finding of an implied threat applies only to counts three through eight, we agree with the Crown’s submission that the criminal association exception rendered duress unavailable to Aravena.
[106] The association exception is part of both the common law and statutory defence of duress. Section 17 excludes from the defence persons who are a party “to a conspiracy or association whereby the person is subject to compulsion”. The common law exclusion is similarly focussed to exclude persons who “knew that their participation in a conspiracy or criminal association came with a risk of coercion and/or threats to compel them to commit an offence”: Ryan, at para. 75.
[107] Under the statutory and the common law defence, the criminal association exception denies the duress defence to those who willingly enter into relationships in which they know they may, by virtue of that relationship, be compelled by threats of bodily harm or death to commit crimes. A person who freely and knowingly subjects herself to the risk that she will be coerced into committing a crime voluntarily assumes the risk of acting under that coercion. Her eventual criminal conduct is, therefore, morally voluntary: Ryan, at para. 77. The requisite moral voluntariness is provided by that person’s decision to willingly place herself in a situation in which she knows she can be coerced into committing a crime. The Crown must still prove any mens rea required for the offence charged. It is at that stage that the accused’s foresight of the crimes she could be compelled to commit may have some relevance to the accused’s liability.
[108] The appellant’s submission that an accused must know the criminal organization might compel her to commit the crime charged finds no support in the Canadian authorities. The Canadian cases focus on the accused’s willing assumption of the risk she might be coerced into committing a crime, and not on the specific criminal activity either foreseen or eventually committed as a result of that coercion. This approach is consistent with the voluntary assumption of risk rationale for the exception: see Ryan, at paras. 75-77; Ruzic, at para. 70; R. v. Li (2002), 2002 CanLII 18077 (ON CA), 162 C.C.C. (3d) 360 (Ont. C.A.), at paras. 20-33.
[109] The English cases stand against the proposition that the criminal association exception excludes duress as a defence only to crimes which the accused anticipated she might be coerced into committing: see R. v. Sharp (1987), 85 Cr. App. R. 207 (Eng. C.A.), at p. 214; R. v. Shepherd (1987), 86 Cr. App. R. 47 (Eng. C.A.), at pp. 51-52; R. v. Ali, [2008] EWCA Crim. 716, at para. 12; Hasan, at paras. 29-39. For example, in Shepherd, the court approved of the following excerpt from a trial ruling:
It is not merely a matter of joining in a criminal enterprise; it is a matter of joining in a criminal enterprise of such a nature that the defendant appreciated the nature of the enterprise itself and the attitudes of those in charge of it, so that when he was in fact subjected to compulsion he could fairly be said by a jury to have voluntarily exposed himself and submitted himself to such compulsion.
[110] In Hasan, the House of Lords considered the ambit of the criminal association exception at some length. The court was asked to confirm the holding in R. v. Baker, [1999] 2 Cr. App. R. 335 (Eng. C.A.), at p. 344, that the exception foreclosed the defence only if the accused knew he might be coerced “into committing criminal offences of the type for which he is being tried”.
[111] The law lords unanimously rejected this submission and declared Baker wrongly decided. Lord Bingham began with a description of the policy underlying the duress defence. He referred to and relied on both Perka and Ruzic for the proposition that the defence of duress must be strictly controlled and limited: Hasan, at para. 22. That control became particularly important in the context of offences committed by or on behalf of terrorists and other kinds of criminal organizations.
[112] After a detailed review of the English cases, Lord Bingham held, at paras. 37-38:
The defendant is, ex hypothesi, a person who has voluntarily surrendered his will to the domination of another. Nothing should turn on foresight of the manner in which, in the event, the dominant party chooses to exploit the defendant’s subservience.
