Madhavi Singh* & Ganesh A Khemka**
Introduction
Recently restaurant aggregators like Swiggy and Zomato launched their standalone private labels for food and meals called Snacc and Bistro respectively. This has caused consternation among restaurant owners who are sellers on these platforms. Restaurants worry that platforms have access to all their data about bestselling food in different areas, price sensitivity, consumer preferences, etc. This would allow them to imitate the business strategies or product offerings of bestselling restaurants, price-cut them and unfairly advantage their private labels in the search results/ listings.
In legal terms, restaurants are afraid that Swiggy and Zomato would leverage their dominant position in the market for restaurant aggregator/ delivery services to unfairly advantage their private-label food services. The National Restaurant Association of India has claimed that such conduct violates the Competition Act, 2002 and has threatened legal action. Meanwhile, Swiggy and Zomato are attempting to allay the fears of their restaurant partners by operating their private labels as a standalone app (outside of the Zomato/ Swiggy delivery apps) and asserting that they do not use their delivery platforms for data harvesting or advertising. Restaurants are unconvinced though and understandably so. Against this backdrop, this article argues that Section 28 of the Competition Act, 2002 should be used to prohibit dominant platform aggregators and marketplaces from acting in a dual capacity i.e., from operating as a seller on their own platform.
The Inherent Conflict Of Interest
Small and Medium Enterprises (SMEs) have long experienced the unfairness that arises when companies act as both the marketplace and seller in the same marketplace. Several investigations have revealed problematic practices in the e-commerce sector including copying successful products, secretly exploiting non-public seller data, offering steep discounts to undercut competitors, and manipulating search results to favor the company’s own products. Elsewhere, I have argued that the Indian competition regulator should prohibit Amazon from acting as a seller on its own marketplace. Unsurprisingly then, the restaurants are wary of the assurances given by Zomato and Swiggy and doubt that they will be treated fairly now that platforms have ‘skin in the game’ as a seller. When a platform is allowed to operate in a dual capacity—as both marketplace and seller, the economic rewards associated with preferential treatment of their private labels are so high that companies will always be induced to find ways to advantage their private labels.
Moreover, platforms are characterized by serious information asymmetries – it is difficult (if not impossible) to detect whether platforms are using non-public seller data to improve their private labels or whether (and how) platforms are manipulating their algorithms. Finally, the relationship of the sellers (or restaurants) with platforms is characterized by unequal bargaining power. If an individual seller or restaurant realizes that their data has been exploited or their products/ business strategy has been copied or their listing unfairly demoted, they have very few recourses available to fight against a well-resourced and sophisticated platform company. Organizing and taking collective action on behalf of sellers is also beset with difficulties.
Inadequacy Of Existing Ex-Post Action
Platforms are further emboldened when they realize that from a cost-benefit perspective, their conduct would probably go unchecked especially since it falls in ambiguous legal territory. Complaints brought by industrial associations like the National Restaurant Association of India or other trade associations often get hamstrung in lengthy investigations and complex legal and appellate processes. Even if subsequently penalized, the penalty, however high, is merely a cost of doing business in return for the protection and entrenchment of their dominant market position. In 2023 the Big Tech companies (Alphabet, Amazon, Apple, Meta and Microsoft) received total fines equaling $3.04 billion, and within seven days of 2024, they had earned enough revenue to pay off all of these fines. Indeed, Apple has been accused of choosing to pay fines instead of complying with a competition order. Essentially, penalties have thus far proven ineffective.
Additionally, the standard ex-post relief imposed at the end of a competition investigation usually dictates the enterprise to refrain from acting in anti-competitive ways again. Such remedies are difficult to monitor and enforce. Tech companies often find technical loopholes or workarounds which allow them to comply with the letter of the order while bypassing it in spirit. For instance, Google’s compliance mechanism for the EU Google (Shopping) order and Apple’s compliance mechanism for the Digital Markets Act (DMA) have both been alleged to be insufficient. Thus, traditional competition law remedies like penalties or behavioral remedies do too little, too late.
Given our experience with digital marketplaces and in the face of such strong economic incentives to engage in preferential treatment, it would be naïve to believe the word of the profit-maximizing platforms or sit around waiting for them to actually abuse their dominant position. Instead, we need to proactively bar them from acting in this dual capacity and take away any opportunity to act in a way that could subvert competition. This is the philosophy behind ‘structural separation’ remedies.
