Aprameya Sheshadri & Anirud Raghav*
Introduction
The emergence of digital markets has immutably transformed the way modern-day regulation is envisaged. From data privacy to competition law, sui-generis regimes are cropping up in a bid to address unprecedented regulatory challenges. Specifically, we focus on the recent Digital Competition Bill (“DCB”), which introduces putative “ex-ante” measures to regulate digital competition. EU’s Digital Markets Act (“DMA”) set the ball rolling by formulating a sui-generis competition regime designed to regulate significant enterprises’ digital services. This model has quickly spread to the United Kingdom, Germany, and Japan, among other countries.
This paper is an attempt to add meaningfully to existing literature and focus on under-addressed issues in the DCB. Centrally, this article will argue that the DCB’s sui-generis regime is unlikely to enable timely intervention and improve contestability. We recognize that the rationale and aims of the DCB are vital and cannot be dealt with under traditional competition models. Thus, we welcome the DCB while highlighting the need for greater conceptual clarity and potential amendments in certain areas.
We structure our discussion as follows: First, we examine ex-ante obligations in the DCB and argue that the enforcement procedures remain largely ex-post, being unlikely to achieve timely intervention effectively; Second, we discuss mandatory data-access obligations in the context of improving contestability – one of the major aims of the Bill. We exhort the incorporation of data-access obligations and dispel any concerns surrounding the same. We conclude by reflecting on the prospects of DCB.
Ex-Ante Approach: A Confused Label
The key quality of the DCB is that it adopts an ex-ante regulatory approach. This section will attempt to clarify the sense in which the DCB adopts an “ex-ante” approach and then argue that most forms of intervention still resemble an ex-post enforcement mechanism so that the goal of timely intervention stands defeated.
The Scope of “Ex-Ante”
Before exploring the concept of an ex-ante approach, it is crucial to outline the key goal of the DCB, which also serves as the rationale for adopting this approach i.e. ensuring timely intervention (p. 34, the Committee on Digital Competition Law (“CDCL”) Report). The inherent peculiarities of digital markets such as multi-sidedness, cross and same-side network effects, zero-price services, and access to vast repositories of consumer data have a unique potential to rapidly fortify existing incumbents’ position. As such, large digital enterprises that are not ‘statutorily’ dominant but which may nonetheless wield the ability to influence markets may therefore escape scrutiny. Thus, timely intervention in the digital markets is essential to prevent the market from tipping irreversibly in favour of dominant enterprises.
The traditional ex-post approach has been ineffective in addressing this mainly due to two reasons: First, the several stages involved in the enforcement proceedings such as the formation of a prima facie view by the CCI, investigation by the Director General, and passing of final order by the CCI are heavily time-consuming. The act does not stipulate time restrictions for the CCI to pass orders. In fact, as noted by the Committee, only three final orders under §27 of the Competition Act have been passed and all of them are under appeal either under the NCLAT or the SC; Second, defining the ‘relevant market’ (a necessary condition for the application of §4 of the Competition Act) and assessing dominance in multi-sided digital enterprises is a complex, evidence-based process, requiring evaluation of substitutes and market strength, which significantly delays grievance redressal.
Despite extensive critiques on ex-ante regulation, surprisingly little attention has been paid to the term “ex-ante” and has been assumed to be self-evident. Despite using it innumerable times in its report, the Committee on Digital Competition Law does not once clarify what ex-ante means. Understanding this term is crucial because it is portrayed as a major reason for facilitating timely intervention. Traditionally, the ex-ante approach initiates enforcement[1] against “externality producers” even before the externality (typically, wrongdoing) occurs. Consider ex-ante measures even within §6 of the Competition Act, 2002 in the form of merger control. The requirement of prior approval typifies an ex-ante approach where mergers are made conditional on the approval of the Competition Commission of India so that even before anti-competitive effects result, they are assessed predictively. An ex-post approach, then, initiates enforcement only after the wrongdoing has already taken place.
