Assessing the Nine Anticompetitive Practices Under the Draft Digital Competition Bill

This publication was produced under Carnegie India’s Technology and Society Program. Research for this paper was made possible in part by support from Google India Private Limited and Amazon Seller Services Private Limited. For details on the program’s funding, please visit the Carnegie India website. The views expressed in this piece are solely those of the author.

Introduction

There are nine anticompetitive practices (ACPs) highlighted under the recently withdrawn draft Digital Competition Bill (DCB).¹ However, each of them is not necessarily anticompetitive per se—many of them are usually seen as per a rule of reason approach; and many of them also have a pro-competitive rationale. Therefore, not all can be painted with the same brush, and not all are equally problematic.

Accordingly, this article will look at each of the nine ACPs, starting off by examining the explanation provided under the Committee on Digital Competition Law (CDCL) report for their perceived anticompetitive nature. It will then look at the redeeming features of each of the ACPs and how their outlawing under the DCB, if implemented, may have an impact on product innovation, consumer welfare, and the cost structure of products offered. Essentially, this article is about regulatory choices and the unintended consequences they may bring about. The article will close with what it views as essential elements of any proposed law on digital competition—the absence of a legitimate business justification defense for the systemically significant digital enterprises (SSDEs) under the current DCB, as well as allowing SSDEs to demonstrate that there can be no less restrictive alternative when it comes to the ACPs in question.

In general,it is clear that no granular-level study was done during the course of the CDCL to look at how a digital economy is demonstrably impacted by any of the nine ACPs. The CDCL report only acknowledges that it engaged in consultations with select stakeholders and there was no impact assessment carried out, unlike in the European Union or the United Kingdom (UK) for the European Union’s Digital Markets Act (EU DMA) in December 2020 and the Digital Markets, Competition and Consumer Bill (DMCC), under the UK’s Competition and Markets Authority (CMA), in April 2023.² It is not clear why this is the case under the CDCL. Perhaps a recent January 2025 report titled “Estimation and Measurement of India’s Digital Economy” prepared by Indian Council for Research on International Economic Relations (ICRIER) for the Ministry of Electronics and Information Technology (MeitY) may provide some insights.³ This joint report looks at how digital markets are often integrated with offline markets and how “the cross-cutting and integrated nature of digital technologies makes the concept of a distinct digital economy difficult to define and measure.”⁴ This may explain why the Ministry of Corporate Affairs, the nodal ministry for the CDCL report, may have had difficulties doing an impact assessment. In this regard, it is interesting to note that the ICRIER-MeitY report states that “new digital industries, which include Big Tech players, other digital platforms and intermediaries, and firms dependent on digital intermediaries, account for nearly 2% of GVA.”⁵ It also highlights that “the highest growth is likely to come from the growth of digital intermediaries and platforms, followed by higher digital diffusion and digitalization of the rest of the economy.”⁶ In other words, it appears that all facets of the digital economy ecosystem, and not just the Big Tech platforms that have been highlighted in the CDCL report, are growing.

Tying or Bundling

There is often an overlap between digital products and other related hardware or software applications which may provide added functionality. Large digital enterprises may choose to tie or bundle their main or ‘must have’ product with complementary products and offer the same as a package to their users, thereby limiting consumer choice and simultaneously foreclosing competition from smaller rival firms in the market. This enables digital enterprises to expand and consolidate their position in adjacent markets as well.⁷

Currently, there is no clear test in the jurisprudence of the Competition Commission of India (CCI) on bundling. The commission does not follow a fixed, formal test for bundling, but it does have a series of steps it considers when looking at whether “tying” is anticompetitive. For instance, in Sonam Sharma v. Apple & Ors., the commission held that:

A tying arrangement occurs when, through a contractual or technological requirement, a seller conditions the sale or lease of one product or service on the customer’s agreement to take a second product or service. In other words, a firm selling products X and Y makes the purchase of product X conditional to the purchase of product Y. Product Y can be purchased freely on the market, but product X can only be purchased together with product Y. The product that a buyer is required to purchase in order to get the product the buyer actually wants is called the tied product. The product that the buyer wants to purchase is called the tying product.⁸

In doing so, the CCI laid down the following ingredients as to what constitutes tying:

a) Presence of two separate products or services capable of being tied: In order to have a tying arrangement, there must be two products that the seller can tie together.

b) The seller must have sufficient market power with respect to the tying product: An important and crucial consideration for analyzing tying violation is the requirement of market power. The seller must have sufficient economic power in the tying market to leverage into the market for the tied product.

c) The tying arrangement must affect a “not insubstantial” amount of commerce: Here, no bright-line test or clear threshold in terms of market share has been provided by the CCI.