The policy of the law must be to discourage association with known criminals, and it should be slow to excuse the criminal conduct of those who do so. If a person voluntarily becomes or remains associated with others engaged in criminal activity in a situation where he knows or ought reasonably to know that he may be the subject of compulsion by them or their associates, he cannot rely on the defence of duress to excuse any act which he is thereafter compelled to do by them.[^9]
[113] The principle of moral involuntariness is given full effect by a criminal association exception that forecloses the duress defence if an accused voluntarily enters into a criminal organization knowing there is a risk he may be compelled by threat of bodily harm or death to commit a crime. The voluntary assumption of that risk provides the choice necessary to justify criminal conviction and punishment. The criminal association exception properly focuses the inquiry on whether an accused willingly chose to subject himself to the risk he would be coerced into committing a crime. Not only is that focus consistent with the principle of moral voluntariness, it also furthers legitimate policy concerns which demand strict limitations on the availability of the defence, especially in the context of crimes committed for or on behalf of criminal organizations.
[114] On any view of the evidence, Aravena willingly put himself under Kellestine’s authority knowing Kellestine could compel him to engage in criminal activity. Having made the choice to submit to Kellestine’s coercion, it does not lie in Aravena’s mouth to argue that when Kellestine in fact compelled him to engage in murder, he had no realistic choice. Aravena had made his choice.
CONCLUSION ON DURESS
[115] The trial judge erred in law by holding that the common law defence of duress was not available to persons charged as parties to a murder. That error caused no substantial wrong or miscarriage of justice to Gardiner, Mather or Aravena, as on the evidence there was no air of reality to that defence on any of the charges.
Issue #2: The Liability of Mather and Aravena under the Constructive First Degree Murder Provisions
[116] Aravena and Mather argue that standing guard as they did could not amount to “an essential substantial and integral” part of the murders. They argue that without that level of participation, the Crown could not establish the causation requirement under the constructive first degree murder provisions: see R. v. Harbottle, 1993 CanLII 71 (SCC), [1993] 3 S.C.R. 306.
[117] After the shots were fired, killing Raposo, Mather and Aravena rushed to the barn where the other Toronto Bandidos were being held. Mather was armed with a gun and Aravena had a baseball bat. The two men stood guard in the barn over the remaining Bandidos as each victim was led outside, one at a time, to be shot and killed. Aravena took possession of the gun for short periods of time. He also supported Kellestine’s false claim made to the captives that there were “back up” shooters on the roof. Aravena fetched drinks from the farmhouse, and assisted in the search of the cars for weapons, keys, cell phones and identification.
[118] Aravena and Mather, together with Kellestine, Sandham, Mushey and M.H. took control of the larger group of eight bikers. But for the subjugation of the victims, they could not have been led outside, one by one, to be executed. Aravena and Mather maintained control over the victims while Kellestine, Mushey and Sandham met outside to discuss next steps, and watched the remainder as two of the farm team accompanied each of the victims outside. The domination of the victims prevented them from escaping or fighting back. The forcible confinement and killings were all part of the same course of events. As observed in R. v. Ferrari, 2012 ONCA 399, 287 C.C.C. (3d) 503 (Ont. C.A.), at para. 87, by Rosenberg J.A., “I do not doubt that in some circumstances, participation in unlawful[ly] confining the victim can also be the foundation for active participation in the killing.”
[119] As the Crown aptly observed, this was an “execution assembly line.” The trial judge did not err in leaving first degree murder with the jury.
Issue #3: Gardiner’s Liability for Murder on Counts Three to Eight
[120] Gardiner did not participate in the unlawful confinement or the shooting. The jury must have convicted him on the basis that he was a party to a planned and deliberate murder. Once he realized that the first two victims had been killed, the jury clearly inferred that he also realized that the plan “if we kill one we kill them all” was to be carried out. He then returned to the farmhouse, and if he followed his orders, would have continued to monitor the scanner. He cleaned out the trunk of one of the vehicles before the last victim was shot and placed in a passenger seat of that car.
[121] Gardiner argues that there was no evidence that he did any act to aid or abet the last six murders and that mere presence was insufficient to make him a party.
[122] The trial judge instructed the jury as follows:
To be liable as an aider or abettor – and this covers both of those second and third categories, it is not enough that a person was simply there when a crime was committed by someone else. In other words, just being there does not make a person guilty as an aider or abettor of any or every crime somebody else commits in the person’s presence. Being a spectator while someone else commits a crime does not make the spectator liable. There is no legal duty to intervene and stop a murder or any other crime.