It would now be useful to consider why the ex-ante approach was chosen over the ex-post approach in digital markets. From the CDCL report, timely intervention appears to be the principal rationale for preferring an ex-ante approach. The report bemoans the time-consuming and resource-intensive nature of ex-post investigations. More importantly, it notes that the market might already have tipped by the time the ex-post investigation is completed. The competitive dynamics of digital markets, including the strong influence of network effects and winner-take-all features, are such that market tipping can occur rapidly and irreversibly. Considering this, it is crucial to effectively stop market tipping prospects in its tracks. Thus, an ex-ante approach was considered the panacea to the market-tipping problem. Moreover, the report ignores the costs accompanying an ex-ante approach – immense monitoring costs. A holistic analysis ought to account for the costs of the preferred approach as well – not merely the costs of the approach it is abandoning.
An Ex-Ante Approach with Overbearing Ex-Post Realities
As noted earlier, the ex-ante approach aims to intervene before the violation occurs. §3 of the DCB places certain entities, identified as SSDEs within the Core Digital Services to a higher standard of scrutiny. This is due to the significant market power that these entities have acquired, driven by key digital market characteristics such as data feedback loops, strong network effects, and economies of scale and scope. The issue does not lie in questioning the rationale behind this heightened scrutiny, which we consider essential in such markets. Rather, we analyse whether the functions carried out by the DCB can be classified as ex-ante or not.
The DCB imposes prohibitory and mandatory obligations upon SSDEs (Chapter III, CDCL Report). Prohibitory obligations are rules designed to prevent gatekeepers from engaging in practices that distort competition, such as bundling services, restricting user choice, favouring their own services, or misusing data without proper consent. Mandatory obligations, on the other hand, impose affirmative duties on gatekeepers to foster fairness, transparency, and competition by enabling interoperability, ensuring data portability, allowing service customization, and providing equitable access to data for business users and end users.
The only requirement triggering these obligations is that the entities providing a Core Digital Service (as delineated in the schedule) must meet certain quantitative or qualitative thresholds, as may apply as per §3 and §4 of the DCB. Until this point, the approach is ex-ante in the sense that these big tech entities are subject to the obligations even before any anti-competitive practice might have occurred. In contrast to an ex-post approach, these entities are subject to obligations merely upon attaining dominance, even if no abuse has occurred. So, why did the DCB hasten to enact an ex-ante approach? The answer lies in the inherent difficulty in digital markets in precisely delineating relevant markets, determining dominance, and identifying abuse—prerequisites inherent in the ex-post framework.
In this context, we argue that the DCB’s ex-ante approach is still heavily burdened with an ex-post reality. Once the SSDEs are identified and the obligations accrue, the enforcement process is still largely ex-post, i.e., a penalty or remedies are provided only upon a violation. Moreover, the enforcement process includes an enquiry into non-compliance with the obligations under Chapter III of the DCB. This enquiry is conducted by the Director General and follows a procedure similar to that of the ex-post one under §26 of the traditional Competition Act. Such an enforcement process is in contrast to other ex-ante regulation regimes including regulation of mergers and acquisitions under the Competition Act itself.
While considering the nature of the intervention under EU’s DMA, a difference emerges when compared to the DCB. Article 18 of DMA provides that the Commission shall conclude the market investigation as to non-compliance of obligations within 12 months. While the CDCL report notes that such an outer timeline is desirable (p.34), the DCB does not provide for any such time limits within which the CCI must complete its investigation. Further, the DMA sets out such specific time limits for multiple other stages of the enforcement process. We argue that such time limits are highly desirable to achieve the stated aim of timely intervention and that the omission of the same from the DCB must be corrected.
The CDCL report also recognises the vast overlap and complementary nature of the “ex-ante” and “ex-post” regimes.[2] However, we contend that while the rationale to have a de novo specialised act for digital markets is well-founded, it might be premature to attribute a large degree of ex-anteness to a bill that still retains considerable ex-post effects and enforcement mechanisms. Here, we emphasize the limited point that the measures under the DCB are no more ex-ante than the obligations under the traditional Competition Act, and therefore, attributing the ex-ante label as a distinctive feature of the Act may be doctrinally confused. Similar concerns have started to emerge amongst practitioners as well. Bearing this in mind, we evaluate whether the DCB effectively achieves two of its primary objectives: timely and expeditious intervention, and enhancing contestability, especially in the context of the fast-paced nature of the market in the era of Artificial Intelligence.