However, the CCI does not have a formal and precise test for bundling. This is unlike other jurisdictions in other countries, such as the United States, which has a case in terms of the “incremental price-cost test” in Cascade Health Solutions v. PeaceHealth.⁹ Under this test, if a plaintiff can show that the overall value of the bundled rebates exceeds any profits made by the defendant firm, that is, the overall rebates provided were below their cost, and these below-cost rebates excluded any rivals from the market, then a valid claim for bundling may be made. By way of an example:

Scenario

A dominant firm (Firm X) sells two products:

●  Product A (for example, general lab services)

●  Product B (for example, specialized cardiac tests)

A smaller rival (Firm Y) only sells Product B.

The breakdown of Firm X’s pricing is as follows:

●  Product A = INR 100

●  Product B = INR 100

●  But if a customer buys both products (A+B), then Firm X offers a bundled discount and charges INR 150 instead of INR 200. Here, the total discount amounts to INR 50.

Applying the U.S. (PeaceHealth case) test

First, the full discount of INR 50 would be allocated to Product B (the product on which Firm Y competes). Therefore, the effective price of Product B would be:

INR 100 - INR 50 = INR 50.

Second, this effective price (INR 50) is compared to Firm Y’s cost to produce Product B. Here, if Firm Y’s cost is INR 60, then it can’t profitably compete, and the discount may be anticompetitive. But if Firm Y’s cost is INR 40, then it can match or beat the price of Firm X, and therefore, the discount is likely legal, because it shows that the defendant firm (Firm X) is more efficient. Under the DCB, the bundle would be anticompetitive merely because it is offered by the SSDE, regardless of which firm might be more efficient.

Furthermore, bundling is not always a bad thing. In Epic Games, Inc. v. Apple, Inc.,¹⁰ the value of bundling was brought out by the court of appeals: “Software markets are highly innovative and feature short product lifetimes—with a constant process of bundling, unbundling, and rebundling of various functions. In such a market, any first-mover product risks being labeled a tie pursuant to the separate-products test.”¹¹ For instance, if we were to look at the freemium model, then bundling may not be possible for many large firms. The freemium model is the model through which many tech companies start off by providing their offerings—and then move to a higher tier—so that they can offer more than basic versions of their services (that are offered under the free version) with advanced features. This could also be called bundling—essentially, one is using two separate products and clubbing them into a single product. This may affect product innovation since the SSDE here will be punished for trying out bundles in the first place.

For instance, the music streaming firm Spotify bundling certain features under its Premium model (which can otherwise be sold separately) could be a problem. Indeed, under a complaint brought at U.S. Federal Trade Commission by the National Music Publishers’ Association, it was stated:

Spotify also has deceived consumers by falsely representing that the ‘bundled’ Premium Plan adds substantial value to consumers and that its Audiobooks Access Plan is valued at close to the Premium Plan. This material misrepresentation that the ‘bundled’ content adds substantial value is likely to mislead reasonable consumers. By bundling audiobook and music services together, Spotify allowed consumers to believe that they were receiving additional audiobook access free of charge. . . . It also has deprived Premium Plan users of any option to access a music-only subscription plan, leaving consumers who want to listen to ad-free music without paying for audiobooks, without any options. . . . Spotify’s ‘bundling’ harms consumers by duping them into paying more for a service that they do not want.¹²

This is a good case of the possible reputational repercussions to poor bundling—in some cases, using the tying product necessitates the use of accompanying complementary products. If the complementary products do not meet certain criteria, the operating quality of the tying product could be reduced, which could have negative implications for its reputation.

