e-Conomics analysed the relation between data, competition and innovation.

Multiple studies for the Ministry of Economic Affairs and Climate Policy

In 2017 and in 2020, the Dutch Ministry of Economic Affairs commissioned Ecorys, e-Conomics, and Radicand Economics to study the relation between data, competition and innovation.

The use of data technologies provides companies with opportunities to improve the quality of their product or service, to offer new information services to consumers, and to personalise products and services. Data analysis may thus benefit us in terms of innovations. However, there are also concerns that the use of data technologies hampers competition. Some argue that the use of data technologies contributes to the rise of dominant positions that are detrimental to consumers and other companies.

Below, we briefly summarise the main findings on the relation between data, competition and innovation.

Data and Competition

The relation between data and competition is complex. On the one hand, employment of data technologies may tip markets into monopolies. So, this implies that there is a negative relation between competition and the use of data. On the other hand, competition in terms of quality (e.g. search engines competing in terms of better search results) drives a strong appetite for data. This implies that there is a positive relation between competition and the use of data. Below we elaborate on these seemingly contradicting relations.

Data and tipping markets

A small lead in terms of data or users can turn (tip) a market into a monopoly. This results from the interlocking and self-reinforcing mechanisms in a platform’s growth engine: network effects, learning effects, and scope economies (these concepts are explained here).

Despite the risk of market tipping, digital markets need not turn into protected monopolies. They are exposed to so-called “moligopolistic competition”. This is a form of dynamic competition that characterises the digital economy. It is a form of competition where a few large companies and many small innovators compete across market boundaries. As a result, these markets remain contestable and dynamic. (Read more about this in “Big Tech and the Digital Economy: The Moligopoly Scenario” by Nicolas Petit)

There is, however, a risk that moligopolists resort to anti-competitive behaviour to protect themselves against these kinds of challenges from other markets, or to leverage their incumbent position to other markets (see also here). Google, for example, has tied Chrome to Android to take over the browser market. It thereby cut off rivals such as Microsoft from observing data on search queries and clicking behaviour, which are essential for developing search engine technologies.

Some have argued that Google should be obliged to share its search data with rival search engines. However, such obligations may neither be effective nor necessary for restoring competition in the search market. Search services are closely integrated with browsers and advertisement services. Given this tight integration, our findings are that data sharing is not effective when anti-competitive tying and bundling strategies remain unaffected. Yet, in the absence of such strategies, it is possibly unnecessary to impose data sharing obligations. Afterall, Microsoft would already have access to relevant search data if Google had not tied its services.

Competition drives an appetite for data

Companies employing data technologies develop a strong appetite for data. Competition can amplify this effect. Providers of advertisement spaces, for example, compete in terms of who provides the most effective ad space. This boils down to who has the most data. So, the more competition Google and Facebook experience on the ads market, the more data they want to harvest in the search and social media market. Likewise, more data stimulates innovation, so that the threat of cross-market competition prompts moligopolists to collect more data.

The relation between data and competition is thus complex, and as a result, so are competition cases involving data. The German competition authority BKA, for example, pursues a case against Facebook. It argues that Facebook is dominant in the market for social media, and that it is exploiting its users by excessively violating their privacy. The BKA considers excessive harvesting of data equivalent to charging excessive monopolistic prices. It regards this an abuse of dominance under competition law.

Undeniably, Facebook gathers extremely large amounts of data. But the comparison with charging excessive monopolistic prices is unsuccessful according to us. The question is whether smaller parties on that market (like Snapchat or TikTok) are behaving differently? Or do they exhibit similar excessive behaviour under pressure of competing forces on the advertisement market? If so, the problem is not one of too little competition in the social media market. It is rather a result of strong competition in the ad space market. The latter implies that competition law is not the right tool for addressing the identified problem of excessive data harvesting.

Data and innovation

Data technology can be employed to orchestrate interactions within the value chain. This allows stakeholders to optimise service delivery and production processes, bypass market power, or increase control. Data may thus contribute to innovations that improve operational efficiency which in turn may contribute to effective competition. Data sharing within (and also across) value chains is essential for unlocking this innovation potential. However, stakeholders may be reluctant to share data with others for fear of losing control over their business. This notably plays a role when data sharing involves big (tech) companies. In other words, there can be a hold-up problem involving different stakeholders. Such hold-up problem calls for contractual certainty. More generally, it calls for trusted data sharing environments.

