Antitrust law and data processing: With big data comes big responsibility
Through rapid digitalization and agile technology, the concept of “data” has become the new raw material of business, regarded as an economic input almost on a par with capital and labor. As modern computing power has expanded, our ability to collect, process, store, and analyze data on a large scale has also raised complex questions about the commercial nature of the accumulated “big data” and the implications for competition in numerous industries across the global economy. In that same vein, as Artificial Intelligence (AI), Machine Learning (ML), and the Internet of Things (IoT) promise to render big data analytics a central feature of virtually every area of commerce, antitrust experts – lawyers, economists, and agencies – are struggling to define big data in the parlance of antitrust and analyze it in light of the world’s antitrust laws.
What s “data” n the eyes of ant trust enforcers?
Before delving into which characteristics of big data are antitrust-related, it is of vital importance to see what exactly “data” is from an antitrust law perspective. At its simplest, “data” is a product and the same antitrust law analysis can be applied to it as is applied to any other product. In a similar manner, services founded on data can be analyzed in the same way under antitrust laws as any other service. The difficulty is that data is not finite in the same way as most products are, and therefore views differ wildly regarding its relative importance to competition. Some argue that large data sets that require a more sophisticated database and software techniques to process (also known as “big data”) is an important barrier to entry because data is difficult to collect, access, replicate, and process. Others assert that “[d] ata-rich companies are not an economic threat but rather an important source of innovation, which policymakers should encourage, not limit” because data – especially consumer data – is readily available, non-rivalrous, and ubiquitous, in that multiple entities can collect and use the same data without raising foreclosure concerns. In the midst of this hot debate, understanding regulatory authorities’ current thinking on the topic is essential for organizations who wish to navigate successfully through a rapidly changing commercial landscape.
What s the key regulatory author t es’ current th nk ng on b g data?
A well-accepted truism is that control of large amounts of data raises the possibility of giving companies an unfair advantage over competitors, allowing them to use their market power to harm consumers and competitors. This is drawing intense scrutiny from antitrust authorities in key jurisdictions around the world. Europe, in particular, has focused significant attention on antitrust and big data: In November 2015, the UK’s Financial Conduct Authority issued a call for input in relation to big data in the retail general insurance sector. In May 2016, the French and German antitrust authorities published a joint paper, titled ‘Competition Law and Data’. Following France and Germany, the Italian Antitrust Authority has just launched its first sector inquiry into big data together with the Italian Communications Authority and the Italian Data Protection Authority. Again, many big data-related issues have arisen during the e-commerce sector inquiry of the European Commission, the final report of which was published on May 10, 2017. Margrethe Vestager, the EU Competition Commissioner, in her January 2016 speech, ‘Competition in a Big Data World’ at the Digital Life Design Conference, acknowledged the growing relevance of data in antitrust law. Later, in another speech on the use of pricing algorithms, she also brought up the issue of collusion risks which may arise from algorithms that are poorly designed or misused by an automated system.
With its antitrust legislation closely modeled on the EU regime, Turkey is also paying ever-increasing attention to the phenomenon. The most recent instance in this respect, the Turkish Competition Authority’s ‘Big Data, Online Platforms and Competition Law Seminar’, organized in conjunction with the Turkish Industrialists and Businessmen’s Association (TUSİAD) on April 25, 2018, indicated that the Turkish Competition Authority will keep a close eye on big data among a broader set of concerns arising from digitalization. The event is notable in that it gives insight into the Turkish Competition Authority’s current thinking on the topic, which is essential to know for organizations that aim to capitalize on big data and analytics.
What could be the potent al ant trust r sks ar s ng from b g data?
The accumulation of data is not problematic per se under antitrust law. However, could owning a significant data set make you a dominant player and therefore subject to added scrutiny? Or can a post-merger combination of data increase your market power, or increase barriers to entry? Does conduct in relation to big data seem to be exclusionary (e.g. exclusive contracts, tied sales, refusal to provide access to data, discriminatory access, discriminatory pricing)? These competitive analysis questions should be answered on a case-by-case basis, combining sector-specific market dynamics with legal knowledge. Still, an effective antitrust case analysis would always require questions in relation to the characteristics of the collected data as well, such as (i) who is collecting the data, how, and on what subjects; (ii) whether comparable data are available from multiple sources; (iii) what is the marginal value of additional data; and (iv) what is reduction in data’s value over time.
Antitrust authorities in key jurisdictions fear that big data can lead to abuse of dominance, especially where undertakings hold unique datasets that cannot be replicated by competitors. This might lead to anticompetitive exclusionary conduct, typically in the form of exclusive contracts for the use of data or otherwise refusal to grant competitors access to certain data. Also, big
data can reinforce an undertaking’s dominant position in another market, for instance in markets where access to a particular data set is essential to enable competition in an upstream/ downstream/neighboring market.
Alternatively, in the application of big data-related technologies (e.g. pricing algorithms), there might be room for collusion among competitors. Partnerships can use algorithms to implement their agreement and fix certain price levels. In other words, pricing algorithms can function as a cartel instrument. Or, partnerships may unilaterally decide to create comparable algorithms aimed at maximizing their profit, leading to parallel market behavior. Although parallel market behavior does not constitute an illegal action per se, it may yield the effect that competitors agree on common price levels more quickly. In relation to merger control, big data acquired entirely by a single company may increase barriers to the entry of new players in a relevant market. The biggest dilemma here is that merger control enforcement in key jurisdictions is triggered based on transaction parties’ turnover, but in the context of a growing digital economy and the increasing numbers of digital start-ups, tech companies often fail to generate high turnovers at first. Although their corporate value may be significant as a result of their degree of innovation, the data sets they have owned, or their market presence in the eyes of sophisticated customers, a merger involving such undertakings may not trigger a merger control review due to their low turnover.
In recent years, the dramatic change in the magnitude and scope of data accumulation and organizations’ increasing ability to process it through modern computing power have put the notion of big data in the spotlight of the antitrust law world. Many authorities have already started to inquire into the sectors that are related to or mostly influenced by big data. Although big data’s actual impact on competition among undertakings is not clear yet, the general consensus is that the ability to generate and process large data sets can be associated to market power, and therefore an antitrust analysis should take account of its impacts (either pro- or anti-competitive).
The traditional antitrust approach can already address many data-related anti-competitive practices, yet there are still uncertainties and regulatory grey zones worth examining in the future. In the interim, undertakings (especially tech companies) should watch out for granting exclusive licensing of or exclusive access to important data sets. Dominant market players should tread carefully when operating their business in order not to face new categories of abuse of dominance claims. On the merger control front, transaction parties should make sure that their ability to accumulate and process large data sets does not lead to the emergence or increase of entry barriers.
Ultimately, companies should use big data as an asset, similar to their use of more traditional assets. In planning strategic big data-related transactions, they must be aware of the characteristics of their specific data (e.g. how, by whom, and on what subjects the data is processed), together with the dynamics of the markets in which they operate. If a particular use of big data might be perceived as having anticompetitive effects, companies must be prepared to justify their conduct to antitrust authorities.