Do big platforms hold up innovation? with Doh-Shin Jeon

Doh-Shin Jeon is professor of economics at Toulouse School of Economics and a part-time research fellow at the CEPR.
An expert in industrial organisation, Doh-Shin’s research has a strong focus on the economics of information technology and Internet platforms: “My ongoing research studies issues such as app store commissions, platform liability, user manipulation by platforms etc”.
He is a member of the expert group of the Observatory on online platform economy of the European Comission.

What do we understand by online platforms and the online platform economy?

My answer to this question is based on the COMMISSION DECISION of 26.4.2018 on setting up the group of experts for the Observatory on the Online Platform Economy.

Online platforms include online intermediation services provided through online marketplaces, app stores, and social media, as well as search engines.

The online platform economy covers all economic activity arising out of actual or intended commercial transactions in the internal market and facilitated directly or indirectly by online platforms, in particular online intermediation services and online search engines.

How important do these online intermediaries prove to be for small and medium scale companies in the actual context of digitalization? Will platform-to-business relations shape the future for smaller players in the market?

According to a recent survey conducted in October 2020 for the European Commission, more than a half of enterprises (59%) in the European Union derived more than 25% of their revenues from online platforms compared to 50% of companies in November 2019. 

Online platform sales exceeded 75% of the revenues of 13% of companies in 2020 compared to 10% in 2019. These figures demonstrate the increasing dependency of business users on online platforms, which may have been accelerated by the COVID-19 crisis and lockdown measures in 2020.

A high share of respondents (61% in 2020) consider that the successful operation of their business is very dependent or completely dependent on online platforms.

Are data asymmetry and winner-take-all dynamics inherent to the online platforms’ economy as of now?

Online platforms are characterized by economies of scale, network effects and economies of scope, which create a tendency for market tipping. 

First, most service/content provided by online platforms is information goods, which have zero marginal cost of production and distribution and a (potentially large) fixed cost of investment. 

This means enormous scalability and hence creates a tendency for market tipping. A late comer may find it difficult to match the service quality of an incumbent as this requires a large fixed cost, which can be recovered only if it gains a large market share. 

Second, regarding network effects, one can distinguish direct network effects from indirect ones. Direct network effects arise, for instance, when the larger the number of users of an online social media, the greater the benefit that a user obtains from joining the social media. 

Indirect network effects arise through complementors (such as applications): the larger the number of consumers using a smartphone platform, the larger the number of applications available on the app store of the platform. For this reason, the Windows mobile phone platform failed to establish itself. 

Third, economies of scope exist when operating across multiple adjacent markets induces a platform to lower its cost or increase its quality. Economies of scope arise when platforms can leverage from one market to another its consumer base, its supplier networks, its brand, its technical capability and its infrastructure. 

Economies of scope induce big platforms to build an ecosystem covering multiple adjacent markets.

And how exactly does data tip the market?

Data reinforces each characteristic of platforms mentioned above and thereby strengthens the tendency of market tipping and the power of platform ecosystems. 

First, most data is generated as a byproduct of users’ consumption activities on online platforms. Once data is generated, the marginal cost of (re)producing data is zero. Hence, data reinforces economies of scale. 

Second, the larger the consumer base of a platform, the larger the amount of data the platform obtains. The larger the amount of data, the greater the benefit (either quality increase or cost reduction). Therefore, data strengthens network effects. For instance, in the case of search engines, network effects exist for rare queries. 

Third, a platform with a large ecosystem operating in multiple markets can merge and combine different data sets to create high dimensional data about its consumers, which allows the platform to gain in-depth insights about diverse needs of its consumers. For this reason, data seems to be the strongest source for economies of scope.

In order to restore competition, one should reduce data advantage of incumbent platforms by imposing data sharing obligations (as is required in search markets by the recent Digital Market Act (DMA) proposal), reduce switching costs and facilitate multihoming by imposing data portability/interoperability obligations.

How does the before mentioned impact research and innovation and SME growth?

When platforms determine governance of their ecosystem of innovations, do they have a strong incentive to promote the innovations of business users who depend on their ecosystems? This is an extremely important question but to the best of my knowledge there is yet not much economic research. Hence, I can provide only fragmented answers.

First, we need to be concerned about hold-up by vertically integrated platforms. Vertical integration is a general phenomenon among GAFAM (Google, Apple, Facebook, Amazon, Microsoft): Microsoft’s Windows office suite is composed of copycats of the previously-existing software programs, Amazon sells products that compete with third-party sellers’ products on its own marketplace and Google and Apple sell apps which compete with third-party apps on their app stores. 

Vertically-integrated platforms have strong incentive to hold up innovations of their business users by copying best-performing products.

And how would they hold up those innovations by business users?

This is facilitated by three factors: (i) platforms have access to data about browsing  and sales of third-party products; (ii) they can manipulate their search/recommendation algorithm in order to give prominence to their own products while demoting competing products; (iii) they can give quality advantage to their own services by providing a higher degree of interoperability with their own other services and by reserving certain functionalities uniquely for their own services. 

This in turn can provide them with a rationale for giving prominence to their own products. For instance, Apple Music is embedded in iOS, with the app pre-installed on all iPhones and the only music-streaming service that is accessible through Siri, Apple’s virtual assistant.

And in the case of Google?

In the case of Google, which has a monopoly position in search advertising and very strong market power in ad tech for display advertising, it combines, for targeted advertising, enormous amounts of data from its own services with data from a large number of third-party websites/apps (such as online newspapers) which rely on Google’s ad ecosystem. 

My recent research asks how such a combination of data affects innovations of business users in the context of online newspapers’ investment in journalism quality. We find that Google has a strong incentive to use the data obtained from newspapers to target ads on other sites (including Google-owned websites/apps), which will lower the advertising revenue of newspapers and thereby their investment in journalism quality.

Third, in my ongoing research studying competition between two smartphone platforms (Apple vs. Android),  we find that multihoming of app developers induces the competing platforms to charge too high a commission on their respective app store (which is higher than the commission that would be chosen by a hypothetical monopolist owning both platforms), which leads to too little innovation by app developers. 

According to the market study into mobile app stores by the Netherlands Authority for Consumers and Markets (2019), since all popular and known apps are present in both smartphone platforms, consumers’ initial choice between the two does not depend on the availability of apps. 

Then, our result implies that competition in the smartphone market between iPhone and Android phone does not translate into competition between App Store and Google Play. Our result is consistent with the remark done by the judge Gonzalez Rogers to Apple CEO Tim Cook in the on-going legal battle between Epic and Apple, “ it doesn’t seem to me that you feel under pressure or competition to actually change the manner in which you act to address the concerns of the developers”.