Does Competition Law Offer a Proper Toolbox to Tackle the Power of Internet Giants and Platforms?

Introduction 

The Internet and platforms operating via the Internet (such as Amazon, eBay, Google, and others) have opened up countless opportunities for manufacturers and retailers, as well as novel and innovative service providers, to promote their particular business. This success goes hand in hand with enhanced consolidation, which is often the result of self-amplifying effects: To be successful, a platform needs to gather a sufficient number of regular users (e.g., a holiday booking platform needs consumers) and of content providers (e.g., of accommodations, flights etc.). The more traction a platform gains on one side (e.g., on the side of the consumers), the more traction it will gain on the other (e.g., on the side of the accommodation providers) and vice versa. 

This phenomenon becomes even further pronounced by the fact that those companies generate massive amounts of data, including highly detailed user preferences and trends. This may allow them to solidify or even further expand their current position. 

Is competition law as it currently stands (here, we are only examining German and EU competition law) indeed the proper means to tackle what feels like a (potential) social-economic misdevelopment? To explore this further, we will first of all cast a glance at the “standard toolbox” competition law is offering to tackle the more “traditional” concerns which are rooted in the pipeline economy. 

Assessing market power in the pipeline economy

In the pipeline economy, companies typically manufactured products, refined products they were sourcing, and/or were active as distributors. The typical and often crucial starting point for a competition law assessment in the pipeline economy was to look at the companies’ power on the supply and/or demand side by assessing their market shares. Where this still applies, it involves a two-step exercise:

(1) It requires a definition of the relevant market. 
(2) It then follows that the establishment of the parties’ market shares should be calculated based on turnover. 

Let’s take the scenario, for example, where a company which is one of three suppliers of headlights for light commercial vehicles (LCV) in Europe intends to acquire another European supplier of headlights for LCV. In this situation, German and EU merger control law demands that we assess (among other things) the parties’ joint market shares.  [1]

1. The parties’ joint market shares will first and foremost depend on whether the market is delineated narrowly or broadly. For example, if the market was defined broadly to cover all lights (rear, front) for all types of vehicles (passenger cars, LCV, and trucks), the parties joint shares will likely be low (assuming that there are various other suppliers of lights for all types of vehicles). However, if the market was defined narrowly to cover headlights for LCV only, the parties’ joint shares would likely be high. The definition of the relevant market in a given case depends on various criteria, in particular on whether customers would (easily) switch to another product (so called demand-side substitutability) if prices for the mentioned product were slightly raised (Small but Significant and Nontransitory Increase in Price, SSNIP-test). This assessment is sometimes complex and hard to predict.
   
   
2. Assuming that the total market value for headlights for vehicles in the EU amounted to € 1 billion in 2017, and the acquirer generated € 0.4 billion with headlights (generated through sales to large OEMs, like VW or Daimler) and the target company € 0.2 billion with the same product, their joint market share would exceed 50%. In a nutshell, this would render it difficult for the parties to receive merger clearances in the EU or Germany. The parties would need to demonstrate, for example, why their customers would still have sufficient alternatives to turn to, why there is no incentive for price increases, that new players could easily enter the market, etc. 
   
   Moreover, (assuming for a moment the merger was lawful) this joint entity (being a market dominant company) would need to ensure that it does not abuse its market power, e.g. by obstructing other companies or by applying dissimilar conditions vis-à-vis its customers. Likewise, if these two companies (instead of merging) were to agree on joint purchasing or joint research and development (R&D) [2]due to their high market shares, competition law would likely deem such an agreement to be invalid and this would expose them to a risk of high fines. 

Of course, market shares are by far not the only criterion for the competition law analysis; they are nevertheless an important indicator. In other words: Low joint market shares (e.g., less than 10% of the entire market) would render it exceedingly difficult for a competition authority to prohibit a merger, an intended cooperation (e.g., R&D), or a distinct unilateral conduct (e.g., the application of dissimilar trading conditions), whereas high joint shares (at least factually) tend to shift the burden of proof to the parties. 

Challenges particular to the platform economy 

Typical Internet “giants”, such as Google, Amazon, or Facebook, but also many somewhat smaller players, work differently. Their business model is focused on creating platforms or entire ecosystems (for purchasing, searching, or socializing), and selling services attached to the platform, or selling access to it. Typically, advertising plays a significant role. To the extent that turnover is generated at all, in contrast to the pipeline economy, this turnover often does not correspond to those sides of the market where, “intuitively”, an “issue of dominance” might arise.

For example, while (arguably) Google is the most important provider of Internet search services in Europe, it does not charge its customers for the provision of these services. Hence, no turnover is generated. Does this mean that Google’s market share is “0” and that therefore all conduct (directly or indirectly) related to the provision of search services is immune to competition law scrutiny? Certainly not! However, it becomes obvious that the more traditional legal and analytical framework, which is rooted in the pipeline economy (see above), no longer works. There is therefore a need for a new or revised toolbox to address the multi-sided and data-rich nature of such business models in a predictable manner. 

Recent legislative changes and current plans for further amendments 

The good news is that legislators and competition authorities are embracing the topic more and more. 

Amendments to German Law

The German legislator introduced amendments to the Act against Restraints of Competition (ARC), which entered into force in June 2017. 

• The legislator inserted new qualitative criteria for the assessment of market power, such as (i) the existence of network and scale effects, typical for platforms, which can lead to market concentration, as well as (ii) access to relevant data, and (iii) the behavior of user groups. While this is certainly a step in the right direction, those criteria leave plenty of room for interpretation and it remains entirely unclear how they should be weighted. This in turn may give rise to significant legal uncertainty.
• The legislator also clarified that a relevant market under competition law may also exist if no money flows between the parties. 
  

Recent study recommends further changes

Moreover, the German Ministry for Economic Affairs just recently published an extensive legal study [3] on whether the ARC requires further amendments to meet the (current and future) demand of the Internet economy and digitalization. The study particularly focused on the current German rules on the abuse of market power by dominant companies. To some extent, those German rules are (even) more rigid than the EU rules which are applicable in parallel. 

• The study critically examined the role of platforms or – more broadly speaking – of so-called “information intermediaries” and found inter alia that “intermediation power” should be anchored as an independent (new) type of power in German competition law. It also suggests the introduction of rules against “tipping”: It recommends the adoption of a new provision, which prohibits platform providers in close oligopolies (i.e. only few market players are left), or platform providers with superior market power to obstruct competition, e.g. via the prevention of multi-homing or switching. While “tipping” can be the result of competition on merit, it should not be further facilitated or induced by certain practices. 
• The study also examined the role of data as a quintessential resource for self-learning algorithms etc. It emphasized that even under current law, companies may not refuse to supply data over which they have exclusive control and which is essential for the petitioner to engage in a certain economic activity. It even contemplates a general data-sharing obligation, something which may, however, be difficult to implement in practice. 
  

Developments at EU level

At EU level, the EU Commission also proposed at number of legislative changes, particularly aimed at creating a fair, transparent, and predictable business environment for smaller businesses and traders when using online platforms. [4]

Dr. Daisy Walzel is a competition law expert and partner located in Cologne, Germany. Daisy advises on all aspects of German and EU competition law, specializing in merger control and competition law compliance. She also represents clients in cartel proceedings and in abuse of dominance cases.