Do the Measures Under the DCB Lead to Timely and Speedy Intervention?
It is trite that digital markets are highly fast-paced and one of DCB’s main objectives is to address the time-consuming nature of the ex-post interventions under the traditional Competition Act. Drawing partly from the above discussion, we argue that this objective is not met for two reasons. Firstly, the DCB aims to reduce the time consumed in litigation by eliminating the need to go into the categorisation of relevant markets, and dominance inter alia. Under the DCB, an exhaustive list of the Core Digital Services is notified and the act applies only to entities operating in these domains. At first blush, this seems to speed up the process. However, this might not be the case. §4 of the DCB allows entities to contest whether their product or service falls within any of the Core Digital Services. This especially allows for delays in classification considering the highly evolving and specialised digital market. For example, AI tech giants may contest that they do not fall under any of the nine Core Digital Services as provided under Schedule I of the DCB owing to the highly specialised nature of the Generative AI services.
Secondly, according to §3(2) of the DCB, to qualify as an SSDE, the entity must satisfy the applicable quantitative thresholds in the preceding three years. We argue that the requirement to meet the relevant quantitative threshold for the immediately preceding three years is under-inclusive, and could prove counterproductive. A three-year period is sufficient to tip the market in favour of the first mover, thereby defeating the thrust of the provision. A recent example is that of ChatGPT, which attracted close to 100 million subscribers within a meagre 2 months of its launch. While this may indeed be an outlier, similar cautions have been issued by other scholars. Considering the main objective of the act is to intervene in time before the market tips, it is important to operate under the caution that evolutions in the digital market occur rapidly. In such cases, it becomes crucial to periodically update the schedule to encompass emerging sectors promptly and provide flexibility. However, a three-year duration for reviewing the schedule appears excessively long in light of the rapidly evolving digital landscape. Thus, we suggest that this time period threshold must be reconsidered.
Improving Contestability – Why do Data Access Obligations Miss Out?
Contestability, a foundational principle in market theory, suggests that as firms approach monopolistic dominance, the scope for competitive freedom diminishes. Evidently, then, contestability an important aim of the DCB, as evident from the Preamble. The DCB takes important steps in furthering contestability by incorporating provisions enabling data portability (§12 DCB) and interoperability (§13 DCB). While data portability improves contestability by allowing users to port their personal data from one company to another without losing any benefits, interoperability allows different digital services to connect and communicate with each other based on a standardized protocol. These provisions mitigate network effects as well as the ensuing consumer lock-in and enable multi-homing, thereby enhancing contestability. While these two provisions inspire optimism, we notice a conspicuous absence of obligations mandating the grant of access to data. This section will address the same.
Why is Data Portability Not Enough?
To understand why only data portability is not sufficient to increase contestability, we must understand that data exists in three different forms i.e., volunteered data, observed data, and inferred data. Volunteered data denotes information deliberately shared by users online, encompassing personal preferences and demographic details. Observed data arises as a consequence of digital platform usage, comprising activities like purchase histories, browsing behaviour, and search queries. Inferred data, on the other hand, refers to insights derived from user behaviour and interactions with digital platforms, often through algorithms and artificial intelligence, to predict preferences or behaviour patterns. Of these three, inferred data is the main contributor to the creation of entry barriers. This is because it is used to forecast demand, develop or improve products and services, or target advertisements and hence comprises an essential input to be able to compete effectively in digital markets.
Data portability allows business users and end users to easily port their data, in a format and manner as may be specified under §12 of the DCB. Thus, it only allows for data contestability on the first two types of data and not on inferred data which as we have already noted, is the main reason for the creation of entry barriers. Thus, increasing contestability requires an effort to categorically allow for data (inferred data) sharing by SSDEs.
What are Data Access Obligations?
