At Youvarajah’s trial DS was called by the Crown as a witness. DS was asked to adopt the agreed statement of facts. Shockingly, he testified that he did not remember signing the document (although he accepted it bore his signature). Further, he testified that the statement was not true. He indicated, instead, that the gun was his own and that he threw the gun into the river after the shooting. Regarding his plea, he indicated that he “did not understand the words ‘acknowledged’ or ‘accurate’ as used by his counsel” and that one of the reasons he pleaded guilty was that he was told he would not have to give any further statement [para 9].
The Crown sought to tender the statement under the principles of KGB. The trial judge denied the Crown’s application on the basis it did not meet threshold reliability. Karakatsanis J outlined the trial judge’s reasons as follows:
None of the safeguards identified in this Court’s decision in R. v. B. (K.G.),  1 S.C.R. 740 (K.G.B.), for reducing the dangers associated with hearsay had been undertaken. The statement was not videotaped; there had been no recital of an oath or affirmation; and the transcript of the guilty plea proceedings was not a suitable substitute to assess D.S.’s demeanour and credibility at the time the statement was made. Furthermore, the statement had been drafted by counsel; it was not offered spontaneously and was not in D.S.’s own words [para 12].
Further barring admission was the fact that cross-examination was largely illusory as DS claimed to have significant memory losses and refused to waive solicitor-client privilege [para 13].
DS’ signed statement of facts was thus not admitted. Youvarajah was acquitted. The Crown appealed. The Court of Appeal allowed the appeal. Youvarajah appealed to the Supreme Court. Karakatsanis J, writing for the majority, allowed the appeal and restored the acquittal.
In considering the appeal Karakatsanis J began with a review of the law of hearsay and admissibility. Inter alia, Karakatsanis J held that prior inconsistent statements may be admitted where the KGB criteria are in place; those include (i) a statement made under oath or affirmation after a warning about possible sanctions; (ii) a video recording of the statement; and (iii) full opportunity to cross examine [para 29].
Karakatsanis J noted, however, that this approach is not the “only means of establishing threshold reliability” and offered the following alternate approaches (not mutually exclusive): (i) the presence of adequate substitutes for testing truth and accuracy (procedural reliability); and (ii) sufficient circumstantial guarantees of reliability or an inherent trustworthiness (substantive reliability) [para 30].
With that review in hand, Karakatsanis J turned to consider the reasons of the trial judge for rejecting the statement. They included the following:
First, DS had a “strong incentive to minimize his role…and shift responsibility” [para 33].
Second, there was no opportunity to observe the demeanour of DS.
Third, the statement was not his words.
Fourth, cross-examination was curtailed.
These factors, Karakatsanis J held, were properly considered by the trial judge and the decision (concluding that threshold reliability was not met) was subject to deference. This conclusion was reached notwithstanding, what the Crown said was sufficient circumstantial guarantees in the plea. On that point, Karakatsanis J offered the following:
In my view, the circumstantial guarantees of trustworthiness asserted by the respondent — the thorough process in creating the ASF, the involvement of counsel, and the solemnity of the guilty plea proceeding — do not establish threshold reliability for the statements from which D.S. recanted, which served to minimize his involvement in the murder and shift responsibility to the appellant. In the circumstances of this case, the formality of the process and the involvement of counsel only provide comfort in respect of D.S.’s statements admitting his own culpability for the murder [emphasis added]; [para 57].
There are a couple of aspects of this decision that, with respect, are peculiar.
First, the majority’s review of the law on hearsay seems unnecessary and unduly restrictive. In Khelawon the Court endorsed – after much academic and judicial commending – a principled approach. Here Karakatsanis J seems to revert somewhat by beginning her consideration of the law by setting out the principles of KGB. In fairness, Karakatsanis J clearly set out that the KGB criteria are not the only means of establishing threshold reliability. Nonetheless, the structured and historical review of the law of hearsay, coupled with the outcome of the case – only a few years after a clear and well-reasoned decision in Khelawon – is peculiar and perhaps confusing.
Second, the majority’s consideration of the solemnity of a plea and involvement of counsel is curious. Ironically, Karakatsanis J outlined admissions against interest as one of the traditional hearsay exceptions [para 20] – exceptions which the Starr Court recognized as presumptively admissible. One might argue that a guilty plea is an admission against interest thus presumptively invoking threshold reliability. This would seem more so where the admission is made with the benefit and presumptively competent advice of counsel.