Structural Separation
Several scholars like Tim Wu and Lina Khan have advocated the use of ‘structural separation’ for digital platforms. Structural separation for digital platforms posits that actors who control the underlying infrastructure should be kept separate from those who control venues of access or other layers of the information ecosystem. This is aimed at eliminating incentives which make unfair or discriminatory behavior appealing in the first place. Structural separation has previously been used in sectors such as railroads, banking, and telecom. In cases like AT&T and Standard Oil, structural separation was used as an ex-post remedy in antitrust enforcement. Structural separation is more effective when implemented ex-ante rather than ex-post because competition once subverted is notoriously difficult to restore, especially in digital markets. Hence, ideally, we would want to prevent the subversion of competition by prohibiting enterprises from acting in dual capacities where they have the opportunity and the incentive to act in unfair and anti-competitive ways.
Power Of The Commission: Section 28
Interestingly, the Competition Commission of India already has the power to take such action and need not rely on statutory amendment or the introduction of a new regulatory framework. Section 28 of the Competition Act, 2002 allows the Commission to divide an enterprise enjoying a dominant position, “to ensure that such enterprise does not abuse its dominant position.” Sub-section (2) of Section 28 further confers on the Commission a wide range of powers (like transfer or vesting of property, adjustment of contracts, etc.) which might be necessary to give effect to the division of the enterprise.
The important question is whether Section 28 be used to enforce structural separation when there is no finding of violation under Section 3 (anti-competitive agreements) or section 4 (abuse of dominant position). For instance, in the recent Zomato-Swiggy controversy, there has been no finding of abuse of dominance yet. Can the Commission still intervene using Section 28 and demand a division of these so-called private label apps like Snacc or Bistro? While there is no judicial pronouncement on this issue and Section 28 has not been used till date, the plain meaning of the text clearly indicates that a finding of violation is not necessary to invoke this section.
The two prerequisites for invoking Section 28 are: 1) the enterprise enjoys a dominant position, and 2) the division is necessary to ensure that the enterprise does not abuse its dominance. Thus, Section 28 can be used even when the abuse of dominance hasn’t occurred, but such abuse might be imminent, and the division is necessary to prevent abuse. As discussed above, the moment a company operating a marketplace/ aggregator proposes to become a seller of the same product/ service for which it is a marketplace/ aggregator, then: 1) the harm to competition becomes imminent, and 2) division is necessary to prevent such harm (because other remedies are inadequate). Thus, when a dominant marketplace or aggregator like Zomato or Swiggy threatens to operate as a seller, the Commission can invoke Section 28 to separate the marketplace and seller entities. A division under Section 28 would ensure actual legal independence and separate operations. Such legal and operational separation is better than the status quo which relies on the largesse of Zomato and Swiggy and asks us to believe that these companies would somehow resist the strong economic temptation to preferentially treat their private labels.
The use of Section 28 even when an ‘abuse’ of dominance hasn’t been established seems like a reasonable interpretation, especially in light of Section 27. Section 27 contains all the other remedies which can be imposed by the Commission in an enforcement action and clearly states that these remedies can only be imposed upon a finding of contravention under section 3 or 4. This language which requires a finding of contravention as a precondition, is absent in Section 28. This interpretation also aligns with the overall purpose of competition law which is not just to correct practices that have already harmed competition but also to prevent them. The entire ex-ante merger regime is based on this premise of preventing harm to competition even before it has occurred. Section 28 operates in the same vein.
Conclusion
Conflicts of interest which arise when a platform operates as both infrastructure/ marketplace and as a participant on the same infrastructure/ marketplace, are pervasive. They are so pervasive that trying to remedy this by continuously admonishing players to not ‘abuse’ their dominance seems fruitless and results in a whack-a-mole situation. Instead, we need to nip such conflicts of interest in the bud and curb opportunities and incentives for players to abuse their dominant position. Fortunately, the Commission already has the tools for such ex-ante structural separation in the form of the long-forgotten Section 28.
*Madhavi Singh (HLS '24, Oxford '20, NLSIU '19) is the Deputy Director of the Thurman Arnold Project and a Resident Fellow of the Information Society Project at Yale Law School.
**Ganesh A Khemka (HLS ’22, NLSIU ’19) is a Delhi-based lawyer interested in antitrust, technology and democracy.
Comments