Trusted data sharing environments

Trusted data sharing environments have rules ensuring transparency and non-discrimination. These governance systems let parties benefit from data sharing without negative consequences. They empower data originators to determine who gets data access. With data originators, we mean individual data subjects or companies to whom the data can be traced back.

Various sectors, such as logistics, healthcare, and in agriculture and horticulture, display examples of data sharing environments. John Deere (the tractor company) and Honeywell (known for its thermostats) are, for example, developing innovation platforms based on data sharing. They follow the examples of tech companies like Facebook, Apple and Google that have strengthened their competitive positions by building ecosystems that facilitate vertical data sharing. Data sharing spurs an exponential growth of correlations to be analysed within an ecosystem. Consequently, it spurs (exponential) growth of innovative apps within the ecosystem. A data sharing platform is thus like any other platform in the sense that it exploits network effects.

Data sharing obligations

Note that tech companies operating data sharing ecosystems may themselves be among the pool of app developers. In that case, they may be(come) reluctant to share data when one of those other app developers becomes a competitive threat. For example, Strava (a platform for sports fanatics) shut down data access for the app Relive, because Relive began offering functionalities that rivalled some functionalities of Strava. As argued before (see ‘digital gatekeepers‘), such behaviour is suspicious from a competition law perspective. In such case, competition law can form the legal basis to impose a data sharing obligation.

Furthermore, parties may owe their competitive position to exclusive access to unique data (e.g. banks, electricity companies). Such parties may be less willing to vertically share data because it may attract competitive threats from outside the industry. Consider the example of Google and Apple entering the banking sector. A reluctance to share may inhibit competition and innovation throughout the entire value chain. In that case, a sector specific obligation to vertically share data may be necessary.

A vision on data sharing

The Dutch vision on data sharing calls for more sharing within and across various sectors. In principle the government prefers data voluntary data sharing. But, when necessary, it can be oblige data sharing. To implement this vision, we suggests some general rules on data sharing governance, alongside possible sector obligations and enforcement of competition law.

Rules on data sharing governance

These general governance rules help to create trusted data sharing environments by addressing the above holdup problem. This stimulates the data sharing within and between sectors.

A first rule is that data originators can control access to data that can be traced back to them. The GDPR already regulates this for individuals. For companies, we could introduce similar regulation. This makes companies less reluctant to share business data.

A second rule is that data is shared on a the basis of non-discrimination and reciprocity. The non-discrimination principle prevents self- preferencing by vertically integrated platforms. The reciprocity principle, as in ‘you get access to (all) my data if I get access to (all) your data’, equalises and maximises the innovation potential for all participants of cross-sectoral data sharing.

Reciprocity

Reciprocity is particularly important as it generates network effects. Afterall, the innovation potential of all participants grows with every additional participant. As the number of participants grows, it thus becomes increasingly attractive for other players (from various sectors) to join the data sharing agreement as well. Moreover, reciprocity has some advantages over paying for access to data. First, a firm’s innovation potential does not grow if it gets paid for data access. Second, financial compensations do not result in network effects.

These general rules may be complemented with sector specific obligations aimed at specific parties owing their competitive position to exclusive access to unique data, such as banks (see above). The general rules on data sharing governance would naturally apply. The reciprocity principle enshrined in these rules prevents outsiders from hijacking the sector by entering the business with a huge data advantage obtained in other sectors. It ensures that that all stakeholders have equal chances to innovate.


Impact

Based on our reports, the Ministry of Economic Affairs has been able to develop a well balanced vision on the role of data in the digital economy. The ongoing European data policy discussions reflect the Dutch vision on data sharing.

Client

Ministry of Economic Affairs and Climate Policy

Output

Accessible reports explaining the relation between data, competition and innovation in the digital economy. The reports suggest a number of policy options for stimulating data sharing and safeguarding competition.

Impact

Based on our reports, the Ministry of Economic Affairs has been able to develop a well balanced vision on the role of data in the digital economy. The Dutch vision on data sharing will be reflected in the ongoing European data policy discussions.

Documents
  • The 2017 report can be found here
  • The 2020 report can be found here
Team
  • Nicolai van Gorp
  • Paul de Bijl
  • Inge Graef
  • Gabor Molnar
  • Roel Peters