Self-Preferencing in Markets with Vertically Integrated Gatekeeper Platforms

Expert Economists Jorge Padilla, Joe Perkins and Salvatore Piccolo wrote an article for the Journal of Industrial Economics. The article discusses the competitive strategies of ‘gatekeeper’ platforms and analyzes a model in which a gatekeeper device-seller facing potentially saturated demand for its device has the incentive and the ability to exclude from the market third-party suppliers of a service that consumers buy via its devices. 


The competitive strategies of ‘gatekeeper’ platforms are subject to enhanced scrutiny. For instance, Apple and Google are being accused of charging excessive access fees to app providers and privileging their own apps. Some have argued that such allegations make no economic sense when the platform’s business model is to sell devices. In this paper, we build a model in which a gatekeeper device-seller facing potentially saturated demand for its device has the incentive and the ability to exclude from the market third-party suppliers of a service that consumers buy via its devices. Foreclosure is more likely if demand growth for the platform’s devices is slow or negative, and can harm consumers if the device-seller’s services are inferior to those offered by the third parties.


The economic significance of online marketplaces, such as Apple’s App Store and Google Play, has increased over time. Apple’s App Store and Google Play earned gross revenues of around €70 billion in 2019, of which almost €10 billion was in Europe. Access to consumers via such platforms has stimulated rapid innovation; over 2.5 million apps are available on Google Play, and more than 1.8 million on the App Store. This is the bright side of their ‘gatekeeper’ role. Because of their broad and loyal customer bases, Apple’s App Store and Google Play constitute critical distribution channels. App developers distributing through them can reach a large number of users at once. But this also allows platform providers to charge app providers significant listing fees and (ad valorem) commissions. On the App Store and Google Play, these amount to 30% of revenues in the first year, and 15% in subsequent years.

Some app developers have complained against these charges. Others have argued that gatekeeper stores restrict their commercial ability so that they cannot avoid these costs. These complaints have attracted considerable policy and regulatory interest and media attention. For example, the European Commission is investigating whether Apple’s rules for the App Store violate antitrust laws.1 In the U.S., Epic Games filed lawsuits against Apple and Google in August, 2020, alleging that restrictions on possible payment methods for apps violate the Sherman Act and harm consumers.2 The Dutch competition authority carried out a market study into mobile app stores in 2019, and recommended further investigation into either ex ante regulation or greater use of competition law in the sector.3 More broadly, reviews of digital competition have supported a more active approach to regulation and antitrust enforcement, including Crémer et al. [2019], U.S. House Committee on the Judiciary [2020] and Digital Competition Expert Panel [2019].

Complainants argue that some platforms exploit their gatekeeper status to extract excessive rents from app developers and/or to favor their own apps to the detriment of their rivals. Some commentators have dismissed these allegations as illogical, arguing that such conduct would be counterproductive for the platforms themselves, since they benefit from the availability of highly valuable third-party apps in their stores. Our paper investigates the plausibility of this argument, seeking to identify when a gatekeeper platform might have an incentive to abuse its market position. To this end, we build a stylized model showing that the incentive of platforms selling devices to abuse their gatekeeper role relates to the evolution of demand for these devices. We set up a two-period game where a monopolist selling a device (e.g., a smartphone) has the option (in the second period) of restricting access to its users by the competitive suppliers of complementary products (e.g., apps) or, in other words, privileging its own product relative to third-party alternatives.

We find that when the growth of demand for the electronic device is healthy, foreclosure in the complementary market is less likely. It becomes more profitable as demand for devices becomes saturated, and the service offered by the device seller is not too inferior compared to the third-party competitors. In our model, foreclosure occurs at equilibrium as an optimal response of the device seller to a slowing down or a decline of its primary business. Under these conditions, foreclosing rivals from the complementary service market enables the device seller to monetize the user base acquired in the first period. Consumers will lose out from foreclosure if the monopolist’s service is inferior to those provided by third-party developers. Such harm is relatively more pronounced as demand for devices becomes saturated.

We also show that, in addition to being detrimental to consumers, the ability to foreclose harms the device seller’s profits ex ante. The device-seller would like to be able to commit not to foreclose in order to increase prices in the first period. Greater profits from a higher period 1 price outweigh the profits gained from foreclosing the service market in period 2. However, there is a time-inconsistency problem; when it makes its decision in period 2, the device-seller may prefer to foreclose to exploit its captive market. Notably, this time-inconsistency problem is more severe when the device seller has a greater incentive to foreclose in the second period — i.e., when demand for devices becomes saturated — which is also the case in which foreclosure harms consumers more. Hence, policies enabling (or forcing) device sellers to overcome the time-inconsistency problem are relatively more important in industries that feature declining demand dynamics.

Several tech giants, such as Apple and Google, seem to have understood the time-inconsistency problem highlighted in our paper, which they have been unable to avoid. For example, when launching the App Store in 2008, Steve Jobs announced that Apple did not intend to profit from it, and that all the money would be given to the developers (Cohen [2008]). Google’s founders initially promoted their search engine’s impartiality on the basis that it was not subject to the potentially harmful influence of advertising (see, e.g., White [2013]).

As of today, however, Apple makes considerable revenues from the App Store,4 and Google monetises its search engine through paid advertising. Such examples are not universal; for instance, Padilla et al. [2017] discuss how Adobe’s commitment to an open standard for pdf’s enabled it to maintain the attractiveness of the format for third-party developers and end users. But they do provide some evidence of time-inconsistent incentives on platforms, which cannot easily be overcome by unilateral commitments. Explicit regulatory constraints may therefore be required to protect consumers against the risk of hold-up. For instance, Gilbert [2021] argues that a mix of antitrust enforcement and regulation may be required to address concerns about the abuse of market power by dominant platforms. Entry by these platforms into aftermarket businesses could be made less harmful by requiring mandatory quality standards to protect consumers and to compensate them for losses in the event of rivals’ foreclosure.

Read the full article here

This article was originally published by the Journal of Industrial Economics and in PDF format here. We would like to thank the Editor and an anonymous referee for insightful suggestions. We are also indebted to Michele Bisceglia, John Davies and Raffaele Fiocco for their comments. Shiva Shekhar provided excellent research assistance. This research was partly funded by the music streaming service Spotify. The views expressed in this paper are the sole responsibility of the authors and cannot be attributed to Compass Lexecon or any other parties.