Trackers on our websites and apps enable multi-sourced data gathering. While numerous operators engage in tracking, a small number of data giants controls the majority of these trackers. This article, available here, considers the rise and growth of this industry, the power it has bestowed on a handful of platforms, and the possible implications for consumer welfare and competition.
Section 2 describes the pervasiveness of third-party tracking.
Third-party tracking is a mechanism through which a company (the third-party tracker) hooks onto another (first-party) website or application and collects identifiable data about users, enabling the tracker to build a comprehensive profile about these users. Tracking may occur both actively and passively. It may offer generic information on usage and webpage visits, or combined and analysed information which enables the identification of the individual.
The gathering of personalised data – through third-party tracking or otherwise – is primarily relied upon for four purposes in the digital realm: to provide data-based (i.e. individualized or targeted) advertisements, to provide individualised services (e.g. personalized search, individualised recommendations, matchmaking or personal training), to charge individualised prices (e.g. price steering, price discrimination), and to sell data sets (data brokerage).
Trackers may be described as branches of a tree, which in large numbers reach most corners of our online landscape. Interestingly, the roots of the tree exhibit a more limited spread, with a handful of companies gathering the data – with Google, followed by Facebook and Twitter, at the forefront. Thus, a handful of platforms that have direct access to user information and occupy important market junctions also engage in third-party tracking which enables them to harvest additional data.
Section 3 looks at the market power implications of third-party tracking, including for merger control.
Extensive data gathering and analysis in digital markets has the potential to create market power. When the tracker is a leading digital platform, third-party tacking may enable it to combine quality data from a variety of sources with information gathered directly on the platform. The scope of third-party tracking should therefore play a role in assessing market power, market dynamics and barriers to entry and expansion.
This is particularly the case as market power can lead to both exclusionary and exploitative concerns. Exclusion may affect potential or existing competitors who may not be able to reach the required volume and richness of data to compete effectively. Such exclusionary practices may take the form of input foreclosure (limited availability of data to other providers) or customer foreclosure (preventing other trackers from operating on the platform). Furthermore, the scope of data may provide the company with a detailed view of the market behaviour of competitors and the behaviour of competitors’ customers. Furthermore, exploitation may emerge from the increased asymmetry in information between users and providers. As a result, price and behavioural discrimination as well as other forms of manipulation may be more easily deployed.
Third-party tracking may also affect the appraisal of proposed concentrations between trackers, as well as between platforms and trackers. Market power could emerge on the data dimension even when the core markets where companies operate do not lead to a significant overlap. Instead, following a concentration possible harm may emerge when the merged entity has the ability and incentive to engage in exploitation or exclusion, facilitated by its tracking capacity. However, because power obtained through the consolidation of data is not necessarily reflected by market shares, the appraisal of the true data advantage stemming from the combinations of third-party tracking capacities is challenging and has attracted, to date, limited attention – as reflected in the light review pursued by the European Commission in Microsoft/Linkedin, Facebook/WattsApp and Google/DoubleClick. The authors suggest that the way these mergers were reviewed and cleared reflect an underestimation of the true (aggregated) data advantage the merging parties acquired thereby – leading legislators to address the issue (e.g. Germany has empowered the competition authority to take into account access to data when assessing market power). Furthermore, the realisation that data concentration is often not reflected through turnover has led to amendments of merger notification thresholds that focus on the value of the transaction, enabling enforcers to investigate mergers in which data – including personal user data gathered through third-party tracking – is key.
Section 4 focuses on exploitative practices.
Exploitation is made possible by asymmetries of information and data collection. This area is challenging from a competition enforcement perspective, reflecting different views as to the scope and role of competition enforcement. For example, the question ‘Is privacy a competition problem?’ has given rise to heated debates in competition circles. Similarly, the role of competition law in preventing price and behavioural discrimination, and in addressing the transfer of wealth between consumers and providers, have both led to lively debates. Inasmuch as these are considered to be problematic issues, the presence of third-party tracking will likely play a role in the analysis of harm, given that it could amplify a tracker’s data gathering capacity, have an impact on data concentration, undermine users’ privacy and autonomy, and increase the provider’s ability to exploit or manipulate consumers.
The authors demonstrate this by reference to two potentially problematic practices: quality degradation and excessive data collection.
- Privacy protection as a dimension of quality has been identified as a possible subject of competition intervention. The argument that a reduction in privacy may ultimately affect the quality of a product has gained increased acceptance. Consumer preferences in this area, however, are not so easily gauged. Competition enforcement will necessarily need to find a way to deal with what has become known as the privacy paradox: while consumers value privacy, they often do not (or cannot) act on this preference. In addition, the perceived lock-in effect experienced by users means that they are usually not able to bypass certain prevalent digital service providers, while so-called tracking walls require users to agree to third-party tracking before accessing a certain website or service.
- Another possible theory of harm concerns the exploitative effects of third-party tracking, and whether it may amount to excessive data collection. Similar to the analysis of excessive prices, one could consider the application of competition law to excessive or unfair data collection. This could occur where users are made to pay an amount of data that greatly surpasses what they reasonably expect. Such an approach is not without complexity, nor controversy. Absent credible benchmarks, it is suggested that competition enforcers could rely on other legal standards, such as those in data protection law.
Section 5 provides a number of concluding remarks.
The amalgamation of multi-sourced data, often with limited user awareness, has significant implications on our privacy, autonomy and welfare. As we consider the implications of third-party tracking, we should not underestimate the rise of multi-sourced data pools, operated by data and analytic giants. The extent to which third-party tracking would affect antitrust scrutiny remains to be seen. As third-party tracking takes place across markets, across platforms and across devices, we might be witnessing a rise in power over consumers even when, seemingly, market power relating to a specific antitrust market is not there (yet). As is often the case with data, the enforcement toolbox goes beyond traditional competition law remedies. In any event, competition law could add an important layer of protection against exploitative third-party tracking – as the ongoing antitrust probe into Amazon’s use of data on its merchant data may demonstrate.
This is an interesting paper that focuses on an under-discussed dimension of the digital economy – even though I have discussed it here in the past. I think this is recommended reading – as is normal for any work by either author.
It is undoubted that data is now squarely in the sight of competition authorities (and legislators), and that theories of harm are being developed in its regard. It should be uncontroversial that third-party tracking should be taken into account in this context. At the same time, I am still unaware of any merger or anticompetitive practice that has been prohibited or sanctioned solely, or even mainly due to data concerns (other than a French decision regarding access to a former State incumbent’s customer database, which was unrelated to the digital economy). Furthermore, the theories of harm being advanced here relate to the use of data more generally – i.e. traditional concerns regarding customer and input foreclosure –, which would seem to apply as long as there is evidence of these effects regardless of whether they relate to data or not. Exploitative abuses, on the other hand, are already somewhat problematic from a pure competition perspective, and become even more controversial when applied to data, as exemplified by the initial reaction to the German Facebook decision.
In other words, I have questions about how much this paper actually advances the debate. Regardless, this is an important area to watch, and this paper is a valuable contribution.