Anti-Steering Provisions

A form of exclusionary conduct practised by large digital enterprises, particularly in the app store market, which prevents consumers and business users from shifting to third-party service-providers which may offer cheaper alternatives with better functionality. Such practices suppress consumer choice and prevent app developers from communicating with users and re-directing them to better substitutes outside the app environment.¹³

The aforementioned conduct has also been referred to as the “anti-steering” provision. A common grievance that has been highlighted when it comes to anti-steering is that app stores have restricted the ability of app developers to inform consumers within an app of the ability to purchase in-app content elsewhere (on a website, for instance). As the CCI pointed out in an earlier case, “the app is the primary and, in most cases, only medium for the app developers to communicate with their users and as such, critical for app developers. The app developer should have freedom to choose their communication channel to interact with their users to promote and offer their services.”¹⁴

However, most app developers choose to redirect their users to outside the app environment of the app store, since the app stores had earlier charged a 30 percent commission on all app purchases as well as in-app ones. Enabling the app developers to provide a route through which users can access the same apps and features without having to pay a 30 percent commission would help provide massive savings for the app developer. Indeed, both Google and Apple have now reduced their commission rates to 15 percent for smaller developers in order to incentivize them to remain within their app ecosystems.¹⁵

However, some commentators have gone as far as arguing to do away with a commission structure altogether and instead going for a flat-fee alternative. However, might this hurt startups even more? With the commission model, app developer startups only have to pay the app store if they make a revenue, which would not be the case with a flat-fee structure. Similarly, even where third-party app stores are now provided as an option for app developers through which to distribute their apps, an app store ecosystem like that of Apple has now come with up two options. First, they could either stick with the old regime of a 30 percent commission, or have the option to pay a considerably smaller commission of 17 percent,¹⁶ and even sidestep the commission altogether where the app is downloaded using a third-party app store. Certain fees would still have to be paid to Apple for its provision of “value Apple provides developers through ongoing investment in the tools, technologies, and services that enable them to build and share innovative apps with users.”¹⁷ However, the catch here is that for these types of third-party app store apps for which no commission is provided to the likes of Apple, the tech giant still reserves the right to charge a flat $0.50–$0.55 per user per year, provided the app in question has a million users/installs, which includes software updates as well.¹⁸

This may sound doable for the app developers once their app goes beyond the threshold of a million downloads/installs. However, it should be noted that most app developers have a freemium model where they do not charge users at all until and unless they want to use select additional features. Therefore, a new model, where potentially no commission is charged by the major app stores, could still prove to be an expensive proposition for startup app developers. Even for larger companies like Spotify—which has a freemium model—paying a hefty annual fee without the corresponding paid subscriptions to offset it may be a risk.

Therefore, it can be argued that anti-steering provisions themselves are a problem in as much as they limit the freedom of app developers to communicate with their users. However, the anti-steering provisions have also been brought up as a competition law issue since they are allegedly enabling app stores of Big Tech firms to extract high commissions from app developers constrained to their ecosystems.

In this regard, there are two observations to be made here.

First, what constitutes a “high” commission is not known as per the current jurisprudence of the CCI. This is because there is no rational basis to measure when prices or commission rates are “high.” As a small caveat, the CCI’s Shamsher Kataria case did mention that the CCI looks at the “concept of unfairness of price” by gauging whether “such price is unrelated to the ‘economic value’ of the product” provided.¹⁹ However, this is a notoriously difficult exercise fraught with challenges that the CCI itself has acknowledged when it observed in the same case that “this can be a particularly difficult task given that an enterprise may have diverse production and market operations which incurs various categories of costs and working out the production costs may raise great difficulties, especially determining what costs should be taken as a basis for calculating the cost-price ratio to show whether the price charged exceeds the costs incurred.”²⁰ There may be lessons from the literature in the area of standard essential patents (SEPs), where SEP holders license their patents to licensees at certain royalty rates under the Fair, Reasonable, and Non-Discriminatory framework.²¹ Here, royalty rates are applied to the price of the final product created by the licensee—this approach is similar to the current practice of commission rates charged by app stores on the purchase price of the product provided by the app developer. However, the difficulty here is that while under the SEP literature, it is possible (though challenging) to disaggregate the value of patented technology licenses from the value of the other components of a licensee’s product, this is not feasible in relation to the app store debate. The app stores provide features to app developers like discoverability, security vetting, and platform infrastructure—how much value can be attributed to these features by app developers is almost impossible to gauge and agree upon.

