Competition Law in Digital Markets

Digital markets do not require a different competition law, but they stretch the existing one. Network effects make positions durable that would elsewhere be temporary; services are supplied at zero price, so the price-based tests do not run; platforms serve several sides at once, so a market cannot be analysed in isolation; and the same firm operates the marketplace and competes within it. Indian enforcement has proceeded by applying Sections 3 and 4 to these conditions, and the question now before Parliament is whether an ex ante regime should be added alongside them.

1. What Is Different about Digital Markets

1. Network effects. Direct effects arise where the service is more valuable as more users of the same kind join it; indirect effects arise where users on one side attract users on another. Both make a leading position self-reinforcing, so market share is a stronger indicator of durable power here than in a conventional market.
2. Data. User data is an input which accumulates with use and improves the service. Whether it is a barrier to entry depends on whether an entrant can replicate it; data that only scale produces is a barrier, data that is widely available or quickly collected is not.
3. Zero price. Many services are supplied without money changing hands. The definition of price in Section 2(o) covers every valuable consideration whether direct or indirect, and attention and data are the consideration. What the absence of a money price removes is the test: substitution must be assessed through quality, data collection or advertising load rather than through a price increase.
4. Multi-sided platforms. A platform serving users and businesses prices the sides differently and often subsidises one entirely. A price rise on one side may be the consequence of a subsidy on the other, so each side cannot be judged alone.
5. Switching costs and lock-in. Accumulated content, learned behaviour, contacts and integration with other services make switching costly even where an alternative exists, and whether users multi-home or single-home is the decisive question.
6. Ecosystems. Connected services, an operating system with an application store, a browser, a payment service and a search engine, in which a position in one supports the others. Indian practice has defined a separate relevant market for each layer, which is what permits an inquiry into leveraging under Section 4(2)(e).

2. Dominance in a Digital Market

The assessment proceeds under Section 19(4), and three of its factors do most of the work: entry barriers, dependence of consumers and the residuary clause, which is how network effects and control of data enter the analysis. Market share remains the starting point, but a share sustained by network effects and switching costs supports a much stronger inference than the same share in a market where customers move freely. The absence of a price does not negate dominance, and the absence of revenue does not either, since the competitive significance of a platform lies in the users it holds.

3. The Recognised Theories of Harm

3.1 Self-preferencing and search bias

Where the operator of a platform also competes with the businesses using it, it may rank, display or default to its own service. The competition objection is not that a firm promotes its own products, which every firm does, but that this firm controls the route by which users reach the market at all, so the promotion excludes rivals rather than competing with them. The conduct is examined as denial of market access under Section 4(2)(c) and as leveraging under Section 4(2)(e). Search bias is the particular case of ranking, and its difficulty is evidentiary: the algorithm is opaque and there is no neutral baseline, so the analysis proceeds by comparing the treatment of the platform's own service with that of comparable third party services.

3.2 App stores and the terms of distribution

• Commission policies. A requirement that developers use the platform's billing system and pay a commission on transactions is examined as an unfair condition under Section 4(2)(a)(i) and as tying under Section 4(2)(d), the tied product being the payment service.
• Anti-steering restrictions. A prohibition on telling users that the service can be bought more cheaply elsewhere prevents the developer from competing on price outside the platform and insulates the commission from competitive pressure.
• Default pre-installation. Requiring a device manufacturer to pre-install a suite of applications, and to place them in default positions, is examined as the imposition of supplementary obligations and as leveraging from the operating system to the applications.
• Anti-fragmentation arrangements. An obligation on a manufacturer not to develop or distribute devices running a modified version of the operating system restricts the development of alternatives and is examined as denial of market access.

3.3 Parity clauses on platforms

A wide parity clause requires a seller not to offer better prices on any other channel, including a rival platform; a narrow clause covers only the seller's own direct channel. The wide form removes the principal way a new platform can enter, which is by charging lower commission and letting sellers pass the saving on, and it has accordingly been examined under Section 3(4) and, where the platform is dominant, under Section 4. The narrow form is easier to justify as protection against free riding on the platform's investment. Such clauses have been examined in the Indian hotel and travel booking sector.

