The regulatory world was stunned two years ago when Ontario’s Divisional Court struck down a registration requirement for certification of teachers. In part, regulators were disconcerted because significant efforts had been made to ensure that the requirement was equitable.
The Divisional Court, relying on statistical data and on research studies from other countries, held that the disproportionate failure rate on the mathematics proficiency test (MPT) by racialized groups (e.g., Black and Indigenous identifying candidates) resulted in the requirement violating the equality provision (section 15) of the Canadian Charter of Rights and Freedoms.
Ontario’s Court of Appeal, while upholding the principle that registration requirements must not have a discriminatory effect, set aside the Divisional Court’s decision: Ontario Teacher Candidates’ Council v. Ontario (Education), 2023 ONCA 788.
The Court of Appeal accepted that a registration requirement would breach the equality provisions of the Charter if:
- It created a distinction based on a protected ground, including through a disproportionate impact of an apparently neutral requirement,
- It imposed a burden (or denied a benefit) that had the effect of reinforcing, exacerbating, or perpetuating disadvantage, and
- It cannot be saved as a reasonable limit imposed by law to achieve an important purpose (i.e., that is demonstrably justified in a free and democratic society).
On the first issue, the Court of Appeal concluded that the data upon which the lower court based its decision was preliminary and incomplete. It consisted of data from only the pilot tests and the first seven weeks of the test’s initial administration. Subsequent data, available to the Court of Appeal, for the remaining five months of the year demonstrated that 93% of candidates from racialized groups were able to pass the MPT (including retakes) compared to 95% of all candidates and 97% of White candidates. Since the MPT was discontinued and candidates could not retake the MPT, “the ultimate disparities in relative success rates between different demographic groups might well be even smaller than the relatively modest differences observed in the December 2021 Data.” There were also concerns about the small numbers of self-identified candidates in various demographic groups that made the statistical data before the Divisional Court less reliable. In fact, the data for the rest of the year showed a marked improvement in outcomes for racialized candidates. The Court of Appeal concluded that it was an overriding and palpable error to make such an important ruling on inadequate data.
The Court recognized that the disparity in pass rates for first attempts was larger. However, candidates could retake the MPT immediately and there was no evidence to support the concern that requiring racialized candidates to retake the test more frequently delayed their registration or caused them to lose out on job opportunities. There was no fee to write or retake the MPT.
The Court noted the relatively “modest” disparity in results. Precedents resulting in judicial findings of discrimination involved a larger degree of disproportionate outcomes. Since the test was terminated (as a result of the Divisional Court ruling) it was possible that candidates who had failed would have retaken and passed the MPT. The immediate and frequent retake policy also included that the regulator would not be informed of unsuccessful MPT attempts.
Despite being in a position to grant the appeal on the first point, the Court of Appeal went on to discuss the second point. Even accepting that there is a “diversity gap” among racialized teachers with the profession, the Court was unable to conclude that, on the record before it, the MPT would reinforce, perpetuate, or exacerbate disadvantage.
The Court noted the efforts made by the test creators to address equity concerns. All test questions were screened on that basis. Adjustments were made to the format (away from traditional multiple-choice questions) and administration (increased availability of test centres) of the MPT to accommodate candidates. The difficulty level of the questions was reduced from Grade 11 (and lower) math to Grade 9 (and lower). An exemption was created for teachers of Native Languages. In addition, likely as a result of the MPT, faculties of education were expanding math instruction within their curricula. Candidates were permitted to attempt the MPT while in school to further reduce the impact of any initial unsuccessful attempts.
As a result, the Divisional Court erred in its reliance on expert evidence of general standardized testing outcomes, especially from studies from the US and the UK.
The Court of Appeal found it unnecessary to deal with the third issue.
The Court also commented that the Divisional Court order was overly broad in that it did not allow for alternate, compliant, math proficiency examinations.
Regulators will still wish to ensure that their registration requirements do not have a disproportionate impact on protected groups that reinforce, perpetuate or exacerbate disadvantage. Standardized tests may require appropriate structure and accommodation. However, the evidence to establish discrimination in an exam or other registration requirement cannot be speculative.