O’Malley: The Prima Facie Test
The law of human rights in Canada is really based on a single case: Ont. Human Rights Comm. v Simpsons-Sears,  2 SCR 536 [O’Malley], often cited as O’Malley v Simpsons-Sears, at para 28, where Justice McIntyre fatefully introduced the prima facie test. His formulation of a prima facie case follows the precedents, which essentially say that there must be some evidence on which a verdict could be entered in favour of the prosecuting party. The idea of “some evidence” can be very slippery.
The more fundamental problem is that the Supreme Court of Canada seems to have overlooked the fact that it is a legal error for a court or tribunal to weigh the evidence in applying the prima facie test. This is trite law, which will be familiar to anyone who has applied for a non-suit. The reasons for the historical position are buried in the law, but the basic principle is that evidence must only be weighed once and all together, at the end of the case.
Justice McIntyre seemed to ignore this in O’Malley, supra, when he implied that the prima facie test requires some level of proof. The same paragraph contains the puzzling statement:
“In some cases [a prima facie case] may be established without evidence; for example, a requirement that all employees work on Saturday in a business which is open only on Saturdays, but once the prima facie proof of a discriminatory effect is made it will remain for the employer to show undue hardship if required to take more steps for its accommodation than he has done.”
One might be tempted to think that the Complainant must actually prove something, though the term “prima facie case” (a notoriously unreliable term at the best of times, as Sopinka and Lederman attest) does not really permit this analysis.
Under O’Malley, human rights tribunals are required to follow a two-stage process at the end of the case. This is problematic in itself. The prima facie test was never intended to be used at the end of a hearing, after the defendant has presented his case. The current law requires too much of Tribunals, who are expected to restrict themselves to the evidence of the Complainant in deciding whether the burden should shift. Anyone who has read the decisions in the area will nevertheless know that a tribunal is expected to apply the exact formula in O’Malley, if it wants to stay out of trouble in the appellate courts.
The problem is that the caselaw has never considered the full effect of the prima facie test in the general run of cases. It is true that there are a few cases in which the complainant cannot establish a prima facie case. But these cases are rare; exceedingly rare in point of fact. In the vast majority of cases, the ruling in O’Malley prevents a Tribunal from weighing the evidence presented by the Complainant on the prima facie test. The evidence must not be weighed. This reverses the entire burden of proof and places it on the Respondent.
The significance of the issue is easily illustrated. The fact that a Complainant has succeeded in passing the prima facie test says nothing about the strength of its case. The evidence may fall egregiously short of what is needed to prove discrimination on a balance of probabilities. The burden nevertheless shifts. A Respondent who decides not to lead evidence will lose the case, even if the case for the Complainant is highly dubious. This is a drastic development. Personally, I find it hard to believe that the Supreme Court intended to change the fundamental rule that the party who brings a case must prove it, without an explicit statement to that effect.
There is no room for subterfuge on these kinds of issues. The usual response is that the Complainant needs to hear the evidence of the Respondent. There are other ways of assuring this, however. Some of the earliest caselaw recognizes that there is a deep and abiding difference between introducing adverse inferences or presumptions into the law, and a procedure that reverses the burden of proof, without any weighing of the evidence whatsoever. Yet that has been the legacy of O’Malley v. Simpsons-Sears: more strange yet, no one seems to have questioned it. One has to wonder why.
In all fairness, McIntyre J. seemed to be contemplating a notion like strict liability, where the prosecution is only obliged to establish the basic elements of the case. That would make sense, since the real issue in O’Malley was the silence of the Respondent. This does not require a reverse onus, however, or dispense with the requirement that the prosecuting party prove his case. Other matters aside, the decision in O’Malley demonstrates why it is important to have practitioners on the court, who are thoroughly familiar with the requirements of litigation.
There is no reason to overstate the situation. Human institutions are fallible, and it is inevitable that the courts will occasionally make mistakes, or choose the wrong words to express their ideas. The real problem is that it can be very hard to fix the error, when it is enshrined in decades of law and practice. It is Respondents who are burdened by the existing law; but a Respondent who decides to pursue the issue can expect to lose at every stop in the appeal process. The Federal Court of Appeal has certainly made it clear, in Lincoln v. Bay Ferries Ltd., 2004 FCA 204 (at para 22), that it will not meddle in the matter.
The use of the prima facie test in the law of human rights is interesting, historically, because it demonstrates how a stray passage from our highest court can inadvertently alter the foundations of the law. I am not sure what the solution is; but there is only one court that can realistically correct the situation, and somehow, one has to hope that the issue will eventually reach the ears of the Supreme Court. There has been enough of subterfuge.