'Big Tech Acquisitions — Towards Empirical Evidence' by Pauline Affeldt and Reinhold Kesler in (2021) 12(6) Journal of European Competition Law & Practice 471–478 comments
Big Tech, commonly associated with the firms Google, Apple, Facebook, Amazon, and Microsoft (GAFAM), makes up the most valuable companies worldwide in 2020. In the ten years leading up to 2020, these five companies alone acquired more than 400 firms, predominantly in the technological sector. However, most of these transactions were not scrutinised by competition authorities as they did not reach the traditional turnover thresholds, whereas those reviewed were not blocked following current merger control procedures. Prominent examples include the Google/YouTube, Facebook/Instagram, Facebook/WhatsApp, and Microsoft/GitHub mergers.
As a result, a number of policy reports voice their concerns about the competitive effects of such acquisitions that target potential competitors but fly under the radar because of the features and challenges of the digital economy. In particular, firms in the digital economy often start to monetise only once they have acquired a large user base, thus not meeting current turnover thresholds for merger investigation. Furthermore, digital industries are typically characterised by multi-sidedness, (indirect) network effects, access to data raising privacy issues, and often zero prices on one side of the market (typically the user side). Competition is then often about non-price outcomes, such as quality of service, data collection, and innovation. Consequently, some of these reports conclude that merger control enforcement needs to be updated to properly account for these particular features. Germany, as one example, already considers the transaction value of the acquisition and in 2021 gave the competition authority power to intervene and prohibit abusive practices when a company has a paramount significance for competition across markets. Other authorities, like the Federal Trade Commission in the United States, announced that they will review all acquisitions made by Big Tech in the past, irrespective of their size.
Besides the discussion about the necessity to update current merger control, authorities are also starting to implement ex ante regulation of dominant digital platforms to complement antitrust intervention and restore competition in digital markets. In late 2020, the European Commission presented its proposal for the Digital Markets Act containing behavioural obligations for large online platforms (so-called ‘gatekeepers’) aimed at reducing entry barriers and ensuring fairness in the relationship between the platform and its different user groups. The United Kingdom recently announced that it will set up a Digital Markets Unit within the Competition and Markets Authority (CMA) to enforce a new code of conduct applicable to platforms with considerable market power (digital businesses with so-called strategic market status), forcing platforms to be more transparent about the services they provide and how they use consumers’ data. The proposal also requires firms with strategic market status to report all acquisitions to the CMA.
A particular case in point for these acquisitions is the market for mobile applications. In 2020, consumers spent 110 billion US dollars on the two major app platforms, whereas worldwide mobile advertising revenue amounted to close to 200 billion US dollars. Examples include the large takeovers of WhatsApp and Instagram by Facebook. While the former acquisition led to Facebook being fined by the European Commission for combining user data, the latter has been in the spotlight of the Big Tech hearing in the US House Judiciary Committee suggesting that one intention of the acquisition was to eliminate a potential competitor. Other prominent examples involve acquisitions of popular apps that were discontinued in the aftermath, such as Microsoft shutting down Wunderlist. However, the majority of acquisitions involve apps that are small and rather unknown. All of these acquisitions take place in a market that is, in principle, characterised by a competitive and dynamic environment with many apps and developers being active. This suggests that acquisitions will not only have an impact on the acquirer and the acquired firm, as often suggested by the current debate, but also on competitors.
Because of these features that are representative for many digital markets, we study big tech acquisitions in the Google Play Store. Specifically, we look at the acquirer’s strategy and the development of the acquired app in terms of prices, data collection, and innovation.
Based on comprehensive lists of all GAFAM acquisitions from 2015 to 2019, we identify more than 50 app acquisitions in the Google Play Store. We then match these with a comprehensive dataset covering almost all apps in the Google Play Store. This allows us to not only observe the acquired apps along with GAFAM as a developer, but also many competing apps. Relevant outcomes include app and in-app prices, updating behaviour, and requested privacy-sensitive permissions. To the best of our knowledge, we are the first to empirically study the effects of big tech acquisitions based on a product-level dataset in a very important online market.
We find that half of the acquired apps are discontinued, which tend to be smaller, less frequently updated, and less privacy-intrusive than acquired apps that are continued. Following the acquisition by GAFAM, the monetization strategy seems to change as apps become free of charge but request more privacy-sensitive permissions. Compared with the whole Play Store, GAFAM seems to target more attractive apps, e.g., with respect to updating, data collection, and demand.
