This article was originally published in the American Bar Association’s International Antitrust Bulletin.

The evaluation and treatment of data and “big data” in antitrust has become an increasingly popular topic of debate and discussion in international antitrust circles. In a global economy that is increasingly driven by disruption and innovation, antitrust authorities have been contemplating when and how traditional antitrust paradigms might apply and whether the existing tools and legislation are sufficient to address any antitrust concerns that may arise.

On September 18th, Canada’s Competition Bureau put out a draft discussion paper for consultation titled “Big Data and Innovation: Implications for Competition Policy in Canada.”¹ With a view to sparking further dialogue with stakeholders, the discussion paper provides the Bureau’s perspective on the role of data (including collection, use and access to such data) in the context of merger review, abuse of dominance cases, cartels and the application of the misleading representations provisions of the Competition Act. Key aspects of the Bureau’s position as outlined in the discussion paper as well as some practical implications are discussed in more detail below.

Role of Data in Merger Review and Monopolistic Practices

Since certain aspects of antitrust analysis (such as market definition and market power) are essential to both merger review and abuse of dominance cases, the role of big data in these two areas is discussed together by the Bureau. Before delving into a more detailed discussion, the discussion paper acknowledges at the outset that “competition policy in Canada does not, and should not, assume that ‘big is bad’.” The Bureau also acknowledges that “firms are increasingly harnessing big data in ways that drive innovation and quality improvements across a range of industries.” As a result, while the Bureau is mindful of the need to avoid “under-enforcement,” it is also mindful of “over-enforcement” which “risks slowing or even stopping such advances.” This need to approach enforcement and intervention in a balanced manner is a welcome acknowledgement and helps to set the overall tone of the discussion paper.

A. Data and Market Definition – Increased Reliance on Direct Effects

Challenges in applying existing tools to platforms or digital markets: The discussion paper outlines the complexity of applying traditional market definition tools to mergers and acquisitions where big data or multi-sided platforms are involved. In particular, the Bureau refers to the difficulties of applying the traditional hypothetical monopolist test (used to assess substitutability between products) where products are made available at no cost or where multi-sided platforms are involved. The Bureau posits that because of such difficulties, in certain cases involving big data or platforms in the digital economy, it “may be appropriate to rely on alternative methods to assess market definition or to forgo market definition as an initial step and focus on direct evidence of competitive effects.”

Direct evidence of anticompetitive effects: The suggestion that the Bureau may skip the market definition step is not a new one (the existing Merger Enforcement Guidelines² refer to this possibility). However, market definition has been undertaken as a first step by the Competition Tribunal in virtually all of the contested abuse of dominance and merger cases to date. In fact, while the Competition Bureau proposed to skip the market definition step in Canada Pipe based on evidence of direct effects, the Competition Tribunal declined to do so.³ In addition, for the same reasons that the Bureau believes the application of the hypothetical monopolist test is challenging in digital markets, any assessment of “direct effects” is also likely to raise numerous challenges for the Competition Bureau.

B. Data and Market Power – Big is Not Necessarily Bad

Data can be a barrier: Again, the Bureau acknowledges that “developing valuable data through competition on the merits does not run afoul of the [Competition] Act even if it results in significant market power.” However, the discussion paper also provides that in certain cases, “access to and control over critical data that serve as an essential input may confer market power” and may also represent a barrier to entry. In particular, the Bureau notes that access to data may act as a barrier as a result of network effects or if a firm has exclusive access to proprietary data.

Data may make market shares less important: The Bureau also notes in cases involving data, market shares, which are traditionally a key input into an assessment of market power, may either overestimate or underestimate a firm’s market power. In particular, the Bureau notes that when assessing the acquisition of a firm with low market share but “troves of valuable data,” agencies should assess post-acquisition incentives and the importance of the data being acquired.

Denying access to data may be exclusionary: With respect to access to data, the Bureau notes that antitrust “does not usually impose on firms an obligation to share data that they have collected and developed.” However, the Bureau also notes that incumbents can try to prevent their competitors from obtaining data that is necessary to compete. In evaluating whether such acts are truly competition on the merits or anti-competitive acts, the Bureau will apply the “no economic sense” test as applied in the Tribunal’s TREB⁴ re-determination decision to the conduct in question. Here, it is interesting that the Bureau has singled out the “no economic sense” test particularly when other possible tests have been discussed and adopted by the Tribunal in the past.

C. Assessment of Competitive Effects

Impact on future competition: The Bureau acknowledges that the assessment of whether a merger or practice is likely to result in a substantial prevention of competition (rather than a lessening of competition) is challenging in cases involving data given the need to understand not just the current uses but also the future or potential uses for the data. As the Bureau also acknowledges, however, any theory of impact upon competition is required by Canadian jurisprudence to be grounded in existing evidence. As the Supreme Court of Canada noted in its decision in Tervita, “factual findings about what a company may or may not do must be based on evidence of the decision the company itself would make; not the decision the Tribunal would make in the company’s circumstances.” From a practical perspective, a company’s business plans and documents will undoubtedly continue to play an important role in the assessment of probable future competitive effects in cases involving data.

Likelihood of coordinated effects: The Bureau notes that where big data is involved, co-ordinated effects may be facilitated by a merger or practice in two ways: (i) by removing an additional constraint on co-ordination in a market that is already conducive to coordination (e.g., there is already transparent pricing data available) or (ii) where data is made more readily available or transparent as a result of the merger or practice. The Bureau cites the example of an acquisition of a maverick firm and notes that such an acquisition may be more problematic in a market where big data would facilitate co-ordination.

