By: Amy Matychuk
PDF Version: When the Burden of Proving Institutional Bias Rests on a Prisoner
Case Commented On: Canada v Ewert, 2016 FCA 203 (CanLII)
Prisons use psychological tests to determine if inmates are likely to reoffend, but are the tests accurate for Aboriginal inmates? In a recent Federal Court of Appeal case, the court found that there was not enough evidence to prove the tests are biased. However, the analysis overlooked a few important factors.
In Canada v. Ewert, 2016 FCA 203 (CanLII) (Ewert FCA), Justice Dawson overruled a Federal Court decision that Correctional Service Canada’s (CSC’s) tests are unreliable when used to assess Aboriginal inmates. She held that Mr. Ewert, a 53-year-old Métis offender serving two life sentences in federal prison, did not provide enough evidence that the tests generate “false results and conclusions” due to cultural bias against Aboriginal people (at para 34). Mr. Ewert argued that these psychological tests do not take Aboriginal cultural differences into account. He alleged that his test scores affected “[his] eligibility for parole, his security classification and his ability to be granted escorted temporary absences” (at para 7). Because the tests generate inaccurate results for Aboriginal inmates, he said, relying on his scores to restrict his freedom was a violation of his rights. Justice Phelan of the Federal Court agreed, finding a section 7 Charter breach and a breach of the Corrections and Conditional Release Act, SC 1992, c 20 (see 2015 FC 1093 (CanLII) (Ewert FC)). However, the Federal Court of Appeal overturned that decision, and ruled that Mr. Ewert had not established on a balance of probabilities that the tests were unreliable.
Justice Dawson’s main reason for overturning the Federal Court ruling was that Justice Phelan failed to require that Mr. Ewert meet the necessary burden of proof (at para 15), which was to establish his claims on a balance of probabilities (at para 19). Justice Phelan ruled that after Mr. Ewert raised a reasonable challenge, the burden shifted to Correctional Services to show they had taken reasonable steps to fulfill their statutory duty to ensure the accuracy of the test. (Ewert FC at para 82). Justice Phelan did not address whether Mr. Ewert had proven his claim on a balance of probabilities.
Mr. Ewert provided expert evidence which suggested that the tests were likely (though not certainly) biased, and Justice Phelan found that this evidence was enough to present a reasonable challenge to the tests’ validity. He held that CSC had breached sections 4(g) and 24(1) of the Corrections and Conditional Release Act by, respectively, failing to make correctional policies responsive to the special needs of Aboriginal persons and failing to ensure that information about an offender is as accurate, up to date, and complete as possible. He also held, based on the same evidence, that the CSC breached Mr. Ewert’s section 7 Charter rights (Ewert FC at paras 75, 113).
Justice Phelan ordered that CSC respond to Mr. Ewert’s challenge of the tests’ accuracy by assessing the tests to ensure their fairness; he also ordered that until CSC assessed the tests’ fairness, CSC could not use the tests on Aboriginal inmates (paras 114-116). In contrast, Justice Dawson held that Mr. Ewert had not established on a balance of probabilities that the tests suffered from cultural bias, and therefore a ruling requiring CSC to ensure the tests’ fairness was not required.
As Justice Phelan had recognized, the available evidence was “a thin record” on which to decide this case (Ewert FC at para 3). Out of the three expert witnesses, the Federal Court largely dismissed two of them for being too narrow and lacking objectivity. Justice Phelan relied on one psychologist, Dr. Hart, for all of the expert evidence in his judgment. Justice Phelan called Dr. Hart’s evidence “balanced, objective, and despite the Defendant’s criticism that he cited no studies, credible” (at para 26). Dr. Hart gave evidence about the types of cross-cultural bias likely to affect CSC’s tests, testifying that they are “more likely than not to be ‘cross culturally variant’” (at para 28). He referred to the “pronounced differences between Aboriginal and non-Aboriginal groups” and said that personally, he would not trust the tests to accurately assess Aboriginal inmates (at para 31.) Dr. Hart concluded that the tests “are not sufficiently predictably reliable for Aboriginals because of the cultural variance or bias of the tests.” (at para 41)
Justice Dawson quoted excerpts from Dr. Hart’s direct examination before the Federal Court, in which he said, “[the bias] may be relatively small and it may be tolerable. But it could actually be large and it may be intolerable. . . . my own professional opinion would be, it would be more likely than not that there is some kind of bias.” (Ewert FCA at para 25) Justice Dawson was not persuaded by this evidence because of the possibility that the tests’ bias against Aboriginal people might be small and inconsequential. Evidence that adequately supported Mr. Ewert’s claim, she held, would have been evidence showing that “cultural bias affected or is more likely than not to affect test usage or the reliability and validity of the resulting test scores in a material way” (at para 27). Dr. Hart’s evidence that bias likely affected the tests in some way was not enough; Justice Dawson wanted proof not only that bias existed, but that the bias was substantial and affected Aboriginal inmates in a negative way.
