A miscarriage of justice is the conviction and punishment of a person for a crime they did not commit and is synonymous with wrongful conviction. Wrongful convictions come from almost every jurisdiction in Canada. Donald Marshall in Nova Scotia, Rejean Hinse in Quebec, Steven Truscott and Guy Paul Morin in Ontario, Thomas Sophonow in Manitoba, David Milgaard in Saskatchewan, Wilson Nepoose in Alberta and Ivan Henry in British Columbia. As notorious as their cases are and as palpable as wrongful convictions should be – the test for miscarriage of justice can be elusive.

The term wrongful conviction appears nowhere in the Criminal Code. But the phrase miscarriage of justice does. It is the principle that vests appellate courts in Canada with the power to overturn a wrongful conviction and either substitute an acquittal or order a new trial. It recently governed the opinions of a three member panel of the Alberta Court of Appeal in its split 2-1 ruling on the trial of Connie Jean Oakes for the murder of Casey Armstrong in Medicine Hat, Alberta.

But, first, what is a miscarriage of justice?

The phrase is not defined in Code s. 686(1)(a)(iii) which simply states that a court of appeal may allow an appeal against conviction, “…on any ground where there was a miscarriage of justice”. To fill out the parameters of the test we have to turn to judge made law. The test has historically been given a broad and liberal construction by the Alberta Court of Appeal.

In R v Duke (1985), 22 CCC (2d) 217 (Alta CA) Justice John W. McClung, writing for a unanimous court at para. 21, held that, “…the determination of whether ‘a miscarriage of justice’ has occurred rests on broader considerations than those attaching to the demonstration of a ‘substantial wrong.’ Proof of actual prejudice resulting from an error of law is not a requisite to a finding that ‘a miscarriage of justice’ has occurred. It may be enough that an appearance of unfairness exists. R. v. Hertrich et al., 137 D.L.R. (3d) 400 (Ont. C.A.), leave to S.C.C. refused 45 N.R. 627.”

And in R v Nepoose (1992), 71 CCC (3d) 419 (Alta CA) Justice Rene P. Foisy stated, at para. 9, that a new trial would be ordered if, “…there was a miscarriage of justice or at least that there was a real possibility that a miscarriage of justice occurred during the trial.”  Most recently, in R v Oakes, 2016 ABCA 90, the case that is my topic here, the majority ruling of Justices Myra Bielby and Frederica Schutz, at para. 11, adopted the opinion in R v Truscott (2007), 225 CCC (3d) 321 (Ont CA) where a unanimous five member panel of the Ontario Court of Appeal stated, at para. 110, that the power to overturn a conviction founded in a miscarriage of justice, “…can reach virtually any kind of error that renders the trial unfair in a procedural or substantive way.”

So, let’s get to what happened here.

Wendy Scott was the sole witness who fingered Oakes for the stabbing death of Armstrong in his trailer. She had cognitive limitations and a learning disability. The jury was told that she had pleaded guilty to the second degree murder of Armstrong and was serving a life sentence when she testified against Oakes. The prosecutor said in her opening address to the jury that Scott’s testimony would be credible because she already admitted to committing the crime. There was no forensic or other evidence tying Oakes to the killing.

The Alberta Court of Appeal subsequently overturned Scott’s conviction because the facts presented to the sentencing judge when she pleaded guilty did not support a conviction for second degree murder. A new trial was ordered – but all that happened after Scott testified and the jury convicted Oakes. The majority put the miscarriage of justice issue this way, at para. 13:

“An assessment of this issue is not limited to the evidence heard by the jury but includes the fresh evidence that Ms. Scott, then believed to be equally guilty of the crime as Ms. Oakes, may not in fact be guilty. This evidence raises concerns that the jury overlooked legitimate credibility issues about Ms. Scott’s testimony, because it was bolstered by her apparent admission against interest: her own guilty plea to the same crime.”

Scott admitted during cross-examination that she previously identified three other people as the murderer. And she lied to the police and at the preliminary inquiry in an attempt to create a false alibi. Her testimony thus required a strong warning from the trial judge that it would be unsafe for the jury to convict in the absence of independent evidence confirming what she said. (See: R v Vetrovec, [1982] 1 SCR 811)

But it gets worse. Scott only identified Oakes after the police suggested to her that (a) her own DNA was found in the deceased’s trailer and (b) Oakes had fingered her as the killer. The police then took Scott to the trailer where she purported to re-enact how Oakes stabbed Armstrong. The re-enactment got off to a good start when Scott took the police to the wrong trailer. In affidavits admitted as fresh evidence on the appeal, the police conceded that both of the statements they made to Scott were lies.