On the other hand, if a person knows that someone intends to commit an offence; and goes to or is present at a place when the offence is committed for the purpose of helping or encouraging the other person commit the offence, his presence can be evidence of aiding or abetting.
Just as mere presence at the commission of an offence does not make a person liable as an aider or abettor, the fact that a person is not present at the commission of an offence does not necessarily mean he is not liable as an aider of abettor. If a person knows that a crime is going to be committed by the principal, and does an act or acts that help or encourage the commission of that offence, with the intention of helping or encouraging the commission of that offence, he can be held liable as an aider or abettor, even if he is not present when the offence is actually committed.
[123] The above-quoted instruction is consistent with that approved by this court in R. v. McLeod, [1982] O.J. No. 59 (C.A.), at para. 2. An accused who is at or near the scene, with the intention of providing acts of assistance if necessary is more than merely present. That conduct can constitute aiding and abetting. The jury was properly instructed.
[124] Gardiner assisted in preparation of the firearms. His job was to monitor the scanner for police activity and warn the rest of the farm team, if necessary. The jury must have found that he knew of the plan to kill all, if one was killed, that he knew after the second killing that the plan would be implemented, and that he was there to follow orders and assist as required. This was a reasonable finding on the record. Gardiner was not there by chance. We do not accept that Gardiner’s conviction on six counts of first degree murder was unreasonable.
Issue #4: The Instruction on Aravena’s Post-Offence Conduct
[125] Aravena admitted in his evidence that he travelled from Winnipeg to Ontario in the hopes of advancing in the Bandidos hierarchy. After the murder, he was elated and excited to have earned a “prospect patch.”
[126] Here the trial judge instructed the jury in the following terms:
There is another potential use of the post-offence conduct that you heard. One of the elements you will be considering as you deliberate is whether the accused committed planned and deliberated murder. Sometimes, evidence as to what a person does after a crime has been committed can constitute circumstantial evidence of a pre-existing plan. To use a gruesome, but real, example, evidence that the body was cut up into pieces after the murder can be connected to the purchase by the accused, before the murder, of surgical instruments, to constitute circumstantial evidence of a pre-existing plan to commit murder and butcher the body. In our case, the evidence relied upon this regard is the evidence of Bandido-related activity by the accused Mushey, Aravena and Sandham in Winnipeg following the demise of the deceased, which it relies on as circumstantial evidence of a pre-existing plan to eliminate the Toronto chapter and elevate the status of the Winnipeg chapter and its members.
[127] Aravena submits that the trial judge ought to have instructed the jury that the evidence of his conduct after the killings was of no probative value as to the degree of culpability for the homicides, as it was equally consistent with guilt of manslaughter, second degree murder and first degree murder.
[128] In R. v. White, 2011 SCC 13, [2011] 1 S.C.R. 433, the court stated at para. 42:
Thus, Arcangioli and White (1998) should be understood as a restatement, tailored to specific circumstances, of the established rule that circumstantial evidence must be relevant to the fact in issue. In any given case, that determination remains a fact-driven exercise. Whether or not a given instance of post-offence conduct has probative value with respect to the accused’s level of culpability depends entirely on the specific nature of the conduct, its relationship to the record as a whole, and the issues raised at trial. There will undoubtedly be cases where, as a matter of logic and human experience, certain aspects of the accused’s post-offence conduct support an inference regarding his level of culpability.
[129] Here, Aravena’s happiness right after the killing suggests a number of logical possibilities:
a) He had obtained the goal to which he aspired;
b) The events at the farm were not shocking or unexpected, but had unfolded in accordance with a plan;
c) He knew about the plan; and/or
d) He aided the killers with knowledge of the plan.
[130] As a matter of logic and human experience, Aravena’s conduct after the killing was relevant to whether he was a party to a planned and deliberate murder.
