Executive Summary: Key Legal and Evidentiary Issues

• Scope of the authorized class action was limited to the gap between theoretical fuel consumption ratings and consumers’ real-world fuel use, not to challenging the validity of the official ratings themselves.
• Expert evidence for the plaintiff (ProLad) on fuel consumption was found methodologically flawed and insufficient to prove that Mazda’s government-approved fuel consumption ratings were erroneous.
• Defence expert evidence (Harrington and Assaf) persuasively supported that SkyActiv vehicles’ fuel consumption was generally consistent with the published ratings under normal driving conditions.
• Advertising materials, including fine-print disclaimers and references to government-approved test methods, were held not to be misleading when viewed as a whole under the “impression générale” test in the Consumer Protection Act.
• No breach of the statutory warranty of conformity (art. 41 L.p.c.) or of misleading representation / omission provisions (arts. 218, 219, 220, 221, 228 L.p.c.) was established on a class-wide basis.
• The class action was dismissed, with Mazda successful on liability and no damages or price reduction awarded, while the representative plaintiff was ordered to pay ordinary court costs but not the defendant’s expert fees (each party bearing its own expert costs).

Factual background and the SkyActiv fuel economy promise

The case arises from a Québec class action brought by consumer Katia Grand-Maison against Mazda Canada Inc. in relation to Mazda3 vehicles equipped with the SkyActiv engine and transmission technology, for model years 2012, 2013 and 2014. The class comprised Québec residents who bought or leased new Mazda3 SkyActiv vehicles from Mazda dealers in the province. The dispute centered on fuel consumption: Grand-Maison alleged that the SkyActiv Mazda3s did not achieve the fuel savings advertised by Mazda and that the fuel consumption ratings used in Mazda’s advertising and promotional materials were either erroneous or misleading. The relevant fuel consumption ratings (the “Cotes”) are set by manufacturers through standardized laboratory certification tests based on protocols derived from the U.S. Environmental Protection Agency (EPA) and incorporated into Canadian federal guidelines. These Cotes are transmitted to Natural Resources Canada (RNCan), appear in RNCan’s Fuel Consumption Guide, and are used on mandatory EnerGuide labels affixed to new vehicles. The Cotes are meant primarily as a comparative tool between different models and are acknowledged in the RNCan Guide as theoretical values that real-world consumption will diverge from due to factors such as driving style, vehicle condition, temperature, traffic, and terrain. Mazda Motor Corporation in Japan conducted the certification tests and established the Cotes for Mazda3 SkyActiv vehicles; Mazda Canada then used those government-approved figures in its marketing in Canada. The SkyActiv technology combined a high-compression direct-injection engine with a six-speed transmission, designed to reduce fuel consumption compared with a conventional Mazda3 powertrain. Evidence showed that SkyActiv vehicles did consume less fuel than comparable non-SkyActiv versions, but the extent of the fuel savings became the central factual and legal controversy.

The plaintiff’s purchase and expectations

Grand-Maison, looking for an economical car, saw Mazda’s advertisements referring to specific Cotes for the Mazda3 SkyActiv: 4.9 L/100 km on the highway and 7.1 L/100 km in city driving for the 2012–2013 models, with slightly lower figures for 2014. She read the small-print disclaimer stating that these Cotes were estimates based on government-approved test methods and that actual results could vary. She also consulted the 2012 RNCan Fuel Consumption Guide, which clearly explains that published ratings are estimates, may differ from real-world use because of numerous driver, vehicle and environmental factors, and are intended as a basis of comparison rather than a guarantee of actual consumption. Despite understanding that exact published numbers would not be replicated, Grand-Maison testified that she expected any variance to be minimal and that, with eco-conscious driving, her real-world fuel use would be comparable to the Cotes. In March 2013, she and her partner purchased a Mazda3 GS-Sky 2012 from a dealer, paying several thousand dollars more than the base GX model. She drove mainly short daily commuting trips of about 7 km each way, mostly on the highway but with a significant city component, and logged 10,000–15,000 km per year. After purchase, she read the owner’s manual, which reiterated fuel-saving driving practices (no extended idling, gentle acceleration, proper maintenance, tire pressure, limiting load, etc.), and she testified that she followed these recommendations. Nonetheless, she soon perceived that her fuel consumption materially exceeded the Cotes, at around or above 9 L/100 km. After multiple complaints, dealer checks, and consumption tests, the dealer confirmed consumption of about 9.46 L/100 km and told her that this was normal for that model, although other customers were also disappointed. Grand-Maison maintained that had she known her real consumption would be so much higher than the Cotes, she would not have purchased the vehicle.