Second, the anti-steering provisions by themselves do not directly lead to higher commissions being charged from app developers. Even with the anti-steering provisions to be removed from all app stores, as mandated by courts in most countries, app developers still have problems—for instance, making upfront payment in Apple’s App Store for their largely free user bases.²² The rationale behind such types of flat fees, if articulated, is that Apple spends a lot of time and effort creating certain developer tools to make this interoperability possible with apps downloaded from third-party app stores that then benefit from Apple’s wide platform base. Anti-steering provisions were accordingly meant to eliminate the free-rider problem. Therefore, even with these provisions now being on the wane, the larger problem of how to derive a fair share of value from the transactions undertaken on these app store platforms remains. Interestingly, certain sectors that have seen the use of commission rates at par with those charged by app stores have been left untouched by the CDCL report—namely, the mobility sector. For instance, the ICRIER-MeitY report highlighted how the commission rates for other sectors in the digital economy hover around 15–30 percent as well, with 30 percent being the rate for the mobility sector.²³ Details of the average commission rates of other sectors are provided below.

Search Ranking, Self-Preferencing, and Platform Neutrality

In cases where large digital enterprises also play the role of retailers on their own platforms, such entities may leverage their dominant market position to favour their own products. They may do so by manipulating the search function to ensure that their products are listed at the top of the search ranking, thus drawing more users. Such conduct undermines platform neutrality which warrants the fair and non-discriminatory treatment of all business users on a digital platform. It also creates a conflict of interest between the dual roles of an enterprise as a platform as well as a business user.

Consumers use keywords on search engines, i.e. terms used to match advertisements with the search queries of consumers. Large digital enterprises may use non-transparent search algorithms and tend to exercise control over the search rankings in order to give preference to sponsored/their own products so as to reduce the contestability of products favoured more by consumers. They may also allow bidding on relevant keywords (which may even be registered trademarks) enabling advertisers to increase their consumer reach. This leads to dilution of the brand power of products and services originally being searched for by consumers and compels companies to spend considerably in order to protect their own intellectual property and outrank competitors on search pages.²⁴

E-commerce site owners, where they serve as both a marketplace and a competitor on that marketplace, may be motivated to leverage their control over the platform to favor their desired vendors or private-label products. Such firms act as information intermediaries to potentially steer consumers away from independent providers. Critics of self-preferencing (whether in favor of platform neutrality or unfiltered advertising) argue for the structural separation of platforms (that is, breaking them up into separate entities) or functional separation of the platform (such as requiring the platform business to be operated separately from the product sales business). In this paradigm, large companies with diversified products that operate both downstream and upstream would be regarded as public utilities to avoid conflicts of interest.

It is hard to argue with the rationale behind the prohibition on self-preferencing. However, the current legislative scheme of the Competition Act, 2002 has sufficed to look into this issue in a timely manner. Given the definitions provided for “SSDEs” under the DCB, it would also encompass many Indian companies (that are no longer startups), particularly in the field of online retail.²⁵ The ICRIER-MeitY report also highlights this by acknowledging that “for the retail sector, most companies are opting for an omni-channel model, of which digital is a steadily growing part. In fact, many companies are developing their own platforms in addition to selling on intermediary platforms. An interesting trend is that e-tailers and e-sellers such as Nykaa, Urban Ladder, and FirstCry that began as completely online models have, over time, invested in physical stores. While the share of online selling is growing, offline formats will continue to exist and grow for some companies.”²⁶ Accordingly, it should be ensured that the prohibition against self-preferencing is applied to all firms, irrespective of their country of origin, in line with the principle of competitive neutrality. After all, “India also has its own tribe of domestic technology-driven businesses that operate using the same playbook of data aggregation, cross-sectoral linkages, acquisitions, and control. However, the general usage of the term big tech in the media and policy discourse in India is almost exclusively reserved for the foreign-owned multinational corporations described earlier.”²⁷

Furthermore, the ICRIER-MeitY Report also reveals specific numbers about the value created by digital intermediaries—approximately INR 46,506 crores (across a few sectors). On the other hand, the overall value created by firms dependent on such digital intermediary firms was higher—at INR 1,54,766 crores.²⁸ This highlights how, given the double-digit percentage commission rates charged by most digital intermediaries, SSDEs may be incentivized to ensure that third-party sellers do well—the higher the value created by dependent firms, the more the commission derived by the SSDEs.