Karakatsanis J offered this in relation to this point:
To the extent that the ASF incriminated D.S., was against his interests, and admitted his own culpability in court, these circumstances provide a compelling inference that those statements were in fact reliable in establishing D.S.’s criminal conduct. However, the underlying rationale for the admissibility of admissions as against the party making them falls away when they are sought to be used against a third party [para 59].
As logical as that appears, the problem with this approach is twofold. First, it presumes that, notwithstanding a full version of events was read out and accepted in court as true and accurate, an accused is not sincere in his acceptance except to the extent it implicates him. Second, it presumes that counsel has given improper and arguably incompetent advice – to the extent he has advised his client to accept facts which may not be true or are “irrelevant” to his own culpability. Third, it ignores that this fact – the possibility of a false allegation in order to obtain a better sentence – is the very area of examination and consideration that will allow a trier of fact the ability to assess ultimate reliability. Wagner J, in dissent, offered the following apt comments in this regard:
The hearsay statement at issue in this case is an ASF which was drafted with the contribution and approval of Crown counsel, defence counsel and the accused. More importantly, it was accepted by a court in the sentencing of a youth for the very serious crime of second degree murder.
An ASF presented in court carries an inherent level of reliability that the judicial system accepts and, in fact, implicitly relies upon in the workings of the administration of justice. It follows that not allowing an ASF to be admitted for its truth in a subsequent hearing, to be weighed and evaluated by the trier of fact, is problematic when viewed through the wider lens of the repute of the system of administration of justice. What does it say about the administration of justice if the courts accept the reliability of a statement to convict an individual and to vary the length of his sentence for a crime as serious as murder, but are unwilling to place that same statement before the trier of fact in another proceeding to be weighed and considered against a recantation of that statement? [Emphasis added]; [paras 106 and 108].
Putting aside these peculiarities, the lessons to be learned from the decision may include the following.
First, pleas in these circumstances should be based upon a “statement” by the accused as opposed to counsel drafting the facts or relying on a police summary.
Second, part of the plea should include a signed waiver of privilege regarding the advice provided in support of the plea, or an examination of the accused on the stand about the fact he received advice and understood its scope and ramifications. This is particularly so given the following comments of Karakatsanis J:
Furthermore, the involvement of defence counsel provides no meaningful check on the danger of an accused acknowledging false allegations against a third party in order to obtain a favourable plea bargain. Counsel have an ethical duty to not knowingly mislead the court. However, it does not require them to verify or investigate the truth of information they present; and the duty is triggered only where counsel has information leading to the “irresistible conclusion” that something is false. See M. Proulx and D. Layton,Ethics and Canadian Criminal Law (2001), at pp. 40-47 and 460 [para 61].
Third, the plea should include the accused either being videotaped – presumably at the police station with an oath – or taking the stand on the plea to accept the facts and acknowledge them as his own accurate recounting of events.
The facts in Youvarajah may be somewhat out of the ordinary; nonetheless the majority’s ruling causes one to pause. It could appear to invite those who are so motivated to take advantage of the criminal justice system through perjury and mischievous pleas of guilt with “substantial” acceptance of facts. It appears to undermine the solemnity of a plea and the scope and ramifications of advice by a lawyer prior to plea. These are some of the concerns likely held by Wagner J, which prompted the comments noted above (paras 106 and 108) and led to the following conclusion:
Regardless of the circumstances of a particular case and the indicia that may be available for the reliability analysis, it is crucial not to lose sight of the truth-seeking function of an admissibility inquiry under the principled approach. In the instance of a prior inconsistent statement, the purpose of the admissibility inquiry is not to make an absolute determination of the reliability of the hearsay statement as compared with the subsequent testimony. Rather, the inquiry is intended to ensure that the statement is sufficiently reliable in order for the trier of fact to assign the appropriate weight to the evidence. As Charron J. states in Khelawon, the “general exclusionary rule is a recognition of the difficulty for a trier of fact to assess what weight, if any, is to be given to a statement made by a person who has not been seen or heard, and who has not been subject to the test of cross-examination” (para. 35). [Emphasis added]; [para 99].