Response from Fideres Partners LLP to the consultation on the revised version of Abuse of Dominance - Enforcement Guidelines

Fideres Partners LLP
11 Old Jewry
London
EC2R 8DU
Tel. +44 20 3397 5161
Fax. +44 20 3397 5170

Monopolistic Practices Directorate
Competition Bureau
50 Victoria Street
Gatineau, Quebec
K1A 0C9

London, 14 May 2018

Dear Madam or Sir,

Table of Contents

Re: Response to the Competition Bureau, Draft Abuse of Dominance Enforcement Guidelines

Fideres welcomes the opportunity to provide the Competition Bureau (“the Bureau”) with our views on their draft Abuse of Dominance Enforcement Guidelines. Fideres is an economic consultancy committed to maintaining fair and transparent markets, and so appreciates the chance to contribute to the development of Canada’s competition enforcement system.

In general, the draft guidelines provide a clear, non-exhaustive list of the range of exclusionary practices available to dominant firms. They also give good examples of empirical tests available to test for market abuse (for example, pricing below ‘avoidable cost’). Fideres also shares the Bureau’s concern to avoid over-enforcement - the need to prevent abusive practices should not disincentivize firms from attempting to achieve a dominant position, which can provide a strong R&D and product improvement incentive.Footnote 1

However, we believe that the draft guidelines could be improved in several important respects. In particular, we recommend that the Bureau:

  • Think more carefully about the newly-achieved market dominance of large tech companies, and what this means for enforcement of the Competition Act;
  • Reconsider the existing statutory limits on fines for antitrust infringements.

We elaborate on both of these points below.

1   Networking Effects, Tech Companies and Abuse of Dominance
(Responds to Section 2 of the Draft Guidelines: “Anti-competitive Acts”)

The draft guidelines provide a non-exhaustive list of possible practices that the Bureau may classify as predatory, exclusionary or disciplinary conduct by a dominant firm. Examples given of exclusionary and disciplinary conduct include margin squeezing, exclusive dealing arrangements, high switching costs and tying/bundling. However, as many commentators have noted, traditional ideas of antitrust violations may need amending as the influence of large technology companies grow.Footnote 2 We note that the draft guidelines contain a somewhat limited discussion of ‘network effects’ and their implications for competition policy in a digital world.

A ‘direct’ network effect means that the success of a platform depends for one user on how many other users use the platform. For example, the utility of using Skype for one user depends on how many others they can connect to.

Figure 1

  • Figure 1 : Direct Network Effects

    User 1's utility from online platform increases the more other users use it.

‘Indirect’ network effects, in contrast, arise when more users on one side of a platform attract more users to the other side. For example, potential buyers using Amazon do not benefit directly from an increase in other buyers but benefit indirectly if the increase in buyers attracts more sellers.Footnote 3

Figure 2

  • Figure 2 : Indirect Network Effects

    More sellers attract more buyers and more buyers attracts more sellers from online platform.

The existence of powerful market players exploiting network effects should not, in and of itself, concern regulators. In fact, large platforms connecting millions of users online have important efficiency justifications.Footnote 4 However, powerful technology companies exploiting their control over internet networking platforms creates new opportunities for market abuse that need to be addressed by competition authorities.

A well-known, recent example concerns Amazon’s dispute with French publishing house Hachette in 2014. After disagreements with Hachette over e-book pricing, Amazon progressively made it more difficult for consumers to buy Hachette books on their platform.Footnote 5 Although Amazon did not ban Hachette books outright, they increased prices and delivery times for them, and took other steps (such as selective advertising) to direct consumers to other publishers.Footnote 6 This is an example of ‘disciplinary’ conduct. Yet it is does not neatly fit into any of the ‘traditional’ categories of exclusionary or disciplinary conduct discussed in the draft guidelines.

Another relevant example concerns search engine giant Google. In June 2017, the European Commission (EC) fined the company €2.4 billion for privileging its own online shopping platform in search results. Specifically, consumers searching for online shopping platforms would see Google’s platform at the top of the search results. This amounted to a direct attempt to exploit the power of networking effects: the commercial success of Google’s platform was strongly dependent on how many consumers visited the page, which in turn influenced how many sellers would sell their products there. The EC found that Google’s platform had no discernible competitive advantage over its competitors, and that privileging it in search results amounted to abuse of dominance.Footnote 7 Like the Amazon case, this does not neatly fit into any categories of abuse of dominance discussed in the Bureau’s draft guidelines.

The Google and Amazon examples illustrate only some of the ways that large technology companies can abuse their dominance of marketplaces on the internet.

In testing whether such practices amount to abuse of dominance, both qualitative and quantitative methods are necessary. In cases of disciplinary conduct by an online marketplace like Amazon, the Bureau can use resale price data to test whether the marketplace is deliberately charging higher resale prices for products supplied by a company it intends to discipline. Statistical tests, such as structural break analyses, can be used to assess whether these prices change significantly when the dominant firm is in a dispute with its suppliers. Other cases will require exploiting qualitative research on how consumers engage with the internet. For example, the EC’s Google ruling relied on a study showing that results on the first page of an internet search receive 95% of all clicks from internet users, as well as a range of consumer surveys.