Generally speaking, data access obligations typically require gatekeepers (or in DCB’s case, the SSDEs) to grant access to various types of data in its possession on request by end users, business users, and other third parties, including competitors. These come in different forms. For instance, within Article 6 of EU’s DMA, there exist provisions mandating the sharing of performance measurement tools, aggregated and non-aggregated data to business users (presumably to enable consumer preference solutions), and providing data to search engine undertakings on fair, reasonable and non-discriminatory terms. In all their different forms, they serve the same purpose – to ensure that competitors can freely access data, and compete on quality and merits. This type of data-sharing obligation features in the EU’s DMA, the UK’s competition regime, and Germany’s law on abuse of economic power. This is to say that there is precedent for data access obligations, making it all the more curious that it should miss out on the DCB. Yet, before prematurely attributing the omission to drafting negligence, we examine what potential reasons might have warranted an intentional omission of the data access provisions.
An Oversight or a Deliberate Omission?
There are at least two reasons why we would expect such a provision to be present in the DCB. First, most of the Act seems to be inspired by Western competition jurisprudence,[3] including the UK’s DMU and the EU’s DMA, both of which stipulate data access obligations. Second, and more importantly, data serves as a critical insurmountable structural barriers in digital markets. Considering that analogous provisions designed to neutralize structural barriers (like data portability and interoperability) have made their way into the Bill, it is inexplicable that data access obligations should miss out. Nonetheless, one might argue that there is reason to believe that this was an intentional omission and not an oversight. Consider that India’s economy is primarily dominated by the service sector, which contributes nearly 54% to the GDP. Incorporating adverse provisions might negatively impact the ease of doing business, which a service-dominated economy can ill afford. Moreover, fears around overregulation and error costs loom large.[4] Innovation and privacy concerns subsist as well. The DCB exhibits a strong bent towards valuing data privacy, as is evident from §12(2) of the bill, which prohibits SSDEs from allowing third parties to access end users’ personal data.
Yet, we submit that valuing data privacy need not be at odds with enabling data access – the EU’s DMA mandates the provision of data access but requires personal data to be anonymised. This way, a balance is struck between data privacy and addressing contestability by neutralizing data’s role as a structural barrier to new entrants. Furthermore, innovation concerns might be unfounded because the obligations relating to data access focus on the data that the gatekeeper collects. This data, by itself, is not a source of innovation. It is merely the subject that innovative techniques use to deliver quality. Thus, even if all the competitors in the world were to have access to the same datasets, the incentive to innovate would not diminish. Rather, the exact opposite is true – since all competitors now enjoy access to the same data, the competition is directly on quality and merits, which would likely accelerate innovation.
Conclusion
In this piece, we have made two central points. First, it is inapposite to apply the label ex-ante to the DCB, considering that the operative procedures and investigations continue to apply in an ex-post fashion. More importantly, we have argued that it is unlikely that the measures will achieve timely intervention, raising questions about whether the primary purpose of the Act is met. Second, we have examined the extent to which contestability is furthered by the DCB, with special emphasis on data portability and data-access obligations. Firstly, we have argued that data portability, by itself, does not suffice because the type of data that actually poses structural barriers, i.e., inferred data, continues to remain exclusively with the SSDE’s. Secondly, we have noted that data-access obligations curiously miss out on the DCB’s scheme and exhort its incorporation. Regardless, the DCB is a welcome and resilient move representing a necessary adaptation to a rapidly changing digital landscape to protect the rights of both business users and end users. In a growing service sector economy, positive implications for ease of doing business, especially for new entrants, bode well for a sound competition ecosystem. Yet, the jurisprudence on ex-ante law is still nascent and we must keep a vigilant eye to see how things unfold.
[1] Enforcement here is used in a broad sense and can include any type of regulatory control exercised (by the CCI) against entities triggering such regulation by satisfying the criteria for ex-ante regulation.
[2] Concerns regarding simultaneous proceedings under both the DCB and the Competition Act have also arisen. This could lead to higher costs for both the enforcement authorities as well as the entities subject to the proceedings, thus leading to high inefficiencies. See page 112 of the CDCL report.
[3] Report of the Committee on Digital Competition Law focuses on the ex-ante provisions in European Union, United Kingdom, Germany inter alia.
[4] Report of the Committee on Digital Competition Law ¶4.1.5 (noting the concern of overregulation), ¶4.3.26 (noting the importance of ease of doing business).
*Aprameya Sheshadri & Anirud Raghav are B.A., LL.B. (Hons.) students at the National Law School of India University (NLSIU), Bengaluru
Comments