The majority came to these conclusions in the application of the test, and the intertwined doctrine of deference to the jury’s verdict, at paras. 38-9:

“…[W]hat we are reviewing here is not simply the evidence led before the jury but also the ‘fresh’ evidence of the setting-aside of Ms. Scott’s own conviction with the resulting impact that may have had on her credibility assessment by the jury had they known this event would occur in the future. This is a key difference in this case, from one where an appellate court is asked to set aside a verdict based on a review of only the evidence led at trial.

…[W]e must consider both the evidence the jury did hear, and the fresh evidence that shows Ms. Scott should not have been treated as a convicted accomplice in the same murder when she testified against Ms. Oakes. If that fresh evidence had existed at the time of Ms. Oakes’ trial, and was presented to the jurors, we conclude that there is a reasonable possibility that the jury would have found that the admitted inconsistencies, lies and self-interest in Ms. Scott’s testimony left them a reasonable doubt that her identification of Ms. Oakes as the murderer was reliable…[D]eference to the decision of the trier of fact cannot play a significant role where that trier of fact was not in possession of critical information when it decided to convict.”

Justices Bielby and Schutz therefore concluded, at para. 59, that this was not a case where an ongoing miscarriage of justice could only be prevented by a directed verdict and a new trial was ordered.

Yet, that is not the end of the story. Justice J.D. Bruce McDonald dissented. It is difficult to discern the crux of the dissent. Certainly, there was no disagreement about the law. It appears that he diverged from the majority on the application of the test for the reception of fresh evidence on appeal. He stated, at para. 66, that the setting aside of Scott’s conviction, “…could not be expected to have affected the result of the trial because it does not decisively bear upon Ms. Scott’s credibility.” He then went on to this conclusion, at para. 70: “The fact that the conviction no longer stands also cuts both ways. We cannot know how it would affect Ms. Scott’s credibility assessment. In other words, we cannot expect it would affect the assessment of Ms. Scott’s credibility either positively or negatively, and therefore, we cannot expect it would have affected the result of the trial.”

There are two mistakes here. First, Justice McDonald is discussing the fourth criteria of the Palmer test. That is, whether the fresh evidence could reasonably be expected to have affected the result. This test is a low threshold but Justice McDonald would impose a higher one of decisiveness in order to reach it. That is not the law. The second criteria of the Palmer test only requires that the fresh evidence bear upon a decisive issue in the trial. It does not require that it bear decisively on that issue. Second, Justice McDonald would impose another high standard requiring appellate judges to know how the setting-aside of Scott’s conviction would affect the credibility assessment of her testimony by a jury. That is not the law either. The test is one of reasonableness not decisiveness; could and not would. In the Supreme Court of Canada’s landmark ruling in R v Palmer, [1980] 1 SCR 759 Justice William McIntryre succinctly put the fourth criteria of the test this way, at p. 11 (QL): The fresh evidence, “…must be such that if believed it could reasonably, when taken with the other evidence adduced at trial, be expected to have affected the result.”

Justice McDonald then veered into a third error where he emphasized, at para. 71, that the facts admitted at Scott’s sentence hearing, “…were essentially the same narrative that she testified to at the appellant’s trial a year later.” Assuming that the agreed facts from Scott’s case were properly admitted at Oakes’ trial (attached to Justice McDonald’s opinion as Appendix A) they were not admissible as a form of prior consistent statement which is the very use to which he put them. The rule prohibiting the use of a prior consistent statement is grounded in concern that evidence may be manufactured by a witness through mere repetition of a consistent lie. (See: R v Keeler (1977), 36 CCC (2d) 8 (Alta CA) per McDermid J.A., at para. 4; and, Watt’s Manual of Criminal Evidence (2012), at p. 219ff)

The case was sent back to the Court of Queen’s Bench of Alberta, in Medicine Hat, where Connie Jean Oakes will try to get a fair trial the second time around.