In discussing the literature on big tech acquisitions the authors state
The literature on the competitive effects of big tech mergers or acquisitions is predominantly theoretical. It is often motivated by the example of incumbent high-tech companies buying up start-ups feared to be emerging or potential competitors. Typical features of digital industries include (direct and indirect) network effects, multi-sidedness, free provision of a service to one side of the market (while typically the advertising side pays), and the importance of data. On the side of the market that does not pay for the product or service, competition is then often about non-price outcomes, such as the quality of service, data collection, and innovation. Accordingly, research on big tech mergers or acquisitions mainly focuses on innovation rather than price effects, while data are not traditionally at the core of most of the analyses.
For these digital industries, Motta and Peitz (2020)7 provide an overview of competitive effects of such acquisitions along with several theories of harm resulting in a call for stricter merger control. The theoretical model, in which an incumbent can acquire a potential competitor, highlights that the competitive effects of the acquisition depend on the likely counterfactual: if the start-up has the ability to pursue its project absent the merger, the acquisition is always anti-competitive. The acquisition can only be pro-competitive if the potential competitor is unable to pursue the project absent the merger and if the incumbent has an incentive to develop the project following the acquisition. Instead, Cabral (2020) provides a more cautious note, highlighting the importance of technology transfer through acquisitions and the discouragement of entrants’ innovation incentives due to strict merger policies. In particular, Cabral (2020) argues that digital industries are characterised by high uncertainty about where the next competitive threat comes from, which lowers the preemption motive for acquisitions. Rather, due to poorly working markets for technology transfer, acquisitions are a means for incumbents to appropriate complementary technology. If this technology is worth more in the hand of the incumbent, the higher acquisition price generates innovation incentives for entrants in the first place.
This relates more broadly to studies on incentives to innovate for both the incumbent and start-up with an intervening competition authority. Letina, Schmutzler, and Seibel (2020)9 show that a prohibition of acquisitions leads to reduced innovation efforts and, as a result, the authors argue that acquisitions should rather be challenged in industries where innovation effects are considered small to justify enhancing competition. In contrast, Fumagalli, Motta, and Tarantino (2020) show that merger policy does not need to be lenient towards all acquisitions of potential competitors. The beneficial effects of acquisitions on innovation can instead be reached by policies pushing incumbents to early acquisitions of financially constrained start-ups. Kamepalli, Rajan, and Zingales (2020) show that acquisitions by the incumbent lower payoff prospects of new entrants and thus discourage them from investing (‘kill zones’). Bryan and Hovenkamp (2020) study acquisitions of start-ups in a model with two incumbents, one leader and one laggard, where the start-up does not have the ability to enter the market. Absent limits on start-up acquisition, the leader will always acquire the start-up to prevent the laggard from catching up technologically. This also implies that start-ups bias their R&D investment towards improving the leader’s technology rather than towards technology helping the laggard to catch up. They propose antitrust intervention in the form of compulsory licencing to laggards in cases of start-up acquisitions by dominant incumbents. Differently from the previously mentioned papers, Katz (2020) models competition for the market rather than in the market. In his infinite horizon entry model, the incumbent and entrant compete for the market for one period, then one of them exits the market while the other reaps monopoly profits until the next entry event. He argues for higher antitrust scrutiny of incumbents’ acquisitions of emerging or potential competitors when competition is for the market; however whether the prohibition of mergers increases or decreases innovation incentives of entrants depends on the characteristics of the concerned market.
Cunningham, Ederer, and Ma (forthcoming) show empirically for the pharmaceutical industry that incumbents acquire (possible) entrants to discontinue the target’s innovation, thus pre-empting potential competition, thereby coining the term ‘killer acquisitions’. However, the relevance of killer acquisitions for digital markets is a priori not clear, as these markets are often characterised by an abundance of products (translating to many entrants) paired with an unpredictability of success that considerably weakens the pre-emption motive.
The only empirical papers studying acquisitions of GAFAM comprise Gautier and Lamesch (2020) and Koski, Kässi, and Braesemann (2020). Gautier and Lamesch (2020) look into acquisition strategies, finding acquisitions between 2015 and 2017 by GAFAM to take place in the firms’ core segments and mostly accompanied by shutdowns. Koski, Kässi, and Braesemann (2020) find entry rates and venture capital funding to be reduced in the target’s product market following big tech (GAFAM and IBM) acquisitions for the period from 2003 to 2018.
To the best of our knowledge, we are the first to empirically study GAFAM acquisitions in the market for mobile applications based on product-level data. The market for mobile applications is a prime example of a relevant online market, as it entails a dynamic environment with many active apps, in which innovation and privacy considerations are more important parameters of competition than price. It is also a market in which different types of acquisitions take place and spillovers can be studied