The authorized class action and causes of action

In 2016, the Superior Court, per Yergeau J., authorized a class action focused on the alleged discrepancy between the theoretical fuel consumption announced via the Cotes and the “consommation réelle” actually experienced by consumers on Québec roads. The authorized issues asked essentially whether Mazda had falsely represented fuel consumption in contravention of Québec’s Consumer Protection Act (Loi sur la protection du consommateur, L.p.c.), whether this fault engaged Mazda’s liability, and whether the class members were entitled to (a) reimbursement of the price difference between SkyActiv and regular Mazda3 models, (b) reimbursement of excess fuel costs, and (c) punitive damages. The authorization judgment accepted that real-world fuel consumption cannot be guaranteed, but noted that Mazda’s advertising emphasized the Cotes themselves rather than their comparative function, potentially creating expectations that could be disappointed if actual performance substantially overshot the ratings. At trial, Grand-Maison advanced two related theories. Her principal cause of action asserted that the Cotes themselves were erroneous—that Mazda’s SkyActiv fuel ratings overstated fuel savings compared to non-SkyActiv Mazda3s—thus breaching the statutory warranty of conformity and the prohibition on false or misleading representations (particularly articles 41, 219, 220(a) and 221(g) L.p.c.). Her subordinate cause of action accepted the Cotes as valid in a laboratory sense but alleged that using them in advertising without more context was misleading, because typical drivers could not realistically achieve similar consumption, engaging articles 41, 218, 219 and 228 L.p.c. and opening the door to remedies under article 272 L.p.c.

Expert evidence for the plaintiff: challenging the ratings

The plaintiff’s case rested heavily on the expert work of two engineers from ProLad, Bellavigna-Ladoux and Meunier. They had general vehicle engineering experience but limited specialized background in fuel consumption certification testing. ProLad conducted two sets of comparative fuel tests between SkyActiv and regular Mazda3s: chassis dynamometer tests in a college lab and on-road tests over a route approximating Grand-Maison’s commute. In both sets of tests, they found that SkyActiv models indeed consumed less fuel than regular models, but the measured savings (roughly 6–13%) were materially lower than the relative savings implied by the Cotes (about 18–23%). From this, they concluded that Mazda’s official Cotes overstated the fuel economy advantage of SkyActiv vehicles and were therefore “erroneous.” Their testing, however, was not intended to replicate the certification tests used to generate the Cotes, and they worked with older, high-mileage vehicles rather than nearly new ones as in formal certification protocols. The court scrutinized their methodology and the inferential leap from differences in comparative results to a finding that the original Cotes themselves must be wrong.

Defence expert evidence: supporting the ratings and real-world performance

Mazda responded with two key expert lines of evidence. First, it called Ryan Harrington, from Exponent, an engineer with deep, specialized experience in regulatory fuel economy testing (EPA and similar protocols). Harrington explained how certification tests work, emphasized that engine efficiency varies with operating parameters, and stated that one cannot reliably compare results from different test protocols unless it is shown that vehicle operating conditions are essentially identical. He highlighted several methodological flaws in ProLad’s lab testing, including: altering the EPA two-cycle procedure in ways that disadvantaged SkyActiv vehicles (e.g., removing the cold-start and dynamic segments where SkyActiv gains are strongest); modifying resistance coefficients rather than using validated values, without proper dynamometer calibration or speed-varying resistance modelling; and failing to demonstrate that the driver followed the prescribed speed-time profile within the tight tolerances required for valid dynamometer testing. For the on-road tests, he noted wide variations in temperature, wind, traffic and trip times, showing that environmental and congestion factors were not controlled and could easily swamp the differences the plaintiff sought to measure. In his view, ProLad’s deviations and omissions were sufficient by themselves to explain why their measured fuel savings differed from those implied by the Cotes. Harrington also pointed to independent EPA and Environment Canada testing of Mazda3 SkyActiv models, as well as an EPA benchmarking study of the SkyActiv engine, all of which were consistent with Mazda’s published ratings rather than suggesting overstatement. Second, Mazda relied on the work of Professor Gabriel Assaf, an expert in transport engineering and vehicle operating costs, who conducted extensive on-road testing of three SkyActiv Mazda3s over varied real-world conditions. Across almost 1,900 km of mixed driving, Assaf found an average real-world consumption of 6.73 L/100 km, very close to the Cotes weighted for the mix of routes used (6.42 L/100 km). Although his methodology had limitations (use of pump shutoff rather than scales, some inconsistency in fuel octane), he convincingly testified that, over the entire dataset and numerous refuellings, these imperfections did not materially affect the overall average. His results supported the proposition that, in typical driving, SkyActiv Mazda3s can achieve fuel consumption broadly consistent with the published ratings. Additional support came from the Eco-Run event organized by automotive journalists and RNCan, where SkyActiv vehicles, including a 2013 Mazda3, achieved fuel consumption better than the Cotes under real-world driving, showing that the ratings were at least technically attainable and not unrealistic on their face.

