On 15 December 2020, the European Commission (“Commission”) proposed the “Digital Services Package”. This package comprises proposals for two directly effective regulations, namely the (1) Digital Services Act (“DSA”) and (2) the Digital Markets Act (“DMA”). The objective of these regulations is to create a “digital safe space” and “fair online competition” in the EU.
Due to its far-reaching terms, the DMA would likely require changes to the business models of big tech. Mandatory ‘big tech benevolence’ may well benefit “business users”, i.e. people or companies that use platforms to provide goods or services to internet users. Because, as economist Milton Friedman famously stated “there is no such thing as a free lunch,” however, changes in big tech’s business models prompted by the DMA might also affect free internet services.
This blog post presents a preliminary analysis of the DMA’s possible consequences for ‘the internet as we know it,’ i.e., for ‘big tech’, smaller internet-based competitors, and European users. The analysis is based on the DMA as it was proposed by the Commission on 15 December 2020; the scope and content of the DMA could still be amended during the legislative process. In the first part, the main prohibitions, obligations and rights set forth in the DMA, and the revenue impacts thereof for ‘big tech,’ are analyzed. The DMA’s possible consequences for smaller players and European consumers are discussed in part two. In the third part, the concept of ‘data as currency’ is clarified. The final part lays out the conclusions.
1. Strategic Business Responses to the DMA
The proposed DMA sets forth data protection and competition-related obligations for large online platforms acting as “gatekeepers” and grants internet users rights vis-à-vis such gatekeepers. If adopted, the DMA will impose a series of restrictions and affirmative obligations on large internet companies and create significant compliance risks (including fines of up to 10% of global annual turnover).
Given the competitive forces at play, big tech will have incentives to minimize the costs of compliance with the DMA’s requirements, and, maybe more importantly, the loss of revenue associated with the obligations it would impose. In some cases, companies might be tempted to examine ways to “work around” the DMA, but doing so might be risky. Anticipating creative strategies, the DMA prohibits “circumvention” – i.e. “any behavior” by the gatekeeper, be it contractual, commercial, technical or other that “undermines the implementation of the obligations” set forth in the DMA. To that effect, it prohibits, inter alia, that gatekeepers “degrade the conditions or quality of any of the core platform services provided to business users or end users who avail themselves of the rights or choices laid down [in the DMA]”. While this remedy may also be imposed in enforcement of traditional competition law, if the DMA’s terms are construed broadly, many changes in a gatekeeper’s business model could constitute circumvention. If it only means that big tech may not retaliate against business users that invoke their rights, the impact would be more limited.
In other cases, they might want to consider adapting their business models or pioneer new business models. To compensate for the loss of revenue, they could increase advertising or prices for advertising on their platforms (which is also subject to restrictions imposed by the DMA, such as the prohibition to combine data from different services unless the users specifically provide consent), increase prices for other services, such as shipping, or switch to a subscription-based model or some other model. They may also substantially reduce the services they offer to users based in the EU; for instance, they may stop investing in the further development of their platforms, or no longer offer certain services, such as a market place for third party vendors, or cloud storage and email services.
Below, some of the main requirements set forth in the DMA, and how they would result in a decrease in gatekeeper revenues, are discussed.
– Price parity
Gatekeepers will have to allow business users to offer the same products or services to users through third party online intermediation services at prices or conditions that are different from those offered through the platform of the gatekeeper. Accommodation booking platforms such as Booking.com and Expedia will no longer be able to demand “price parity”, which requires that hotels offer the same or a better room price on the platform as they offer on other online intermediary sales channels. As a result, hotels will be able to offer lower prices via other, cheaper channels. The hotels and users have an incentive to do so, because they both benefit from avoiding having to pay higher platform commission fees to, for instance, Booking.com. Consumers would be able to search for hotels on the large platforms, identify the hotel they prefer, and then look for cheaper prices elsewhere. As a result, the large platforms would lose revenues.
– Use of non-public data
Gatekeepers will no longer be allowed to use non-public data from business users they were able to gather through their platform (such as the number of sales made via their platform, the revenue generated via their platform, delivery and shipping data, etc.). Marketplace platforms and retailers, such as Amazon, will no longer be allowed to make marketing decisions based on non-public data from vendors on their platform. This obligation appears to have been inspired by the remedy of ‘firewalls’, which has been imposed by the Commission in the enforcement of traditional competition law. For example, Amazon would not be permitted to use the sales-related data of top sellers on its platform (e.g., data relating to product positioning) to improve its own product offerings. As a result, Amazon would lose the revenues deriving from the value of third party data use.
– Access to data
Gatekeepers will have to provide competitors and/or customers access to the following data:
- Gatekeepers must provide business users, “free of charge”, with “effective, high quality, continuous and real-time access” to and use of the data related to their business.
- Gatekeeper search engines, such as Google, will have to provide third party search engines access to ranking, query click and view data of organic and advertised search results generated by the gatekeeper’s users, on “fair, reasonable and non-discriminatory terms.”
