A review of amendments to the DMA by Parliament and the Council
On 15 December 2020, the European Commission published a draft Digital Markets Act (‘DMA’), proposing new rules for the largest online platforms. A year later, on 15 December 2021, the European Parliament (‘Parliament’) adopted the final amendments that it will take into the trilogue negotiations. In parallel, the European Council (‘the Council’) reviewed the DMA and adopted a ‘General Approach’ text on 26 November 2021.
While many of the amendments provide more protection for business users, they often overlook the negative impact on end-users. In light of this, the Computer & Communications Industry Association (‘CCIA’) asked Oxera to review the three texts, focusing on the unintended consequences for consumers that could result from proposed changes to: (i) data separation; (ii) product and services integration; (iii) interoperability and interconnection; and (iv) business model choice.
We found some amendments by Parliament and the Council that improve the balance of the proposed obligations. For example, amendments to Articles 5(a), 6.1(c) and 6.1(f) would add exemptions to the obligations around data processing and interoperability, allowing platforms to continue protecting platform integrity, end-user data protection, and cyber security. This would benefit consumers by enabling innovative safety and security features such as Microsoft’s Defender suite, which combines data from across its ecosystem to isolate threats.
At the same time, a number of amendments introduced by Parliament risk worsened outcomes for consumers over the long term.
First, amendments to recital 46 make an unrealistic demand of platforms by requiring that the ‘less personalised’ alternative is of the same quality as the personalised service, resulting in all users receiving the ‘less personalised alternative’. This will degrade the consumer experience if, for example, Amazon cannot personalise its service based on a user’s search history, related offers, and product reviews from across its ecosystem—even if that user has agreed to their data being combined.
Second, amendments to the scope of Article 6.1(d) and restrictions on product integrations in recital 48 would hamper product improvement and reliability while worsening the ‘out-of-the-box’ experience for users. For example, Google Search draws on data and functionalities from other Google services (such as Maps or News) to offer users an improved interface and more accurate results.
Third, extending the scope of Articles 6.1(f) and 6.1(c) to include access to more platform functions for third parties would increase governance risks
by preventing platforms from balancing the degree of interoperability and access granted to third parties in ways that optimise security, quality and trust. For example, Apple’s restrictions on the use of Mobile Device Management tools in consumer-facing apps helps preserve user privacy while still offering contestability through APIs.
Fourth, amendments to recital 49, requiring all services to be commercially viable on a standalone basis, would undermine cross-subsidisation by platforms and jeopardise ad-funded businesses. For example, Facebook’s ad-funded business model cross-subsidises the introduction of new functionalities on the consumer side. Uncertainty over whether changes to services will be considered improvements to a core platform or new services that must be viable on a standalone basis could inhibit innovation and the deployment of new features.
Fifth, mandating free interoperability in Article 6.1(f) would reduce innovation incentives, meaning fewer features and functionalities for consumers. Innovators must be permitted to share in the value of their innovations if they are to have an incentive to incur the costs and risks required to develop them. Undermining this will ultimately reduce choices of products and services for consumers in the future.
Finally, light amendments to Articles 5(c) and 5(e) maintain the obligation to allow business users to steer end-users off-platform, which puts the viability of commission-based business models at risk. If this is replaced with a less efficient alternative, consumers could end up worse off.
Applying all of these restrictions to every gatekeeper platform as a one-size-fits-all solution would undermine many platform business models, forcing changes that could lead to increased prices for both businesses and consumers.
We recommend that throughout the trilogue process policymakers seek out the most flexible approach possible, maximising the scope for enforcers to make a holistic assessment of the market context in which obligations apply, and to assess the impact of those obligations on consumers and businesses.
Amendments that increase the possibility for regulatory dialogue will also help reduce legal uncertainty and minimise the risk of future litigation. At the same time, by recognising that some obligations must be tailored to a platform’s specific economic context, the DMA can avoid the unintended consequences that would worsen outcomes for consumers.