Data Usage and Advertising Policies

Large digital enterprises have access to vast stores of user data which they use to innovate and improve their own products, thereby entrenching their position in the market. Network effects enable them to attract more users and generate and accumulate more data, thus creating a profitable feedback loop. The personal data collected by large digital enterprises may be used for profiling consumers and may also be sold to advertisers seeking to curate targeted online services and products, leading to concerns surrounding data privacy. This creates barriers to the entry of new players in digital markets which do not have access to these enormous repositories of data, thus distorting the level playing field for smaller digital enterprises and restricting competition in the market.

Large digital enterprises appear to be dominant at each stage of the advertising technology chain, i.e. demand, supply and exchange and their revenue models are primarily based on monetisation through advertisement revenue. There appears to be increasing market concentration, consolidation and integration across many levels in the ad-tech supply chain which gives the incumbent platform an unfair edge over the market. They may require advertisers to use ‘web crawlers’, i.e. programs used by search engines to collect and extract large amounts of consumer activity and data, on their websites which further allows them to provide targeted advertising. Large digital enterprises with dominant search engines or app stores reportedly also tend to prevent third-party app stores / apps from advertising on their platform citing vague and arbitrary concerns thereby limiting their consumer reach.²⁹

Section 12(2) of the DCB further elaborates on the above passages and states that SSDEs shall not, without consent, “intermix or cross use the personal data of end users or business users collected from different services including its Core Digital Service.”

To start off, this line of argument overlooks how a certain portion of end-users may be willing to part with their data to secure access to the various services of an SSDE. Furthermore, when it comes to targeted advertising as a result of profiling consumers, one of the problems often highlighted in the past has been that “it is unusual, to say the least, for a single company to represent both sellers and buyers in the same market”—a clear reference to Google’s presence across the ad-tech stack (explained below).³⁰ However, this presence arguably assists in giving a clearer picture of something called “attribution,” which is the clear linkage between the act of an end-user to visit a website of the advertiser or purchase the product of the advertiser and a response to the viewing of a particular digital ad placed by the advertiser. Google’s vertical integration helps its ability to accurately measure these metrics and accordingly arrive at the payment to the host of the ad, which would also be in the interest of transparency. Of course, if required, the host/publisher of the webpage where the ad is hosted can always request an audit regarding the performance of ads.

The ad-tech stack has “buy-side” tools used by advertisers to buy digital ads and “sell-side” tools used by publishers to sell digital ad space. The reason behind this is that “millions of businesses sought to advertise on the Internet, but they did not have anywhere close to the time or resources needed to negotiate individual contracts with each publisher.”³¹ Google has a presence across the buy-side tools and the sell-side tools in the ad-tech stack. It also owns a key ad exchange (AdX). These ad exchanges are seen as crucial to intermediating between the advertisers’ ads and the publishers’ webpage inventory by conducting real-time auctions, matching auction winners with the available publisher. Usually, “between $20 to $40 of every $100 spent on digital advertising goes to the ad tech companies that develop and operate these tools.”³²

One of the grievances against SSDEs consolidating their position across various stages of the ad-tech stack was that they had policies that allegedly prohibited web publishers using their sell-side tools (for their inventory) from setting higher price floors for the SSDE ad exchange as compared to other ad exchanges. This policy allegedly “increased the number of impressions AdX won and the revenue it received, while decreasing impressions won and revenue received by third-party exchanges.”³³ However, there are a few problems with this assertion, which focuses only on the web publishers that are keen on selling their inventory at as high a price as possible. But it should be noted that there are other players as well—advertisers and the consumers themselves. Each of them may benefit from the current arrangements. Relevant market delineation in antitrust markets is always challenging, but the trend is increasingly moving toward multi-sided markets (like the ad-tech market) requiring a holistic consideration of all sides of the market, on a case-by-case basis, before progressing to make an evaluation of any competitive harm.³⁴

When it comes to targeted advertising in general, however, it can be ensured, through the DCB, that in trying to arrive at an accurate profile of an end-user, SSDEs should not be able to make the use of their core digital service (CDS) or its certain functionalities by an end-user, as conditional upon the end-user’s consent to provide data across all services of the SSDE.