- Gatekeepers will have to provide their advertisers (customers) and publishers (suppliers) with the prices they charge and pay for the publishing of a specific ad and for each of the advertising services they provided. The proposed DMA does not explicitly require that this be free of charge or on fair, reasonable, and non-discriminatory terms.
- Gatekeepers will have to grant their advertisers and publishers access, “free of charge” to the performance measuring tools and information necessary to independently verify the “ad inventory”.
Mandatory access to a gatekeeper’s data will have two direct consequences: (1) the gatekeeper will have to invest in systems to make this data available to these parties, and staff and maintain these systems, which will increase their ongoing compliance cost, and (2) the gatekeeper will lose its competitive edge and its competitive position will deteriorate, as its competitors will have access to the same data, in some cases even without having to bear any cost for such data. Thus, unless gatekeepers are somehow compensated for the costs associated with providing access to the data or the data itself, and for the loss of revenue associated with the exclusive use of the data, the decrease in revenues would cause them to invest less in the affected businesses, and move into other markets. As a result, the market might become less competitive, as fewer projects would be invested in. Further, the DMA would make it harder for any aspiring, successful company to become a gatekeeper, and may itself serve as a barrier to entry. This may explain why ‘big tech’ was not alarmed by the Commission’s proposal, but, by and large, welcomed it.
Only where an affirmative obligation permits it, a gatekeeper may require that a business user pay a fair, reasonable and non-discriminatory price. Where the access service must be provided free of charge, pricing is not an option available to the gatekeeper. In that case, the gatekeeper would have to absorb the associated costs as a cost of doing business, or shut down the pertinent part of its platform (note that this option would not be available if doing so would constitute “circumvention”). If the sum of the costs associated with providing the data exceeds the value of the data to the gatekeeper, the latter would have an incentive to stop collecting the data altogether – the practice would appear to be allowed by the DMA as it imposes only an obligation to “provide access”.
– Circumvention of gatekeeper’s platform
Gatekeepers must allow business users to promote offers to and conclude contracts with end users acquired via the gatekeeper’s service that does not require the use of the gatekeeper’s platform, and they must allow end users to use any content, subscriptions or features purchased outside the gatekeeper’s platform, on the gatekeeper’s platform.
Platforms such as Apple’s App Store and the Google Play Store generate revenue by charging a percentage on the price of in-app purchases by users in the apps downloaded from their app stores. For example, the purchase of a newspaper subscription or a game’s currency. To circumvent the fee charged by Apple or Google, users could be directed to the app developer’s own website (or other third party site) for any purchases. As a result of the DMA, this practice would have to be allowed, since restrictions on the possibility to purchase outside of the app and on the use of content purchased outside the app, would be prohibited. Consequently, Apple and Google might lose revenues from such sales.
– Interoperability with gatekeeper software
Gatekeepers will have to provide business users and providers of ancillary services access to and interoperability with their operating systems, hardware and software features that are used or are available to users in the provision of ancillary services by the gatekeeper.
Providers of ancillary services and business users will have access to the underlying technology and platforms, including the ancillary services, of the gatekeeper. Ancillary services include payment services, fulfilment (i.e. storage, processing and delivery service) and advertising services. For instance, Apple would have to allow other digital wallets besides Apple Pay on iOS devices, and would have to provide Apple Pay to all third party vendors on iOS. The DMA also requires third party access to the platform’s ‘hardware’, which is not defined in the DMA. Hopefully, the intent is not to include non-electronic hardware such as warehouses, trucks, etc. As a consequence, platforms’ revenue from digital wallets’ transactions or delivery fees may decline.
– Fair conditions
Contractual conditions for app developers to access gatekeepers’ app stores will also have to be “fair and non-discriminatory”. The conditions on which gatekeepers grant access to their app stores, such as the Google Play store and the Apple App Store, will be subject to the European Commission’s scrutiny and perception of what is “fair” and “non-discriminatory.” A key issue here may become the amount of commission charged by the app stores. It has been claimed that Apple’s 30% fee is unfair. Whether the Commission will take a position on this issue remains to be seen, but with the DMA the risk that it will impose limits on commission has increased. This results in uncertainty for big tech about the compliance of their business models.
2. The DMA’s Effects on Smaller Competitors and Users
According to the DMA’s Explanatory Memorandum, gatekeepers’ “unfair practices”, as described above, lead to “higher prices, lower quality, as well as less choice and innovation to the detriment of the European consumer.” How the Commission knows that these effects occur, is unclear. The evidence provided in the impact assessment is derived from theories, case studies, and modeling, and selective, opaque, and hard to verify. In addition, the Commission acknowledges the limitations of the methodologies employed; for instance, to calculate consumer benefits resulting from the DMA, researchers relied on “a series of assumptions which may not necessarily be true” and “a hypothesized consumer behavior model, which may not correspond to real choices.”
