BC Court of Appeal Confirms Low Bar for Certification: Lessons from McKinsey & Company v British Columbia
On June 25, 2026, the British Columbia Court of Appeal released its decision in McKinsey & Company v British Columbia, dismissing McKinsey's appeal from an order certifying a class action against it as part of the broader opioid litigation in Canada. The decision is significant for both consultants and other professionals, but it is also important for anyone tracking the “one-step versus two-step” debate over how plaintiffs prove commonality at certification. The Court of Appeal used the case to firmly entrench British Columbia's functional, single-inquiry approach – and to explicitly part ways with the Federal Court of Appeal's reasoning in Jensen v Samsung Electronics Co Ltd.
Background
The action is a companion proceeding to the Province of British Columbia’s long-running opioid class action against pharmaceutical manufacturers, wholesalers, and distributors. This proceeding targets McKinsey, which the Province alleges provided marketing and sales advisory services to various opioid manufacturers and distributors. The Province says McKinsey developed strategies in the United States – including campaigns minimizing addiction risk and promoting higher dosing – that were also designed for, and used in, the Canadian market, giving rise to claims in conspiracy, common design, breach of the Competition Act, and statutory “opioid-related wrongs” under British Columbia's Opioid Damages and Health Care Costs Recovery Act.
McKinsey did not dispute that the proposed common issues were, in the abstract, “common.” Its sole argument was that there was no evidentiary basis to tie those issues to McKinsey at all, because the evidence showed McKinsey was never retained to advise on opioid marketing in the Canadian market specifically. The Province’s evidence, by contrast, relied heavily on documents from the U.S. Opioid Industry Documents Archive, McKinsey’s own description of itself as “a single global partnership,” internal communications referencing engagements in “North America” and meetings with Purdue Canada, and expert opinion suggesting pharmaceutical marketing methodology does not vary meaningfully between the U.S. and Canadian markets.
The Decision Below
Justice Brundrett, the same case-management judge who certified the Main Action, certified the claim against McKinsey, concluding the Province had established “some basis in fact” for commonality under section 4(1)(c) of the Class Proceedings Act. He found McKinsey’s evidence – principally an affidavit from a former senior McKinsey partner denying any Canadian opioid-marketing engagements – was not “definitive,” given its reliance on hearsay and an unclear scope of document review. Justice Brundrett treated McKinsey’s denial of direct Canadian involvement as a merits question reserved for trial, not a certification-stage bar.
Critically, the judge characterized the Province's claim as resting on “more expansive allegations of agency, common design, and civil conspiracy” – meaning the Province did not need direct evidence that McKinsey performed marketing work in Canada specifically; it needed only “tenable links” suggesting McKinsey’s U.S. work product could plausibly have been built upon for the Canadian market. He found that threshold met, particularly with respect to Purdue, and certified the action.
The Court of Appeal’s Decision
Writing for a unanimous panel, Justice Fisher upheld the certification order on both grounds of appeal. On the question of law – what “some basis in fact” requires – the Court reaffirmed that certification is concerned with the form of the action, not its merits. A plaintiff need not prove that the underlying facts of the claim actually occurred. The Court held that the evidentiary burden goes only to establishing that an issue is capable of common resolution across the class – not to establishing that the alleged conduct happened.
The most consequential part of the judgment is the Court’s treatment of the “one-step versus two-step” controversy. McKinsey argued, relying on the Federal Court of Appeal’s decision in Jensen, that plaintiffs must satisfy a two-part test: first showing some basis in fact that the alleged issue exists, and second that it is common. The British Columbia Court of Appeal rejected this framework outright, holding there is “no utility” in continuing to debate whether a one-step or two-step test applies. Instead, the Court endorsed a single, functional inquiry: whether there is a minimum evidentiary foundation supporting the existence of a common issue, with “existence” typically implicit in the pleaded cause of action itself. The Court expressly declined to follow Jensen’s framing that the commonality requirement exists to “root out unfounded and frivolous claims,” noting that such language belongs to merits-testing exercises like motions to strike, not to certification’s procedural screening function.
This puts British Columbia in clear, acknowledged tension with the Federal Court and with parts of Ontario’s approach. The Court distinguished – rather than rejected – the results in the Ontario Court of Appeal’s Lilleyman v Bumble Foods and the British Columbia Supreme Court’s O'Connor v Canadian Pacific Railway Limited, both of which denied certification where there was literally no evidence of the alleged conduct. The Court characterized those cases as illustrating “no basis in fact,” not authority for a formal two-step test. McKinsey’s case was different: the Province’s broader theory of agency and common design meant evidence of U.S. conduct, combined with expert evidence on the cross-border uniformity of pharmaceutical marketing, was capable of supporting a Canadian common issue even without direct proof that McKinsey performed Canadian-specific work.
Applying this to the record, the Court found no palpable and overriding error in the certification judge’s assessment. The Court also endorsed the certification judge’s refusal to treat McKinsey’s denial as decisive, holding that weighing conflicting lay and expert evidence is squarely a trial function, not a certification-stage one.
Key Takeaways
Three practical implications stand out for organizations with Canadian exposure.
For consulting, advisory, and professional services firms with cross-border operations, this decision confirms that, at least in British Columbia, a “we were never retained for Canadian work” defence may no longer defeat certification on its own – particularly where a plaintiff’s theory is based on agency, common design, or group enterprise liability rather than direct Canadian conduct. Internal characterizations of a firm as a unified “global partnership,” cross-border staffing references, and any documentary trail suggesting domestic affiliates received or adapted foreign work product can constitute sufficient “tenable links” at this low threshold, even where the underlying facts remain genuinely contested. Firms should expect that sworn evidence denying direct Canadian engagement – however sincere – will be treated as raising a merits dispute for trial rather than a basis to resist certification.
Second, companies facing class actions in multiple Canadian courts should be aware that what works in one jurisdiction may not work in another. This decision is essential reading for any organization litigating, or anticipating, class proceedings in British Columbia. By explicitly rejecting the Jensen two-step framework and emphasizing that British Columbia’s commonality test is not designed to screen out “unfounded” claims, the Court has widened the gap between British Columbia’s certification jurisprudence and that of the Federal Court (and arguably parts of Ontario). Counsel advising clients with national or multi-jurisdictional class action exposure should not assume that arguments successful in resisting certification federally, or in Ontario, will translate to British Columbia.
Finally, this decision reinforces that British Columbia courts will continue applying highly deferential appellate review to certification rulings made by experienced case-management judges, especially in complex multi-defendant proceedings like the opioid litigation. Defendants seeking to overturn certification face a steep climb unless they can point to a clear error in principle.