Deep Discounting

Large digital enterprises are more prone to employ predatory pricing tactics which involve setting prices below cost so that competitors can be excluded from the market, subsequent to which they raise prices to recoup their losses.³⁵

An interesting example of how deep discounting has been invoked can be seen in the advertising space. For instance, it has been observed by some in the ad-tech space that “Google’s profit margin is higher when it does not have to pay a publisher; marginal cost on its own properties is zero.”³⁶ Further, Google cross-subsidizes competitive functions of the ad-tech stack with monopolized functions in order to raise its rivals’ costs. Specifically, it has been argued that “another opportunity made possible by Google’s occupation of the entire ad tech stack is the ability to charge low prices at one end of the stack, to drive out competitors, while charging high prices at the other to counterbalance any losses.”³⁷ Incidentally, in India, both the CCI and the National Company Law Appellate Tribunal (NCLAT) cases saw a similar argument being made—against (i) margin squeeze and (ii) cross-subsidization.

The CCI case saw commission rates of 30 percent being charged from other apps but low charges levied on certain apps of its own like YouTube. It was noted by the CCI that “the monetization model of Google is based on cross-subsidization by Google where the 3% of the apps offering paid apps or IAPs (In-App Purchases) are made to bear the entire cost of the Play Store, even though all the apps are using similar services of the Play Store. Therefore, the question to be determined is whether it is reasonable and fair for these 3% of the apps to bear the 100% cost of the Play Store.”³⁸ On the other hand, one of the allegations (as highlighted during the NCLAT appeal) was that “Google has also been adopting discriminatory practices. Google is paying only fee of 2.35% to payment processor with regard to its app ‘YouTube’ whereas it imposes service fee of 15 to 30% on other apps. Discriminatory practices result in competitive disadvantages to the competitor of Google in downstream market by increasing their costs.”³⁹

In both cases, the CCI held that information available on record was not sufficient to give a finding on the monetization model, and therefore did not do so.⁴⁰ But what this showed was: (i) the cross-subsidization model in itself is not anticompetitive; (ii) detailed data is required to make out the claim of discriminatory pricing; and (iii) even though YouTube may not have been levied a commission by Google (since YouTube was free, and commission is predicated on a fee being charged by the app concerned), it did contribute in terms of advertising revenues, as pointed out by the CCI.

The above finding on the validity of the cross-subsidization model is apparently aligned with the CDCL report observation at the beginning of this section on recouping losses as well, that is, the monopolistic firm would only recoup losses in the market (in which they were incurred) later on. In this author’s opinion, cross-subsidization would not help. The very fact that there is cross-subsidization assumes that the so-called monopolist is not engaging in profit-maximizing behavior in other markets. In such a case, it would also be difficult to recoup losses from the monopolistic firm’s other profitable markets since, if one were a true monopolistic firm, they would already be maximizing profits in all their markets. Therefore, recouping losses incurred due to cross-subsidization, which would require such a monopolistic firm to further increase their prices in other profitable markets, would not be possible.

Exclusive Tie-Ups

Large digital enterprises may enter into exclusive arrangements with business users or sellers of products and services to not deal with other enterprises, thus denying them market access. Exclusive agreements can be classified into two kinds: agreements which exclusively launch a specific product on a platform or under which a platform provides only one particular brand in a specific product group.⁴¹

Exclusive tie-ups are largely seen as pro-competitive under the current Competition Act, 2002, where they are judged on a rule-of-reason basis, that is, after evaluating factors like potential efficiencies, consumers, and any risk of market foreclosure.