Without sufficiently robust empirical evidence, it cannot be concluded confidently that the DMA would benefit internet consumers by causing lower prices, higher quality services and products, or more innovative services and products. In a recent position paper, the Spanish competition authority (CNMC) warned that without “knowledge about the theory of damage and efficiencies” an ex ante regulation with a list of prohibited practices “is not advisable,” because “the probability of error is high.” It also pointed out that “given the volume and/or growth of these markets,” the cost might be high. In a similar vein, the Irish government called on the Commission to “justify the need for ex ante intervention” by demonstrating that “innovation is being stifled by so called ‘gatekeeper platforms’ and that digital markets are not contestable due to exclusionary behavior.” The Irish competition authority recalls that many Member States recently blocked attempts to include ‘blacklisted’ practices without merit and evidence of economic harm.” It “remains to be convinced that this evidence base exists now.”
This criticism reflects a broader issue with the DMA’s empirical basis. It is likely that the DMA would advance the interests of business users to the detriment of ‘big tech’. Where a law merely advances the position of inefficient smaller competitors relative to more efficient, larger companies, however, it may become counter-productive and adversely affect the interests of consumers. The question therefore is whether the DMA would advance the competitive position of small players to the detriment of European consumers – the DMA, revives ‘the antitrust paradox’.
In the impact assessment study, the Commission refers to the “waterbed theory”, holding that if law restricts a specific practice in one area, big tech will respond through similar behavior in another area. As examples, the Commission suggests that gatekeepers may increase hardware prices or delivery fees to compensate for the loss of revenue caused by the DMA. Under the waterbed theory, only by regulating the entirety of big tech’s scope of digital business (where insufficient competition occurs) counter-productive adverse effects anywhere in the digital space will be effectively countered. The waterbed effect therefore justifies broad regulation and broad powers for the Commission to intervene. The Commission conceives of the relevant waterbed, however, as being limited to e-commerce, excluding free internet services.
Nowhere in its Explanatory Memorandum does the Commission discuss the tradeoff between the interests of business users and those of individual users. Individual users include consumers and other users; it is entirely conceivable if not plausible, that lower revenues from e-commerce will cause big tech to reduce its free internet services. The Commission’s claim that the consumer surplus post-DMA will increase by 6% (or 13 billion euros) per year, which claim is subject to the qualifications stated in the first paragraph of this section, does not take into account a potential decrease in consumer surplus derived from free internet services. As illustrated below, once the free internet services are included in the analysis, consumer surplus after the DMA may be less than it is right now.
The mechanism by which free internet services may decrease, is straight forward. As big tech faces decreased revenues and increased compliance costs under the DMA, it may decide to adapt its business models. To recuperate the lost revenues and incremental costs caused by the DMA, gatekeepers will have to tap into other or new sources of revenue, or cut costs elsewhere. For instance, big tech could increase advertising space, increase subscription or transaction fees, move to a subscription-based model, or discontinue certain services.
Directly or indirectly, these costs will be borne by European users in the form of reduced free services (either qualitatively or quantitatively), or by consumers in the form of increased prices. The DMA would be an illustration of economist Milton Friedman’s famous statement that ‘there is no such thing as a free lunch.’
3. Data as Currency
The DMA can also be analyzed from the perspective of the economics of information. In the digital economy, data can serve as currency – in an economically real sense, internet users “pay” for services, such as email and video streaming, with their data. Data is valuable to online platforms, as they can use it to improve their advertising services by targeting more likely purchasers of products and services. While it has been argued that big tech may overcharge internet users by collecting large amounts of data from them, the point here is simply that user data has become a valuable asset.
By not permitting self-interested use and requiring free supply of data to business users, the DMA devaluates this currency and thus lowers the price paid by users who use free internet services. Unavoidably, the restrictions on data use, and affirmative obligations to give it to competitors, and to acquire specific user consent in order to combine a user’s data across services (including third party services), make data less valuable to gatekeepers. In other words, the DMA also has a price-regulating effect. As a result, big tech might reduce the level of services provided for free, either in terms of quantity or in terms of quality. This implies that in order to obtain the same level of free internet services as users currently enjoy they may have to pay in real money, not in data.
By devaluing data, the DMA could cause big tech to collect less data, process less data, use less data, and invest less in data-driven ventures. Indeed, a possible effect of the DMA may well be that innovative services that do not yet exist, will never be invented. This invisible effect typically escapes the attention of regulators, because the models do not capture it. Nevertheless, the things that will never come to be, may be part of the costs the DMA imposes on the economy.
As we have seen, the DMA would have implications for the internet as it currently functions. Because big tech could no longer use data of business users to its advantage, and will have to provide various services free of charge (or upon fair and reasonable terms) to business users, it will not only incur additional compliance cost but also lose revenue.
To make up for the loss of revenues due to the DMA, big tech may have to adapt its business models. Because consumers could no longer pay with their data, this avenue would be closed down. The shortfall could be made up through higher prices of transactions on their platforms, higher user fees, reduced free services to consumers and users, or any combination of these. It is conceivable that the level of free services provided by big tech may be the first casualty of the DMA.
All views and opinions expressed in this blog post are solely those of the author.
 Why the EU uses the term ‘act’ is unclear. EU legislation takes the form of regulations, directives or decisions. See Article 288 TFEU.