From: D. Daniel Sokol
To: John Pecman, Commissioner of Competition
Re: Big Data and Canadian Competition Policy
Date: October. 10, 2017
The Canadian Competition Bureau recently released its draft discussion paper on Big Data. The Bureau strikes the right tone in noting that there are both potential benefits and potential drawbacks to Big Data from a competition standpoint. It makes two fundamental points:
- Canadian competition law already has the proper toolset to analyze the competitive impact of Big Data issues; and
- Analysis must be on a case specific basis to match facts to theory to minimize the impact of an incorrect approach.
These points suggest nuance for what is a politically charged issue.
The Bureau begins its analysis by noting the intricacies of Big Data given that the purpose of competition law is not to regulate but to act as a backstop to the market as the primary “regulator” of business behaviour. As a restatement of existing policy, the Bureau observes that “big is not bad” but that anti-competitive conduct through the assertion of market power is the basis for potential intervention. Further, the Bureau explains that Big Data has the potential to change the market in fast and at times unpredictable ways.
Data in its many forms have been traditional issues that have been addressed by the Bureau in cases such as: Google, TMX, Nielsen, and Tele-Direct. What makes Big Data analysis more complex are the four Vs: volume, velocity, variety and value. The paper did not expand on the difficulties of extracting value from Big Data. This is critical because to extract value, it is not the data that matters but what one does with it.
The Bureau correctly notes the complexities that Big Data analysis creates for purposes of market definition and the assessment of market power for two-sided markets. Yet, such assessments are possible, such as the recent Microsoft/Linkedin, Amazon/Wholefoods, and Facebook/WhatsApp mergers, none of which were blocked on competition law grounds. However, the Bureau’s discussion on barriers to entry is incomplete as there is insufficient analysis of customers switching across platforms and low switching costs that exist for many online platforms. Nor is there sufficient discussion that data can get stale, that certain types of data are non-exclusive and non-rivalrous, or that highly differentiated platforms need highly differentiated data.
The discussion on efficiencies that Big Data create correctly acknowledges the improved quality and enhanced innovation that Big Data offers, such as with Google Maps and the Tom Tom/Tele Atlas merger. For this reason, vertical mergers typically do not present competition law challenges.
Big Data and collusion take up a considerable amount of the paper. The Bureau properly acknowledges that Big Data is an issue of degree rather than of kind for hard-core collusion and conscious parallelism. Big Data may expand the challenges of facilitating practices but the Bureau identifies that this must be decided on a case-by-case basis. The remainder of the paper addresses deceptive marketing practices. The paper does not reflect that other Canadian institutions also regulate data protection.
Overall, the report reflects a cautious and well-reasoned approach to understand a complex topic.
D. Daniel Sokol is University of Florida Research Foundation Professor and University Term Professor at the University of Florida Levin College of Law and Senior Of Counsel at Wilson Sonsini Goodrich & Rosati.
To send a comment or leave feedback, email us at blog@cdhowe.org.
