Competition Bureau Releases Big Data White Paper for Public Comment
September 26, 2017
On September 18, 2017, the Competition Bureau (the “Bureau”) published a white paper for public consultation titled “Big data and Innovation: Implications for competition policy in Canada”. The white paper draws from the Bureau’s recent abuse of dominance investigations involving big data considerations, and also considers US and European developments in order to identify challenges raised by big data in the context of criminal cartels, mergers, and misleading advertising cases.
At the outset, the Bureau recognizes that competition enforcement needs to “strike a balance” that does not stifle innovation driven by the collection and use of data and legitimate competition. The Bureau considers that the existing legislative framework under the Competition Act (the “Act”) is largely effective in meeting the new challenges posed by big data. Nonetheless, given that the use of big data is new and developing at a fast pace, the white paper identifies challenges of analyzing big data cases under the Act.
Competition and Privacy
One question raised by competition investigations involving big data is the role of privacy and data security concerns. The Bureau indicates that while such considerations are relevant to a broader debate about big data, its mandate is limited to addressing conduct that harms competition. The white paper, however, does stress that when firms compete with respect to privacy safeguards or transparency in respect of how their data may be used, privacy represents a significant non-price dimension of competition. In mergers and abuse of dominance investigations, the Bureau may therefore consider whether a transaction or conduct has an adverse effect on consumer privacy as a non-price dimension of competition. The Bureau recognizes that anti-competitive effects on consumer privacy may be difficult to express, which raises challenges where the parties put forward an efficiencies defence thereby requiring the Bureau to quantify anti-competitive effects.
The misleading advertising provisions of the Act apply to representations made to the public to promote any business interest, even indirectly. The white paper indicates that the Bureau may review representations that mislead consumers with respect to the type of data collected, the purposes for which the data are collected, and how the data will be used, maintained and erased. The Bureau will assess whether consumers are provided with the information necessary to make informed choices about data collection.
Notably, the Bureau’s mandate in respect of data is broader than that of the Office of the Privacy Commissioner of Canada. The Privacy Commissioner is limited to matters of privacy – which deals only with personal information – whereas the Bureau can examine competition matters related to all types of data.
Big Data and Pricing Algorithms – Criminal Risk
The Bureau is mindful that competition enforcement should not chill innovative and procompetitive uses of big data. While the use of pricing algorithms to monitor and adjust pricing may lead competitors to unilaterally adopt similar or identical pricing or business practices, such “conscious parallelism” behaviour is not criminal under the Act. Although big data may lead to conscious parallelism and thereby soften competition, the Bureau indicates that mere conscious parallelism will not be subject to criminal investigation.
However, the Bureau may review parallel behaviour accompanied by practices that facilitate, or may be an indication of, an agreement between competitors. Such “facilitating practices” may be reviewed under the criminal cartel provisions of the Act, or the civil provisions prohibiting anti-competitive agreements between competitors. The Bureau believes that big data can lead to various activities that could constitute facilitating practices. For example, the white paper indicates that disclosing a pricing algorithm to competitors or disseminating pricing information using a digital platform in a concentrated industry could facilitate anticompetitive agreements. It is therefore important for firms to design and implement algorithms to minimize the risk of a potential criminal investigation.
The Bureau indicates that it will not hesitate to pursue big data related conduct where there is an underlying agreement between competitors to fix prices, allocate markets or restrict output. For example, competitors may agree to adopt the same pricing algorithms to maintain prices, or share inventory data to facilitate an agreement to restrict output.
Big Data in Mergers and Abuse of Dominance Cases
The white paper considers how big data can affect the Bureau’s usual analytical tools and remedies. Even where the Bureau anticipates that there could be challenges, it does not suggest the need for any significant departure from its current approach.
- Market Definition: Big data is frequently used by businesses that offer multi-sided platforms (e.g., a social media network that is free for users, but that charges advertisers). The Bureau recognizes that the interaction between all sides of a platform needs to be considered when defining relevant markets. For certain cases involving big data or platforms in the digital economy, the Bureau may focus on direct evidence of competitive effects, rather than trying to define the relevant market as an initial step.
- Market Power: Assessing market power in big data investigations may raise challenges. First, pricing on all sides of a platform may need to be considered when considering pricing as an indicator of market power. Additionally, due to the rapid pace of change and innovation associated with big data businesses, market shares may not be as indicative as they are for assessing traditional businesses. Access to and control over data may confer market power where such data is an essential input for rival firms to compete, or where network effects are present.
- Purpose and Business Justifications: The Bureau will consider business justifications when a firm is preventing competitors from having access to data that is necessary for competitors to compete. To determine whether such foreclosure is anticompetitive conduct, rather than legitimate competition, the Bureau may use the “no economic sense” test from the TREB case and assess whether profits that are unrelated to the foreseeable anticompetitive effects of the conduct are greater than the costs incurred in pursuing the practice. The white paper only briefly mentions that intellectual property rights will be examined as part of big data investigations, and does not provide any new guidance.
- Competitive Effects: Given that big data is often used in the production of goods and services, mergers and business practices involving big data may involve vertical competition issues. Big data may also create difficulties in prevention cases as this requires an understanding of the future use of the data. Algorithms and bots increase transparency in the market, and the Bureau will consider whether they facilitate coordinated effects. Assessing and predicting competitive effects in a context where dynamic competition is important raises inherent difficulties associated with the measurement and quantification of innovation.
- Efficiencies: The Bureau predicts that dynamic efficiencies (i.e., efficiencies resulting from better product offerings and new production processes) may become more prominent due to the rapidly innovating marketplace.
- Remedies: A unique feature of data is that it can be exploited by multiple users and thereby divested without depriving the seller of its use as well. The Bureau’s preference for structural rather than behavioural remedies remains, but it recognizes that big data could lead to novel remedies. For example, divested data could become stale over time, requiring divesting parties to periodically hand over updated data.
The white paper provides useful initial guidance on how the Bureau may apply the existing framework under the Act to various competition considerations related to big data and algorithms that have been examined so far. However, the big data era is still in its infancy, and the white paper leaves open many questions as to how the Bureau would analyze conduct in light of the big data challenges it identifies. It therefore remains to be seen how those competition issues will develop, and how the Bureau will respond. The white paper is open for comment until November 17, 2017.