Exclusive tie-ups also eliminate the free-rider problem in many cases since they enable a manufacturer to prevent dealers from free-riding (on its promotional efforts) by aiming to incentivize them to focus only on its own brand. Also, in exclusive tie-ups, manufacturers will bolster their distributors because “other brands will not be able to take a free-ride on the supplier’s investment by selling through the same distributors.”⁴² Exclusive tie-ups can “stimulate suppliers to put more time and effort and money behind their channels of distribution, because . . . they do not have to worry about divided loyalties where they are wasting their effort.”⁴³

Accordingly, exclusive tie-ups should continue to be judged as per the rule of reason, as captured here:

Every agreement concerning trade, every regulation of trade, restrains. To bind, to restrain, is of their very essence. The true test of legality is whether the restraint imposed is such as merely regulates and perhaps thereby promotes competition or whether it is such as may suppress or even destroy competition. To determine that question, the court must ordinarily consider the facts peculiar to the business to which the restraint is applied, its condition before and after the restraint was imposed, the nature of the restraint, and its effect, actual or probable. The history of the restraint, the evil believed to exist, the reason for adopting the particular remedy, the purpose or end sought to be attained, are all relevant facts.⁴⁴

The DCB arguably does away with this.

Restricting Third-Party Applications

Large digital enterprises tend to prevent users from accessing or using third-party applications other than their own. They may do so by using exclusionary anti-steering policies (which have been discussed above) including curbing the installation of third-party applications.⁴⁵

The issue of restricting third-party applications, including third-party app stores, was the focal point of the CCI order concerning Google’s Play Store policies. Here, the sticking point was that other large players were not able to come up with their own app stores due to what was perceived as the anti-fragmentation agreement (AFA) clause. Under the AFA clause, the original equipment manufacturers (OEMs) were prohibited from taking any actions and/or allowing any third party to do so that may have resulted in the fragmentation of Android, that is, modified versions of the original version of Android.⁴⁶

However, this does not explain why OEMs and mobile phone manufacturers themselves were not able to open up their own app stores on their own manufactured devices. Yes, it is possible that they (i) viewed Google’s Play Store as a “must-have” app and didn’t want to “ask Google to exempt them from pre-installing Play Store,” and (ii) they did not pre-install any competing app store for fear that Google may see that as “fragmentation” and in breach of the AFA. A significant part of this could have been due to the term “fragmentation” not being defined under the AFA, which may have led to many OEMs not entertaining other app stores on their devices.

At the same time, the AFA was a regular contract like any other and would have been subject to a judicial interpretation in case any dispute arose between the OEMs and Google. After all, the AFA was not held by the CCI to a contract of adhesion. However, in some cases, the “problem” of restricting third-party applications was one of the placement of the app and not that of pure restriction per se. For instance, other app developers such as Mozilla were not content at even being pre-installed. As highlighted in the CCI’s Play Store order, the most that a majority of OEMs were “able to offer Mozilla was to pre-install Firefox as a secondary web browser option on the home screen” and not “as the default.”⁴⁷

Besides, sometimes there is substitutability between the app store in question and the other platforms that app developers could consider as alternatives. For instance, in Epic Games, Inc. v. Apple, Inc., the court found that “Epic presented no evidence regarding whether consumers unknowingly lock themselves into Apple’s app-distribution restrictions when they buy iOS devices. A natural experiment facilitated by Apple’s removal of Epic Games’ Fortnite from the App Store showed that iOS Fortnite users switched about 87% of their . . . iOS spending to other platforms—suggesting substitutionality between the App Store and other game-transaction platforms.”⁴⁸ This is to say that substitutionality between the app stores and other platforms ought to be looked into.

From the above ACPs, which have come up before the CCI, it appears that platforms may sometimes abuse market power, which may require targeted interventions. However, it should be deliberated whether those instances should be dealt with under competition law or under a new DCB altogether. Any market regulation must address “why,” “how,” and “when” collectively to ensure sensible outcomes. Over-regulation risks stifling innovation and returning to a pre-liberalization era regulatory mindset. Furthermore, ex ante laws should only be considered when overwhelming evidence shows that ex post measures are insufficient. As per this view, deflecting from enforcement priorities under the present Competition Act, 2002, and clamoring for the introduction of a new law is an easy way out. An ex ante law may not necessarily be directly ruled out for India. However, this needs to be done right and it should be ensured that there is a nuanced approach to regulation. For instance, examples are replete of the draft DCB allowing too much leeway to the CCI—Schedule I to the DCB currently allows the CCI to designate any service as online intermediation service, and for tailoring bespoke obligations for each such service. This may defeat the very purpose of ex ante regulation, which is not intended to operate in the manner of a case-by-case evaluation.

