Vertical restraints under Indian Competition Law: whither law and economics

The correct welfare assessment of vertical agreements in competition law is a difﬁcult craft. The more mature jurisdictions such as the EU and the US have struggled to develop the optimal framework. This article scrutinizes vertical agreements cases of the Competition Commission of India (CCI) that is now 11 years old. The objective is to assist the CCI in strengthening its legal and economic framework vis-a`-vis vertical agreements in its formative stage. The scrutiny of some leading cases revealed three types of inﬁrmities in CCI’s cases. First, there are some legal ambiguities in the interpretation of the legislation. Secondly, the economic analysis is largely incoherent and truncated. Thirdly, there is a visible overreliance on the EU jurisprudence that does not conform to the legislative scheme of the Indian Competition Act, 2002. The article draws out some lessons towards the end.


I. INTRODUCTION
In the marketplace, vertical agreements are a popular way to distribute products and services instead of direct sale or vertical integration. Vertical agreements also impose certain restraints on the parties. These restrictions are often pro-competitive as they a. tie-in arrangement; b. exclusive supply agreement; c. exclusive distribution agreement; d. refusal to deal; e. resale price maintenance, Shall be an agreement in contravention of sub-section (1) [sub-section (1) of Sec 3 sets out provision for prohibiting all anti-competitive agreements, which are rendered void by sub-section (2)] if such agreement causes or is likely to cause an appreciable adverse effect on competition in India.
In turn, section 19(3) of the Act enunciates upon Appreciable Adverse Effect on Competition [hereafter, AAEC] by noting that [t]he Commission shall, while determining whether an agreement has an appreciable adverse effect on competition under section 3, have due regard to all or any of the following factors, namely:a. creation of barriers to new entrants in the market; b. driving existing competitors out of the market; c. foreclosure of competition by hindering entry into the market; d. accrual of benefits to consumers; e. improvements in production or distribution of goods or provision of services; or f. promotion of technical, scientific and economic development by means of production or distribution of goods or provision of services.
While for horizontal agreements, section 3(3) sets out an exhaustive list, section 3(4) lays out an inclusive list of vertical agreements that may have an AAEC. 6 A look at the factors mentioned in section 19(3) makes it clear that clauses (a)-(c) are the 'negative factors' that restrict the competitive process; whereas, clauses (d)-(f) are the 'positive factors' that enhance the efficiency of the distribution process. 7 The CCI notes that 'whether an agreement restricts the competitive process is always an analysis of the balance between the positive and the negative factors listed under section 19(a)-(f)'. 8 It can be seen that section 19(3) also provides discretion to the CCI, in that in its assessment the CCI is free to account for 'all or any of the . . . factors'. This discretion is economically unjustified. A net effect on competition can be assessed only by taking all the factors (both positive and negative) into account. The CCI is mindful of this and hence in practice takes account of all the factors together. 9 In the case of horizontal agreements, section 3(3) raises a presumption of AAEC. 10 No such presumption of AAEC is raised in the case of agreements falling under section 3(4), and the burden of proof lies on the CCI. 11 The CCI has also made it clear that with respect to vertical agreements, it follows the rule of reason standard, 12 which is the standard that the EU and the US also follow for vertical agreements. 13 Here it is also relevant to distinguish the legislative scheme of sections 3 and 19(3) of the Indian Act from Article101 of the Treaty on the Functioning of the European Union (TFEU) as, in general, Indian Competition Act is modelled on the lines of the European competition law framework. 14 Article 101TFEU that provides a framework for competitive scrutiny of agreements bifurcates the restriction of competition and efficiency defence into Articles 101(1) and 101(3) of the TFEU, respectively. This bifurcation has been criticized by some commentators, as the division of labour between these two sub-sections has not been clear. 15 The Indian legal framework proscribes all such vertical agreements that have a net AAEC determined against the criteria mentioned in section 19(3) (a-f) of the Act. This way, the legislative structure of section 3 follows section 1 of the Sherman Act 1890, where anti-and pro-competitive elements are weighed at once to determine the net effect on welfare. 16 Another point of departure from Article 101(1) TFEU is that agreements in India do not have to be put into the 'by object' 17 or the 'by effect' 18 infringement categories. Once again, the identification of 'by object' and 'by effect' cases has been a source of constant controversy. 19 As per the India Act, the entire assessment is 10  III. LEGAL ANALYSIS-SOME AMBIGUITIES This part highlights certain ambiguities in the legal analysis in the CCI's decisions. Admittedly, it is not possible to draw clear boundaries between legal and economic analysis in competition law. Almost always, economic principles inform legal provisions. For the sake of clarity, however, the analysis in this part discusses instances where the interpretation of legislative provisions gave rise to ambiguities. These ambiguities may give rise to enforcement problems.