3.4 Data and privacy

Two distinct theories arise. The first is data exploitation: a platform uses data generated by the businesses operating on it to compete with them, or combines data across services in a way rivals cannot match, so that the advantage flows from the position rather than from competition on the merits. The second is privacy degradation: where a service is supplied at zero price, a unilateral worsening of the data terms is an increase in the effective price, and the Commission has proceeded on the footing that it may be examined as the imposition of unfair conditions under Section 4(2)(a)(i). The interface with the Digital Personal Data Protection Act, 2023 is one of coexistence rather than conflict, since the two statutes ask different questions about the same conduct, but the risk of inconsistent remedies is real.

3.5 Algorithmic pricing and collusion

Three situations must be separated. An algorithm used to implement an agreement already reached raises no new question: the agreement is proved in the ordinary way and the algorithm is the instrument. Competitors using the same pricing algorithm or the same provider is the hub and spoke case, which the amendment of 2023 addresses by presuming participation where an enterprise not in the same trade actively furthers the arrangement. The third is unresolved: independent algorithms observing each other may learn to sustain supra-competitive prices with no communication between the firms, and Section 3 requires an agreement, so conduct with no meeting of minds falls outside it. In Samir Agrawal v. Competition Commission of India, (2021) 3 SCC 136 the Supreme Court affirmed that algorithmic fare setting by cab platforms did not, on those facts, establish a horizontal agreement among drivers.

3.6 Killer acquisitions and ecosystem consolidation

The acquisition of a nascent rival whose product would have competed with the acquirer's, made in order to discontinue or absorb it, escaped the asset and turnover thresholds entirely because such targets have neither. The deal value threshold in Section 5(d), notified with effect from 10 September 2024, was introduced principally for this reason: a transaction exceeding two thousand crore rupees is notifiable where the target has substantial business operations in India, and the small target exemption does not apply to it.

4. Interoperability and Remedies

The remedies sought in digital cases are mostly behavioural, and interoperability is the recurring demand: that a platform permit rival services to connect to it, that data be portable so that a user can leave without losing what it has accumulated, or that a developer be free to use an alternative billing system. Such remedies attack the source of the durable position, which is the lock-in, rather than the conduct alone. Their difficulty is that they require continuing supervision, technical specification and a mechanism for resolving disputes about compliance, which is why the settlement and commitment routes under Sections 48A and 48B have become the practical vehicle for obtaining them, as the Android television settlement of April 2025 shows.

5. Ex Ante Regulation: The Current Position

The case for an ex ante regime rests on three propositions: that enforcement after the event is too slow for markets that tip quickly, that the same conduct recurs across cases so that case-by-case adjudication wastes effort, and that once a market has tipped no remedy restores what was lost. The case against it is that obligations imposed in advance on designated firms regulate conduct without proof of harm, may forbid conduct that is efficient, and require the regulator to define categories that technology outruns.

1. The comparative model. The European Union's Digital Markets Act designates gatekeepers by quantitative criteria and imposes obligations and prohibitions on them directly, including on self-preferencing, data combination, anti-steering and interoperability, enforced without any finding of dominance or of effect.
2. The Indian position. A Committee on Digital Competition Law was constituted by the Ministry of Corporate Affairs and submitted its report in 2024 with a draft Digital Competition Bill, which would designate systemically significant digital enterprises by quantitative thresholds and impose obligations of the kind described. The draft was released for public consultation.
3. Its status. The draft Bill has not been introduced in Parliament and is not law. Digital conduct in India continues to be governed by Sections 3 and 4 of the Competition Act, 2002, with merger control extended to digital acquisitions by the deal value threshold. An answer should state that position first and describe the draft as a proposal.

6. Related Topics and Provisions