French and German competition authorities publish joint study on Data and Competition

The meaning of data and Big Data

The paper first focuses on the meaning of data and “Big Data”, which are usually subject to various definitions. The two competition authorities therefore intend to categorize the different types of existing data. Classification could be made according to:

• the type of data (information on individuals, entities, etc)
• whether the data are structured or unstructured, or
• ways of collecting data (data provided by customers on a voluntary basis, data collected by observing the user’s behaviour, data generated by inferring new information with existing data).

The role of data in the economy

The paper then points out the positive role of data in the economy, which enable for companies to improve their products and services, to exploit new business opportunities, to better target existing and potential customers.

The role of data in competition analysis

According to French and German competition authorities, collection and exploitation of data may raise different competition concerns.

Firstly, the paper highlights that collecting and using massive volumes of data may be considered as a barrier to entry if competitors are not able to do so.

The paper further mentions the risk of “snowball effect”, stating that the collection and exploitation of data by leading companies may reinforce their market power, enabling them to provide better services, and attracting de facto more customers, and therefore simultaneously marginalize smaller competitors.

Secondly, the paper stresses that, thanks to the development of digital economy, the customer will benefit from greater transparency, and will be able to compare prices and goods characteristics easily. However, the paper warns about the well-known issue that market transparency may also facilitate collusion and limit competition.

Indeed, companies holding data may be tempted to adopt exclusionary practices, such as refusing data access to competitors, or granting limited access to competitors. Companies could also exclude competitors by concluding exclusive agreements, impose tied sales, etc. Finally, since data make it possible for companies to better know the customers’ preferences, it could also be used to settle price discrimination between clients.

Finally, the paper raises the concern of reduction of privacy protection resulting from the collection and the use of personal data. Even if it does not fall within the scope of competition, it could nevertheless be taken into account as an element of context for the competitive analysis.

The assessment of Big Data and market power

The key part is developed in Section IV of the paper, and aims at assessing Big Data and market power.

The paper focuses on the specificity of the online industry, which is the industry that collects the most data. According to French and German competition authorities, the analysis must be undertaken under a case-by-case basis, and take account for different features such as:

• the relevant market, which may be difficult to determine since the online market are usually multi-sided (one company is active on different customer’s markets)
• the network effects, which refers to “how the use of a good or service by a user impacts the value of that product to other users” and which could either be pro or anticompetitive
• the existence of multi-homing, consisting for a customer to use several providers to get the same service, or
• the market dynamics, and the possibility for new entrants to enter the market and develop their activities.

The paper then provides an overview of the previous cases that considered data as an advantage, either in “non-digital” or “digital” markets.

The paper also raises the issue of the scarcity of data, highlighting that the scarcer the data are, the greater the competitive advantage will be for the companies owning data. The paper points out the two elements that could contribute to increasing the availability of data to the greatest number of operators, and tackle the risk of anticompetitive effects resulting from unequal access to data: 

• In principle, data are “non rivalrous”, which means that access to data is not reserved to one specific company but can rather be accessed by various companies, and should contribute to limiting the data price increase. However, access to data can be limited (significant investments to build a platform, to provide a secured storage, etc).
• In principle, data are easily available, and owing data would therefore not be likely to increase one company’s market power. However, not only the availability but also the variety of data has increased in the past few years, and an assessment of the substitutability of the different types of data should be included in a competition analysis.

Finally, the paper insists on the importance of analysing on a case-by-case basis whether or not data must be collected on a large scale and/or scope to procure a significant competitive advantage.

This paper is a step forward in the analysis of Data and competition issues. This paper will probably provide key tools to the French Competition Authority in its sector inquiry that it is about to launch on data, commercial strategy and competition, and to the Bundeskartellamt in its already launched inquiry against Facebook.