The good, the bad (and the sometimes ugly) amendments to the Digital Markets Act: Part 1
In the next weeks, EU lawmakers will have to agree on the changes they want to see in the European Commission’s plans to regulate the digital economy under the Digital Markets Act (DMA).
Some of the proposed amendments, if adopted, could radically transform the flagship services offered by Google, Amazon, Facebook, Apple and Microsoft.
The DMA is complicated, and one can wonder, what are “the good, the bad, and the ugly” ideas lawmakers are proposing? The committees for Industry, Research and Energy (ITRE) and Economic and Monetary Affairs (ECON) have certainly come up with a few in their draft opinions, so let’s have a look. Note, “good” amendments help make sure that the rules can work in practice, and will have their intended effect. Some “bad” amendments could have counterproductive or unpredictable effects when implemented. Meanwhile some “ugly” amendments carry the very real risk of harm to the quality of core platform services in Europe, and the many consumers and business users who enjoy them.
Industry, Research and Energy Committee
Amendment 1 starts the opinion off nicely, adding in a reference in recital 1 to the important role that digital services and core platform services play in helping small and medium sized enterprises (SMEs) and start-ups compete with bigger players (in particular by benefiting from a platform’s efficiencies of scale and scope).
Under Article 16(3) the Commission proposed a relatively low limit, under which a gatekeeper would be considered to engage in “systematic” non-compliance if it breaches the DMA three times within five years. Amendment 78 would lower the ceiling even more, to two non-compliance decisions with no time limit. Given the troublesome lack of regulatory dialogue, or any obligation on the Commission to clarify ambiguities in the text, this opens the door to disproportionate penalties, forcing companies to take an extra-cautious approach in compliance and limiting any balancing of the trade-offs inherent to practical implementation of the DMA’s mandates.
The DMA already includes online intermediation services, online search engines, and operating systems, online social networking, video sharing platforms, number-independent interpersonal communication services and cloud computing. Amendment 5 would add smart TVs and IPTVs, digital voice assistants and platforms that use integrated voice assistant technologies, mobile payment services, web-browsers, video and audio on demand services, digital services which allow the creation of, processing of, accessing or storage of data in digital form, including software as a service, digital lockers and online advertising services. This would be a remarkable expansion of the scope of the regulation to a variety of platform services, and other digital services and even hardware devices which are neither “important gateways for business users” (the logic of the DMA), nor for which there has been any identified competition issues, nor for which there has been any impact assessment on the particular obligations (though to be fair, the impact assessment for the existing core platform services is filled with unsubstantiated assumptions as well).
Economic Affairs Committee
Again, Amendment 1 starts off nicely, changing recital 9 to emphasise the need to stick to the DMA’s legal basis, and avoid national fragmentation. This amendment would add that “the Commission should be able to prevent the adoption of national measures based on stricter national laws that are inconsistent with this Regulation or with a decision adopted by the Commission”. This would help the DMA achieve its stated objectives.
Amendment 110 would change Article 39(2) so that the DMA would start applying three months after it enters into force, instead of the current six. The DMA contains far-reaching and ambiguous obligations in Articles 5 and 6 that will apply to a diversity of core platform services, with a one-size-fits-all approach, and the potential for high fines unrelated to any quantum of harm. This will encourage a very conservative and risk-averse approach to implementation, and carries the potential for overcompliance and the degradation of core platform services. Shortening timelines would only exacerbate this.
Amendment 57 would change Article 6(1)(f) to add a “fair, reasonable and non-discriminatory” (“FRAND”) standard for access conditions. Mandating interoperability to previously closed layers of each core platform services’ technical stack is already pretty extreme (both from a technical perspective, and from an economic incentives perspective). Mandating access be on FRAND conditions is a recipe for endless litigation. Anyone familiar with the enforcement of FRAND in the licensing of standard essential patents, or on the difficulty of specifying the provision of access conditions in the telecom sector would hear the alarm bells ringing. “FRAND access” is a moving target, ambiguous and doesn’t actually provide any legal certainty. What it will certainly do is generate lots of work for lawyers, economists, and lobbyists arguing that whatever is provided, is not FRAND. While the DMA is already being used to resolve disputes between big corporate interests, the focus should be on having a regulation that actually meets its intended objectives, and helps make Europe fit for the digital age.