Issue #5: The Ruling that M.H. was a Confidential Informant in Respect of Prior Dealings with the Winnipeg Police
[131] At trial, the Crown sought a ruling prohibiting the defence from cross-examining M.H. about his previous relationship with police. Counsel for Aravena, Mushey and Mather who responded to the Crown’s motion took the position that M.H.’s confidential informer privilege had been waived or lost by his subsequent conduct. On the return of the motion, counsel for Sandham submitted for the first time that he wanted to hear oral evidence on the motion from M.H. and his police officer handler and raised the question of whether M.H. was previously an agent instead of a confidential informant. Sandham had not filed any material on the motion, nor had he asked the Crown to have the two witnesses present. Neither witness was there, although counsel for M.H. had flown to London to assert his claim to informer privilege.
[132] The Crown was prepared to argue the motion on the basis of the material filed, the relevant portions of the preliminary inquiry transcript. The trial judge ordered that the motion would proceed on the basis of the material filed by the Crown and defence.
[133] He did not err in exercising his discretion in this manner. A trial of this duration, complexity, and number of parties and witnesses requires careful management. Trials which spin out of control and end up lasting far longer than necessary deprive other litigants who are ready to proceed of their day in court.
[134] M.H. asserted a claim to privilege in relation to his contacts with Winnipeg police before the shooting of the eight Toronto Bandidos. He expressly waived any privilege related to his contacts with police after the shootings.
[135] Counsel for Mushey indicated in submissions to the trial judge, “I have no submissions to make, Your Honour, as to whether or not he was an informer or agent. I’m standing here before you intending to make my submissions on the assumption that he is a confidential informant.”
[136] Counsel went on to argue that since it had been revealed at the preliminary inquiry that M.H. was an informant with a long standing relationship with a particular officer, the informer privilege no longer applied. He argued that there was no need for the protection offered by the confidential informer privilege, as M.H. was in a witness protection program. He wanted to question M.H. about his prior contacts with police as a confidential informant with a view to undermining M.H.’s credibility.
[137] Counsel for Aravena submitted that since M.H. had told a police officer other than his Winnipeg handler that he had acted as a confidential informant in the past, he must be taken to have given up all claim to the privilege. Counsel for Mather agreed with the submissions by Mushey. Kellestine did not wish to pursue this line of questioning. The other counsel had no submissions to make on this issue.
[138] The issue as to whether an individual is a state agent, as opposed to a confidential informant is a question of mixed fact and law, and is owed some deference: R. v. N.Y., 2012 ONCA 745, 294 C.C.C. (3d) 313, at para. 114.
[139] As observed in R. v. G.B., 2000 CanLII 16820 (ON CA), 146 C.C.C. (3d) 465, [2000] O.J. No. 2963 (C.A.), at para. 10, “[i]n general terms, the distinction between an informer and an agent is that an informer merely furnishes information to the police and an agent acts on the direction of the police and goes ‘into the field’ to participate in the illegal transaction in some way.”
[140] Here the trial judge expressly concluded that M.H. was a confidential informant and not an agent in his dealings with the Winnipeg police. He indicated in his ruling on the confidential informant motion, R. v. Sandham, [2008] O.J. No. 5804 (S.C.), at para. 13, that, “[t]he evidence is clear and uncontradicted that M.H. has indeed been a confidential informant since approximately 1997” and, at para. 19, that “the evidence is that M.H. was only recruited as an agent for purposes of the investigation of these matters before the court.” These findings were reasonably open to the trial judge on the evidence before him and the arguments by counsel. There was no evidence that M.H. participated in the illegal activities which were the subject of his reports to police.
[141] The evidence which Mushey points to for the first time, on appeal, that on one occasion the handler asked M.H. for a copy of floor plans of a clubhouse, or on occasion asked for specific information, does not amount to sufficient evidence of direction as to undermine the trial judge’s conclusions.
[142] The appellant also argues that the handler coerced M.H. into acting as an informant. The evidence established that M.H. provided information to police in the expectation that his participation would remain confidential. There can be no doubt that he was risking his life in providing the information. Even if there were misconduct on the part of the handler, which we do not necessarily accept, the need for protection of the confidential informant is not diminished.
[143] The appellants do not pursue the argument here that the confidential informant privilege was waived or lost.