Policy terms and statutory provisions at issue

Although the case did not involve an insurance or indemnity “policy” in the contractual sense, it turned centrally on statutory and regulatory frameworks governing consumer representations and fuel consumption data. Under article 41 L.p.c., goods must conform to declarations or advertising made about them; advertising binds the merchant or manufacturer. The plaintiff argued that if the Cotes overstated fuel savings, or if real-world consumption significantly exceeded them, the vehicles were not “conform” to Mazda’s messages. Articles 219, 220(a), and 221(g) L.p.c. prohibit false or misleading representations, including falsely attributing an advantage or a particular performance characteristic to a product. Article 218 L.p.c. instructs courts to assess the “impression générale” of a representation, not just its literal wording. Article 228 L.p.c. bars merchants from omitting important facts in representations—facts that could reasonably influence a consumer’s decision to contract. Finally, article 272 L.p.c. sets out the remedial toolbox (rescission, price reduction, damages, punitive damages) when a prohibited practice is established. In parallel, federal testing guidelines and RNCan’s Fuel Consumption Guide played a critical contextual role. The guidelines require automakers who advertise fuel consumption to use only ratings derived from approved laboratory tests, and the RNCan Guide expressly warns that no test can simulate all real-world conditions, that actual consumption will differ from ratings, and that ratings are primarily a comparative tool. These regulatory parameters shaped both what Mazda was allowed to say and what a reasonable consumer could infer about fuel consumption claims.

Court’s analysis of the principal cause of action: alleged erroneous ratings

The court first held that the plaintiff’s principal theory—that the Cotes themselves were erroneous—exceeded the scope of the action as authorized at the certification stage. Yergeau J.’s authorization judgment focused on the “hiatus” between theoretical Cotes and consumers’ actual fuel consumption, not on the intrinsic validity of the Cotes or the correctness of Mazda Motor Corporation’s certification testing. By turning the trial into an indirect attack on the official ratings, using ProLad’s alternative testing to allege that the Cotes were wrong, the plaintiff introduced a new cause of action that had not been authorized and should have been brought, if at all, by way of a modification of the authorized proceeding with fresh authorization analysis. Even assuming this new cause were properly before the court, the judge found it unproven. Given the regulatory context, independent EPA and Canadian testing, and Harrington’s evidence, the court concluded that there was no sound evidentiary basis to displace the certified Cotes. ProLad’s methodological flaws and the impossibility of direct comparison between their tests and the original certification regime meant that their results could not reliably show that Mazda’s published Cotes were wrong. Moreover, even if differences in comparative savings were accepted, ProLad’s data could as easily suggest that regular Mazda3 ratings were understated as that SkyActiv ratings were overstated. On this record, the plaintiff failed to show on a balance of probabilities that the official fuel consumption ratings for Mazda3 SkyActiv vehicles were erroneous.

Court’s analysis of the subordinate cause of action: alleged misleading use of ratings

Turning to the authorized, subordinate theory, the court applied the two-step test from the Supreme Court’s decision in Richard v. Time to determine whether Mazda’s advertising representations were false or misleading under the L.p.c. First, the court asked what general impression the advertising would give to a credulous and inexperienced consumer. Mazda’s print ads, brochures and website repeated the Cotes in small data boxes and often used prominent slogans emphasizing fuel savings and efficiency (“Less fuel, more fun”, “Drive more, fuel less”, “Make every kilometre count”). At the same time, every ad that used the Cotes included a fine-print disclaimer stating that the figures were estimates based on government-approved test methods, that real-world results may vary, and, in some instances, that the ratings were for comparison purposes only. The RNCan Guide—which Grand-Maison herself read—reinforced that message by explaining the experimental nature of the ratings and the many factors that cause actual consumption to deviate. The court accepted that fine print in small characters may visually suggest lesser importance, but held that the text was legible, intelligible and sufficiently clear that it could not be ignored in assessing the overall impression. Considering all elements together, the judge held that a credulous, inexperienced consumer would (1) understand that SkyActiv vehicles promise lower fuel consumption than conventional Mazda3s, (2) reasonably expect that, in general, real-world consumption would be broadly comparable to the Cotes, but (3) also be aware that there can be exceptions and that real-world consumption may differ depending on conditions and usage. Second, the court compared this general impression to reality. Evidence, chiefly from Assaf and the regulatory testing record, demonstrated that SkyActiv Mazda3s do, in fact, consume materially less fuel than comparable non-SkyActiv models and that, in ordinary driving, they can achieve real-world consumption broadly consistent with the published Cotes. The plaintiff’s contrary expert evidence did not validly undermine this conclusion. Grand-Maison’s personal experience, while acknowledged as genuine, was found to be an outlier driven by her specific usage pattern—very short trips on a cold engine—which literature and expert evidence show can dramatically increase fuel consumption. No evidence established that this pattern was typical of the class at large. In light of these findings, the court concluded that the general impression created by Mazda’s advertising—that SkyActiv Mazda3s generally deliver fuel economy in line with the rated figures, subject to variation—was consistent with the real-world performance of the vehicles. Consequently, Mazda’s advertising did not constitute false or misleading representations under articles 219, 220 or 221 L.p.c.