Are there anticompetitive agreements that are neither horizontal nor vertical? The CCI in Mr Ramakant Kini and Dr L.H. Hiranandani Hospital 21 held that section 3(1) is independent of sections 3(3) and 3(4) of the Act. This implies that there can be anti-competitive agreements that are neither horizontal nor vertical. This case was centred on deciding the legality of an exclusive agreement between the Opposite Party (OP), 22 a super speciality hospital, and Cryobanks International India (Cryobanks), which was the provider of stem cord banking facility. As per this agreement, only Cryobanks was allowed to enter the premises of the OP and collect stem cord right after the birth of a child. The complainant (or informant as used in the decisions of CCI), who had availed the maternity services at the hospital, had entered into an agreement with another provider of stem cord collection facility, namely Life Cell. On being denied to let Life Cell collect stem cord, the complainant brought a case under sections 3(4), and 4(2)(a)(i) and 4(2)(c) of the Act.
Instead of treating the agreement between the OP party and Cryobanks as a vertical agreement, the majority decision of the CCI suggested that such an agreement was neither a vertical nor a horizontal arrangement between the parties. The CCI developed its reasoning to argue that horizontal and vertical agreements are merely a sub-species of anticompetitive agreements under section 3(1) of the Act and are in no way exhaustive of the scope of this section. 23 Thus, it carved out a third category that the principles that should be followed to determine whether a practice is restrictive by object or by effect are clearer than commonly conceded by commentators'. 21. of agreements that do not fall in either of the boxes. It is contentious if the legislative scheme of section 3 permits this interpretation. 24 An agreement that is neither horizontal nor vertical, may pose challenge with respect to the burden of proof. As noted above, while with respect to horizontal agreements, the Indian Act raises a presumption of illegality, for vertical agreements the CCI in its decisional practice resorts to the rule of reason analysis where it discharges the initial burden of proof. In case an agreement that is neither horizontal nor vertical, but still falls within the scope of section 3(1), it is not clear as to who will discharge the initial burden of proof. Thus, the appropriate way to incorporate such agreements within the meaning of section 3(1) is by way of a legislative amendment that could identify the initial burden of proof with respect to this category of agreements.
The recent Competition Law Review Committee (CLRC) report, citing the Hiranandani case, advocated inclusion of 'an express provision in Section 3 of the Act to comprehensively cover all kinds of anti-competitive agreements that may not strictly fall within the categorisation of either a horizontal or a vertical agreement'. 25 The Draft Competition (Amendment) Bill, 2020 also proposes an amendment to section 3(4) that would capture '[a]ny other agreement amongst enterprises or persons including but not restricted to agreements amongst enterprises or persons at different stages'. 26 As the amendment has proposed amendment in section 3(4)provision related to vertical agreements-it can be assumed that the CCI will discharge the initial burden of proof with respect to this category of agreements. Remarkably, however, as there are no international precedents on this category of agreements, it is difficult to say if the initial discharge of burden of proof by the antitrust agency is the optimal solution in all cases related to the third category of agreements. Arguably, there can be such agreements, even within the third category, that can be presumed to be illegal based on their adverse effect on welfare. Different degrees of market power for sections 3(4) and 4 cases The Indian legislature has chosen to provide a separate category for concerted tying between the manufacturers of different products. Explanation (a) to section 3(4) of the Act states that a 'tie-in arrangement' includes any agreement requiring a purchaser of goods, as a condition of such purchase, to purchase some other goods.