[144] The trial judge did not err in concluding that M.H. could not be questioned about his prior contacts with Winnipeg police.
Conclusion
[145] This was a difficult and lengthy trial. This court has had the benefit of careful and thorough arguments presented on behalf of the appellants and the Crown. We are satisfied that the appeal must be dismissed. The trial judge fairly and carefully instructed the jury. His charge was well structured, and equipped the jury to grapple with the evidentiary and legal issues presented. His failure to hold that the common law defence of duress was available to persons charged as parties to murder caused no prejudice to any of the appellants.
[146] The appeals are dismissed.
Released: “DD” “APR 16 2015”
“Doherty J.A.”
“G. Pardu J.A.”
“I agree David Watt J.A.”
[^1]: M.H., a Bandidos accomplice in the murders, became a Crown witness in exchange for immunity.
[^2]: In distinguishing between perpetrators and parties in the defence of duress, Paquette overruled R. v. Carker, 1966 CanLII 39 (SCC), [1967] S.C.R. 114. Carker had described the defence of duress in s. 17 as an exhaustive codification of the duress defence. New Zealand appellate courts in interpreting a provision that is very similar to s. 17 have also held that the statutory provision is exhaustive and applicable to perpetrators and parties alike: see R. v. Akulue, [2013] NZSC 88, at para. 22, fn. 45. It can be fairly said that the justification either in policy or in the language of s. 17 for the distinction drawn in Paquette is far from compelling.
[^3]: Murder has always been excluded from the statutory defence in Canada: see The Criminal Code, 1892, S.C. 1892, c. 29, s. 12; see also Criminal Code Bill Commission, Report of the Royal Commission Appointed to Consider the Law Relating to Indictable Offences (London: Eyre and Spottiswoode, 1879), at p. 68.
[^4]: In R v. Sheridan, [2010] O.J. No. 4884 (S.C.), the trial judge rewrote section 17 to permit a duress defence to a person charged as a perpetrator in a murder.
[^5]: Contrary to this line of authority, Fish J.A., writing for the Quebec Court of Appeal, in R. v. Langlois, 1993 CanLII 3594 (QC CA), [1993] R.J.Q. 675, 80 C.C.C. (3d) 28 (C.A.), at pp. 31-32, noted in passing that the common law defence of duress had not been available for murder since the pronouncement of the House of Lords in R. v. Howe, [1987] A.C. 417 (H.L. Eng.).
[^6]: S. Yeo, “Revisiting Necessity” (2010) 56 C.L.Q. 13, at p. 20. Some commentators prefer the phrase “normative involuntariness”: see D. Paciocco, “No-one Wants to be Eaten: The Logic and Experience of the Law of Necessity and Duress” (2010) 56 C.L.Q. 240, at p. 250. See also, Stuart, at p. 489.
[^7]: In Latimer, at para. 40, the court left open the question of whether necessity could provide a defence in a homicide situation. The court opined that “[i]t is difficult … to imagine a circumstance in which the proportionality requirement could be met for a homicide. We leave open, if and until it arises, the question of whether the proportionality requirement could be met in a homicide situation.” The proportionality requirement in necessity, unlike duress, is a purely objective one.
[^8]: See Payam Akhavan, “Should Duress Apply to All Crimes? A Comparative Appraisal of Moral Involuntariness and the Twenty Crimes Exception Under Section 17 of the Criminal Code” (2009) 13 Can. Crim. L. Rev. 271, at pp. 277-78, 282-84 (pointing out that the penal codes of France and Germany do not exclude duress as a defence to murder yet no evidence suggests that the availability of duress in those jurisdictions facilitates organized killing); W.R. LaFave, Substantive Criminal Law, 2d ed., vol. 2 (St. Paul: Thomson West, 2003), at p. 81 (noting that Alaska, Arkansas, Connecticut, Delaware, Hawaii, New York, North Dakota, Pennsylvania, South Dakota, Texas and Utah allow duress as a defence to murder by statute).
[^9]: English law, unlike the Canadian law, applies an objective test in considering the accused’s knowledge that the criminal organization might compel him to commit a crime: see Ryan, at para. 80.