Non-price effects include impact on privacy: The discussion paper notes that in addition to assessing the impact of the practice or merger on dimensions such as innovation and quality, another important non-price effect to be evaluated is the impact on privacy. While consideration of such effects is again, not new, the Bureau also acknowledges that quantifying such non-price effects is challenging and “could have consequences for provisions of the Act that include an efficiencies defence.” While the Bureau is clearly signalling the difficulties associated with the assessment and quantification of anti-competitive effects (as required by the Tervita decision) particularly where non-price effects are involved, the discipline imposed by quantification is a necessary hurdle to protect against over-enforcement.

Big Data and Cartels: Collusion in an Age of Algorithms

Algorithms and agreements: A hot topic in antitrust circles has been whether the use of algorithms and big data might facilitate tacit co-ordination as well as collusion that perhaps may go undetected. While the Bureau acknowledges that big data “may introduce more efficient and powerful ways to implement and manage a cartel . . . it does not constitute a new kind of activity.” The Bureau notes (correctly in our view) that even with the increasing sophistication of the tools involved, the offence is still rooted in the agreement itself. That is, if companies agree to use algorithms to implement an agreement with respect to pricing (as occurred in the UK and U.S. cases involving the sale of posters via Amazon⁵), then while such an agreement uses new tools to implement and monitor the agreement, the underlying offence is no different. With these statements, the Bureau appears to be denouncing suggestions made by certain commentators regarding the need to consider amending antitrust laws to encompass a broader concept of collusion.

Conscious parallelism is not criminal: Certain antitrust academics have suggested that the increased ability to use big data to monitor and react to competitors’ pricing data may lead to increased acts of parallel pricing which in turn, dampen competition. However, the Bureau notes that “big data is likely to introduce a difference of degree rather than a difference of kind when it comes to conscious parallelism.” The Bureau also notes that there is “broad consensus that the unilateral monitoring and responding to data collected on one's competitors is legal” and that altering the existing framework would be “unworkable.”

Big data and facilitating practices: The discussion paper notes that to the extent that big data and algorithms are used to engage in facilitating practices, such activity may be an indicator of the existence of an illegal price-fixing agreement and thus, may be examined by the cartels provision or alternatively, may be captured by section 90.1 of the Act (a civil provision which prohibits agreements between competitors that result in or are likely to result in an anti-competitive effect). However, given that section 90.1 also requires an agreement or arrangement, it is not clear that this provision could apply to situations where the facilitating practice falls short of an agreement.

Application of Misleading Representations Provisions to Data Collection

While the Bureau has historically applied the misleading representations provision to “practices where consumers were misled into purchasing a product or service,” the discussion paper notes that the “era of big data may warrant devoting greater attention to representations that mislead consumers into giving away their information.” This statement is an important signal as to the direction the Bureau intends to take with respect to statements and disclosures involving the collection and use of data.

Misleading statements re: collection and use of data: The discussion paper notes that when companies make misleading representations to consumers about how their data will be collected, maintained, shared or erased, such statements can be pursued under the misleading representation provisions. Here, the Bureau gives the example of the FTC’s proceedings against SnapChat for representing that its snaps would disappear forever when recipients could use several simple methods to save snaps indefinitely.

Inadequate disclosure may also be misleading: Not only will the Bureau examine whether the disclosure provided in respect of data collection and use is truthful, it will also examine whether the disclosure provided is sufficient for consumers to make an informed choice to provide such data. More specifically, the Bureau states that “companies are putting themselves at risk when they collect information that consumers would not expect to be collected in the normal course of business and only disclose this material information in terms and conditions that are likely to be overlooked by consumers.” The Bureau also states that “the collection and use of data that go beyond what consumers would reasonably expect increases the likelihood of deception.” From a practical perspective, however, this approach raises questions about what consumers reasonably expect re: data collection in today’s digital era. In addition, whether nondisclosure or inadequate disclosure would be sufficient to be “material” (which has been defined by the courts to mean whether the representation influenced a purchasing decision) remains to be seen. Finally, as noted by Canada’s Office of the Privacy Commissioner or “OPC” in its comments⁶ on the Bureau’s discussion paper, the collection of personal information without consent is already within the scope of the OPC’s mandate under Canada’s federal privacy sector legislation, the Personal Information Protection and Electronic Documents Act (PIPEDA). The OPC’s comments regarding its desire to work with the Bureau in this area suggest that perhaps the Bureau had not made the OPC aware in advance of its intention to increase its enforcement efforts in the privacy area. Regardless, the potential for the Competition Bureau to use the significant penalties available under the misleading advertising provisions means that companies should carefully consider the adequacy and correctness of its disclosure regarding the collection or use of data.

Conclusions and Implications

The discussion and debate regarding the role of big data in antitrust is likely to continue to evolve both in Canada and elsewhere. The Bureau’s discussion paper covers a number of areas and provides a useful first glance at the Bureau’s perspective on these evolving issues. While the paper does not, for the most part, represent a significant departure from the Bureau’s established policies or approaches, companies doing in business in Canada may consider taking the following practical steps:

• Companies using algorithms to monitor competitors’ pricing or make dynamic pricing decisions should continue to make pricing decisions unilaterally and avoid communications with competitors regarding approaches to pricing whether through algorithms or other electronic tools.
• Reviewing existing privacy policies and disclosure with respect to the use, collection and disposal of consumer data to ensure that adequate and proper disclosure is provided up front (and not simply in terms and conditions that, according to the Bureau, could be overlooked).
• Avoid the use of data collection that would be unexpected (for example, the collection of a consumer’s location data where a consumer purchases an app that provides another purpose).
• When evaluating the antitrust implications of a possible merger or other practice where one of the parties has significant data, take into account the effects of the proposed conduct on non-price effects such as privacy as well as quality and innovation.