Assessment of conformity, omissions and the consumer protection framework

On the statutory warranty of conformity under article 41 L.p.c., the court reiterated that this provision is engaged only where there is a significant non-conformity between the advertised attributes and the product’s actual performance. While Grand-Maison’s vehicle consumed significantly more fuel than the Cotes, the evidence showed this was attributable to her driving pattern rather than any systematic mismatch between the advertised SkyActiv performance and typical consumer experience. The plaintiff offered no reliable class-wide evidence (e.g., surveys, usage studies) establishing that members generally experienced real-world consumption materially higher than the Cotes. Nor did ProLad’s work provide a scientifically valid bridge from certification figures to typical on-road consumption. The court therefore found no proven class-wide defect of conformity. On article 228 L.p.c. (omission of an important fact), the plaintiff initially argued that Mazda had failed to disclose the “true” real-world consumption of the vehicles. The court rejected this as impossible in principle: real-world consumption is inherently variable and cannot be distilled into a single “true” figure; the law cannot require manufacturers to disclose a non-existent fixed value. At trial, the plaintiff reframed the alleged omission as Mazda’s failure to disclose either that consumers could not realistically achieve Cotes in normal use, or all the factors affecting consumption. The court held that both theories failed. Evidence from the Eco-Run and Assaf’s tests showed that it is indeed possible to match or beat the Cotes in real-world conditions, so Mazda could not be said to have omitted an impossibility. As to the many factors affecting consumption, these are largely generic to all vehicles and not determinative of a consumer’s decision to buy a Mazda3 versus another car; they are therefore not “faits importants” within the meaning of article 228. The truly important facts in this context are that the Cotes are lab-derived comparative estimates and that real-world results may vary—facts which Mazda repeatedly disclosed in its materials. The court accordingly found no violation of articles 41, 218, 219 or 228 L.p.c.

Remedies, costs, and overall outcome

Because no breach of the Consumer Protection Act was established, the court did not need to grant any remedies under article 272 L.p.c. It nevertheless commented on the plaintiff’s claimed relief. The proposed class-wide compensatory damages for excess fuel costs were based on ProLad’s assumption that all class members drove 20,000 km per year for seven years on a fixed city/highway split. The court found no evidence supporting these usage assumptions and noted that the representative plaintiff herself drove significantly fewer kilometres and retained her car for only four years. Even if liability had been found, the court would have considered ProLad’s damages model unreliable and would have had to arbitrate any compensation on more conservative and individualized bases. The alternative remedy of a price reduction equal to the full difference between SkyActiv and base models also failed conceptually. Many of the price differences reflected added equipment and features, not just the SkyActiv drive-train, and the evidence showed that SkyActiv did deliver genuine fuel savings compared to non-SkyActiv models. Hypothetically, the judge suggested that if SkyActiv’s benefit had been proven smaller than advertised, an appropriate reduction might have reflected only the diminished portion of the fuel-economy advantage, not the entire technology premium. On punitive damages, the court held there was no evidence that Mazda intended to mislead consumers or acted with serious disregard for their rights; Mazda operated within a tightly regulated regime that required it to use government-approved figures when advertising fuel consumption, and it included appropriate disclaimers. Punitive damages would not have been warranted even had a technical breach been found. On costs, the court dismissed the action with costs against Grand-Maison in the ordinary sense (court costs), but expressly departed from the usual rule that the losing party pays all expert fees. Defence experts’ fees exceeded CAD 400,000, an amount the court considered prohibitive and incompatible with access-to-justice goals underlying class actions. Each party was therefore ordered to bear its own expert costs, while ordinary court costs followed the result. In the final analysis, Mazda Canada Inc. emerged as the successful party. The class action was fully dismissed, no class member—including Grand-Maison—received any compensatory, punitive or price-reduction damages, and no specific monetary sum was awarded against Mazda; only unquantified ordinary court costs were ordered against the plaintiff, with each side paying its own expert expenses, so the total financial amount ordered in Mazda’s favour cannot be precisely determined from the judgment.