The box below summarizes our recommendations:

Recommendation 1 for Draft Enforcement Guidelines

  • Add a comprehensive discussion of the importance of network effects in abuse of dominance claims;
  • Add a summary of qualitative and quantitative methods that can be used to test for abusive conduct by tech companies.

2   Fines as Deterrents for Abusive Firms
(Responds to Section 4 (B)(ii) of Draft Guidelines, “Administrative Monetary Penalties”)

The purpose of an administrative monetary penalty (AMP) in Canadian competition law is to “promote compliance with the [Competition] Act, not to punish the respondent for anticompetitive conduct”.Footnote 8 Leaving aside debates on the merits of punitive fines,Footnote 9 an AMP should, at the very least, deter dominant firms from abusing their market position. However, the Competition Act only allows for a maximum fine of $10 million, or $15 million for each subsequent order.

Fideres believes that such maximum fines do not offer adequate deterrence for large companies seeking to abuse their dominant position. Consider the table below, which outlines Canada’s largest companies by market capitalization as of January 2018. For illustrative purposes, we include what percentage of their total market capitalization a $10 million fine would constitute.

Table 1: Largest Companies in Canada by Market Capitalization
CompanyMarket Capitalisation ($ billions)% of Total Market Cap.
($10 Million Fine)
Royal Bank of Canada1220.008%
Toronto-Dominion Bank1110.009%
Bank of Nova Scotia800.013%
Enbridge Inc.690.014%
Canadian National Railway Company630.016%
Suncor Energy Inc.620.016%
Bank of Montreal530.019%
Canadian Natural Resources Limited450.022%
Canadian Imperial Bank of Commerce440.023%
TransCanada Corporation430.023%
Source: Statista, Bloomberg

If, instead, the $10 million fine were assessed relative to total sales affected by the abusive conduct (an approach followed in other jurisdictions - see below), it would likely constitute an even smaller percentage than those listed above. Clearly, a maximum fine of $10 million (or $15 million for each subsequent order) is not an adequate deterrent for companies whose market capitalization numbers in the tens or hundreds of billions of dollars.

The Bureau should consider adopting a criterion of proportionality in its enforcement guidelines, by setting the maximum fine as a percentage of sales affected by the infringement. In Europe, fines for guilty companies can be as large as 30% of the sales affected by the competition infringement (but may not exceed 10% of the overall annual turnover of the company).Footnote 10 Other regulators around the world follow a similar approach:Footnote 11

Table 2: Maximum Fines for Abuse of Dominance in other Jurisdictions
Jurisdiction Max. Fine for Abuse of Dominance Infringements
Brazil 0.1-20% of affected sales
European Economic Area 30% of affected sales; max. 10% of annual turnover
Japan 6%-10% of affected sales
South Korea 3% of affected sales
Source: OECD, Mondaq, Thomson Reuters

This approach allows for larger companies to incur much higher penalties than the absolute limits prescribed in Canada. In the Google case referenced above, the EC applied this method to arrive at an overall fine of €2.42 billion. The EC and Korea’s Fair Trade Commission also recently fined technology company Qualcomm €997m and W1.3 trillion (approx. €700m) for abusing its dominant position in the modem chip market.Footnote 12 This approach allows competition enforcement to be tailored to the size of the guilty company, which current Canadian law does not.

Alternatively, the Bureau may wish to consider setting fines equal to the full extent of the ‘harm’ generated bythe abusive conduct. This requires modelling how the affected marketwould have evolved in the absence of abusive conduct by a dominant firm. For example, if a dominant firm engages in predatory pricing to force out existing competitors, data on competitors’ past profits and revenues before the start of the infringement can be used to model what profits these competitors would have achieved in a fair market.Footnote 13 This is known as a ‘comparator-based’ approach. Several other statistical techniques and theoretical approaches (such as Bertrand and Cournot models) can be used to model how markets evolve in the absence of abusive conduct by a dominant firm.Footnote 14

The box below summarizes our recommendations:

Recommendation 2 for Draft Enforcement Guidelines

  • Scrapping absolute limits on fines for anticompetitive infringements;
  • Setting fines either as a percentage of sales affected bythe anticompetitive infringement, or to the full extent of the ‘harm’ generated by the abusive conduct.

Yours faithfully,


Alberto Thomas, Partner
Fideres Partners LLP

About Fideres

Fideres is an economic consultancy focused on antitrust and competition work, with offices in North America and Europe. Fideres has extensive experience in identifying and analyzing anti-competitive behavior including economic analysis of price fixing, monopolization and abuse of dominance claims, as well as associated damages modelling.

Over the past 9 years, Fideres has been at the forefront of antitrust investigation in prominent cartel cases including major financial markets cases: LIBOR, Foreign Exchange, ISDAfix as well as cartels in the healthcare, shipping, food and technology sectors. To date, Fideres has acted as economic experts in private competition civil claims with aggregate damages in excess of $30bn.

Fideres is also assisting a number of regulators worldwide in their investigations and has given evidence before the U.K. Treasury Select Committee as part of their ongoing investigations into these areas and in particular with respect to the gold market.

For more information, please visit www.fideres.com