Absence of Legitimate Business Justification Defense

It is notable that during the recent NCLAT order on Google allegedly mandating its Google Play Billing System (GPBS), the NCLAT held that Google making the use of GPBS mandatory and exclusive for processing of payments for apps and in-app purchases was devoid of any “legitimate business interest.”⁴⁹ However, the same rationale—also called legitimate business justification (LBJ)—has not been provided under the DCB for any of the ACPs. Legitimate business justification explains that business efficiencies and antitrust law is not concerned with less efficient rivals, as discussed earlier in the bundling section. Furthermore, it has been seen innumerable times that the CCI protects the process of competition and not the competitors. Even in the commission’s case regarding the Google Play Store’s AFA and other agreements, Google had submitted various LBJs for its agreements, most of which were assessed by the CCI. The commission indeed felt that Google had a “legitimate interest in licensing its apps only for those devices which meet the minimum requirements set by it.”⁵⁰ However, it also deemed that the LBJ had to be proportionate.⁵¹ This kind of leeway, where the adjudicating authority can look at the business rationale offered behind a practice, is missing under the DCB.

Absence of “Less Restrictive Alternative” Defense

In Epic Games, Inc. v. Apple, Inc., Epic Games contended that Apple had acted unlawfully by “restricting app distribution on iOS devices to Apple’s App Store, requiring in-app purchases on iOS devices to use Apple’s in-app payment processor, and limiting the ability of app developers to communicate the availability of alternative payment options to iOS device users.”⁵² However, after a trial involving almost 900 exhibits, the district court rejected Epic’s claims, principally on the grounds that “Epic failed to propose viable less restrictive alternatives to Apple’s restrictions.” Even the American Innovation and Choice Online Act (which has been quoted several times in the CDCL report), mentions that “gatekeeper” firms have an affirmative defense if they can demonstrate that their conduct in question was narrowly tailored and “could not be achieved through less discriminatory means.”⁵³ This may be particularly relevant for practices like many of the ACPs, where there are both pro- and anti-competitive effects.

Conclusion

The draft Digital Competition Bill represents a bold and interventionist approach to regulating digital markets in India, premised on the assumption that certain large-scale practices by SSDEs are inherently anticompetitive. However, as this article has hopefully shown, many of the nine identified ACPs are neither per se harmful nor easily disentangled from legitimate business strategies that promote innovation, user experience, and efficient market outcomes.

While the CDCL report raises valid concerns, it does not adequately distinguish between harmful market foreclosure and vigorous competition, nor does it offer sufficient empirical grounding to support sweeping prohibitions. In fact, in multiple cases—bundling, deep discounting, and self-exclusive tie-ups—there are well-established global and domestic precedents that stress a contextual, effects-based, and evidence-driven analysis, and not blanket presumptions of harm.

Critically, the DCB’s omission of a legitimate business justification and less restrictive alternative defense removes essential safeguards against overreach and could inadvertently penalize efficiency-enhancing conduct. It shifts the burden unfairly onto SSDEs, many of which may include Indian firms, and risks replacing market discipline with regulatory discretion.

In this light, any forward-looking digital competition framework for India must balance the need for accountability with the need to preserve incentives for investment, differentiation, and user-centric design. Rather than sidestepping the Competition Act’s flexible and jurisprudence-rich architecture, the next legislative endeavor to regulate competition in digital markets should build upon the lacunae and gaps highlighted in this article—with clearer thresholds, limited presumptions, and defenses grounded in economic logic. After all, the aim of competition policy is not to punish size or success but to ensure that markets remain contestable, dynamic, and ultimately beneficial to consumers.

The author would like to thank an anonymous expert for their review of the manuscript.