This definition, however, does not make it clear that contrary to a unilateral tying, that may constitute abuse of dominance, a concerted tie-in arrangement takes place between two different parties who manufacture different products. For instance, manufacturers A and B agree to sell their products X and Y, respectively together. As the products are not identical or similar, this arrangement is logically seen as a vertical agreement. There are limited international precedents on making vertical tying a separate provision in the legislation. In the US, courts have not distinguished as such between unilateral and concerted tying. 28 In Australia, such agreements where a customer is required to buy two separate products, provided by two different suppliers, together are termed as third-line-forcing agreements. 29 It is important to distinguish tying as vertical agreement and tying as abuse of dominant position. The distinguishing feature between two, aside from the presence of an agreement in the former, is the different level of market power required for these two offences. It is true that even with respect to vertical agreements, the upstream and downstream firms should have some degree of market power for the agreement to have anti-competitive effects. Only when both firms have significant market power can the agreement result in anti-competitive effects in the distribution chain. It is for this reason that the EU Vertical Block Exemption Regulation (VBER) requires the downstream firm as well to have a market share of at least 30 per cent. 30 When the market power reaches the threshold of dominant position, and the upstream firm acts unilaterally, this practice is assessed as abuse of dominance. In the US, for a case of tying as a vertical agreement under section 1 of the Sherman Act, a plaintiff does not need to prove the type of market power that is required for a section 2 monopolization or attempted monopolization claim. 31 In one decision, however, the CCI treated the market power requirement under vertical agreements same as under abuse of dominant provision, arguing that the allegation of vertical tying and exclusive supply agreement could not be sustained as the Informant had the possibility to switch to vendors other than the OP. 32 This reasoning is better suited to assess dominance where the market power reaches the threshold where switching to competitors is not easy. 33 In a vertical agreement, unless one of the parties is dominant and acts unilaterally, it is possible, in principle, for the other party to switch to other vendors. If the same degree of market power is required to trigger the application 28  Vertical restraints under Indian Competition Law 7 of vertical agreements, a large number of anti-competitive vertical agreements will escape the competitive scrutiny, which will result in loss to consumer welfare.

IV. ECONOMIC ANALYSIS-DEFICIENT AND INCOHERENT
This section investigates the strength of economic analysis in some vertical restraint cases. These cases were decided under section 27 of the Act, where the OP was found to have entered into an anti-competitive agreement, and thus required deeper economic analysis before imposing liability. In particular, the following analysis looks at the relevant market definition, theory of harm, assessment of foreclosure, and assessment of countervailing efficiencies (factors in section 19(3)(d)-(f)). Taken together, this framework can help determine any harm to consumer welfare.

Mr Ramakant Kini and Dr L.H. Hiranandani Hospital, Powai, Mumbai
(2014) 34 The Informant, in this case, alleged that the OP, that was a multi-speciality hospital, restricted a third-party firm from collecting stem cord after the birth of babies. As per the Informant, the hospital had an exclusive agreement with a firm named Cryobanks. 35 This is one of the early cases where the CCI adopted a form-based approach. The CCI did not specify the nature of the agreement between the hospital and Cryobanks, and agreed that there can be anticompetitive agreements that are neither horizontal nor vertical. 36 The CCI assumed foreclosure only based on the exclusive nature of the agreement, instead of determining whether it led to actual anticompetitive foreclosure in the relevant market. 37 There is no clear theory of harm. A coherent theory of harm is missing also for the reason that for the CCI, this type of arrangement was neither vertical nor horizontal.
While the Director General (DG) had calculated the market shares of the hospital, Cryobanks and its competitors, the CCI did not calculate the percentage of the market that was foreclosed. Instead, the CCI noted that the hospital's choice of Cryobanks was neither based on a competitive criterion nor was it a result of quality consideration. Rather, this choice was completely based on the amount of commission paid by Cryobanks to the OP. 38 This finding played a major role in holding the agreement anti-competitive.