From a purely technical standpoint, Justice Dawson’s ruling that Mr. Ewert had not established his claim on a balance of probabilities was fair, given the single reliable expert witness and the lack of other evidence. However, before ruling against Mr. Ewert, she ought to have considered the lack of additional evidence available to him. In order to present more evidence to support his claim, he would need an assessment of the psychological tests at issue. As Dr. Hart testified, no research exists that has assessed these tests for cross-cultural bias (Ewert FCA at para 23). Dr. Hart discussed the three ways of assessing these tests, but added, “[i]t is hardly practical for an individual litigant to engage in this type of analysis. Given the CSC’s legislated mandate . . . it is an activity more appropriately commissioned by CSC.” (Ewert FC at paras 34 and 35).
The only way Mr. Ewert could obtain evidence to better support his claim was to wait for CSC, his adversary, to assess the psychological tests for cross-cultural bias and to provide him with the results. To date, CSC has not performed these assessments, which leaves Mr. Ewert in a difficult position: Justice Dawson ruled that he needed more evidence, and the only way he can obtain more evidence is to wait for CSC to perform assessments on its own psychological tests. CSC has very little incentive to perform these assessments because to do so would be to risk creating evidence to support Mr. Ewert’s case.
Justice Dawson should have considered that it would be near impossible for Mr. Ewert to obtain better evidence against CSC than he already had. Justice Phelan recognized this evidentiary difficulty in his Federal Court ruling; he ordered that CSC perform the test assessments before using the tests on Aboriginal inmates. He did not declare that the tests are cross-culturally biased; he only required that CSC produce evidence that would establish whether or not Mr. Ewert’s claim was legitimate. Justice Dawson, conversely, did not require CSC to do anything, leaving Mr. Ewert and other Aboriginal inmates no better off with respect to potentially culturally biased psychological testing. As discussed above, CSC is unlikely to feel motivated to perform these test assessments without a court order: if the assessments show that the tests are biased, CSC will have invited a suit against itself.
However, in an interesting turn of events, CSC is supposedly already assessing the tests for bias. Justice Phelan referred to two previous challenges Mr. Ewert made to the tests’ validity, Ewert v Canada (Attorney General), 2007 FC 13 (CanLII), and Ewert v Canada (Attorney General), 2008 FCA 285 (CanLII). In the 2007 ruling, Justice Beaudry noted that CSC told Mr. Ewert “the study undertaken by its Research Branch in 2003 regarding applicability of the actuarial assessment tools to its Aboriginal inmate population” was still incomplete, and they could therefore not respond to his complaint (at paras 62 and 66). Justice Beaudry urged CSC to, when the research was complete, “explain to the Applicant the initiative undertaken by the Research Branch and the results obtained, if any” (para 67). Neither Justice Phelan nor Justice Dawson’s ruling mentions this study. Based on these previous cases, when Justice Phelan ordered that “the Defendant [conduct] a study that confirms the reliability of [the tests] in respect to adult Aboriginal offenders,” he was ruling that CSC perform research it had already alleged to be ongoing (para 114).
In normal circumstances, a case with so little evidence and so few witnesses to support the plaintiff’s claim should not succeed, and Justice Dawson’s ruling would have been both reasonable and justified. However, this is a case involving a vulnerable prison inmate attempting to enforce his rights against a powerful government actor. In order for Mr. Ewert to successfully challenge the tests’ validity, he needs evidence about cross-cultural bias that requires assessments that only CSC is capable of generating or providing. Justice Phelan’s ruling would have put pressure on CSC to complete assessments of tests that it alleged were ongoing as long ago as 2003 (2007 FC 13 at para 62). If these assessments established that the tests are fair, as CSC argues, Justice Phelan’s judgment would not have prevented CSC from continuing to use the tests on Aboriginal inmates. Justice Dawson’s judgment, however, removes this judicial pressure and allows CSC to continue using potentially biased psychological testing on Aboriginal inmates rather than holding CSC accountable to ensure Aboriginal inmates are treated fairly. Though Justice Dawson’s judgment may be technically correct, it is unfortunate that Mr. Ewert’s claim failed based on a lack of evidence that only his adversary, CSC, was capable of remedying.
To subscribe to ABlawg by email or RSS feed, please go to https://ablawg.ca
Follow us on Twitter @ABlawg