The CCI noted that for the purpose of analysing anticompetitive agreements under section 3, there is no requirement to identify the relevant market and the CCI is only required to 'see if the agreement has anti-competitive effects in any market and this market may be the market of the product/service of any party entering into the agreement'. 39 In future cases, the CCI corrected this thinking. The facts of the case suggest that the arrangement between the hospital and Cryobanks can be seen as an exclusive dealing agreement (or as the Indian Competition Act puts it as exclusive supply agreement as per section 3(4)(b) (Fig. 1). Indeed, one dissenting member analysed the facts within the framework of exclusive supply agreements. 40 It was also possible to see this arrangement as a vertical tying (section 3(4)(a) of the Act) between the hospital and the stem cell collection firm. Admittedly, the determination of these cases either as a vertical tie-in arrangement (under section 3(4) (a)) or as 'exclusive supply agreement' (under section 3(4) (b)), will not drastically alter the theory of harm and the subsequent rule of reason analysis under section 19(3) of the Act. Indeed, the EU Guidelines on vertical restraints mention that '[t]ying may also constitute a vertical restraint falling under Article 101 where it results in a single branding type of obligation (see paragraphs 129 to 150) for the tied product'. 41 If one sees stem cord collection as a process, then stem cell banks and maternity hospitals fall at different levels of the value chain. For the collection of the umbilical cord within the first 10 minutes of the childbirth, the stem cell bank has to have access to the premises of the hospital. 42 In somewhat similar facts, where a hospital had an exclusive agreement with a firm of anaesthesiologists, Justice O'Connor in his concurring opinion in Jefferson Parish analysed the case as exclusive dealing. 43 The DG delineated the relevant product market as 'provision of maternity services by Super Speciality Hospitals within a distance of 0-12 km from the Hiranandani Hospital . . . '. 44 In this market, the DG noted that the OP had a market share of 62 per cent. 45 The upstream market, in this case, is the market for the 'collection of stem cell cord at hospital premises'. Furthermore, in its supplementary report, the DG submitted that the market share of Cryobanks in Mumbai was 34.54 per cent in 2011-12. Cryobanks and its competitor Life Cell together held 67 per cent of the market share. 46 The DG in its investigation also found that Cryobanks had exclusive tie-up with other hospitals as well. 47 In this case, it was likely that a substantial part of the downstream market had been foreclosed to competing suppliers and potential suppliers, as more than 62 per cent of the downstream market that had high entry barriers had been tied to Cryobanks. Although the exclusive agreement between Cryobanks and Hiranandani Hospital was for one year, it was renewable at the option of the parties. 48 Even though, the hospital offered justification for this exclusivity, in principle, the cumulative effect of such agreements needed to be analysed in order to assess effects on consumers. 49 The CCI summarily denied any efficiency gains as untenable. 50

Ghanshyam Dass Vij v Bajaj Corp Ltd & Ors. (2015) 51
The feeble economic analysis continued in this case as well that the CCI decided soon after the Hiranandani case. In this case, the Informant, who was a distributor of Ayurvedic and general health products of various companies, made the allegation against the OP, a Fast-moving Consumer Goods (FMCG) company, that the latter imposed vertical restrictions on the sale of hair oil. The DG report found the violation of sections 3(4)(c), 3(4)(d), and 3(4) (e) (exclusive distribution agreement, refusal to supply and resale price maintenance).
The decision by the CCI appears to be confused so far as the existence of vertical restraints is concerned. While the decision, in the beginning, agreed to the presence of such practices and thereafter embarked upon the analysis, towards the end of the decision the CCI raised doubts over the actual implementation of the alleged anticompetitive agreements. 52 Equally troublesome is the economic analysis. While, in principle, the CCI recognized that vertical agreements needed to be prohibited only if they have appreciable adverse effects on competition, the actual analysis did not do justice to this assertion. Instead of defining the relevant market and calculating market shares both on the supply and the distribution level, the CCI merely noted that the opposite party was one among many hair oil suppliers on the market and hence did not have the position of strength in comparison with other brands. 53 Even though 'other equally and better brands' 54 were present, they might not have been in the same relevant market as the product of the OP. this case that the parties should have a significant market position in order to cause appreciable adverse effect on competition. Analytically, they should have substantiated it by defining the relevant market, calculating the market shares of the parties and identifying entry barriers; this, however, was not done. The CCI did not have the occasion to advance to the stage of efficiency analysis as no harm to competition was found in the first place.

Shri Shamsher Kataria and Honda Siel Cars India Ltd. and Ors. (2014) 55
This case remains one of the most challenging cases before the CCI. This case is remarkable also for the reason that the CCI borrowed extensively from EU competition law. In this case, the Informant had alleged abuse of dominance in the spare parts 'aftermarket' by some car manufacturers that also acted as Original Equipment Manufacturers (OEMs) for spare parts and other tools for repair works. It was also alleged that the OEMs had entered into anti-competitive exclusive distribution agreement and selective distribution agreement with Original Equipment Suppliers (OESs) and authorised dealers, respectively. This article only deals with the latter allegation.
The exclusive distribution agreement between the OEMs and OESs (under section 3(4)(c) & (d)) restrained the supply of spare parts by the OESs to the aftermarket. As a result, it was alleged that the supply of spare parts in the aftermarket was adversely affected. Through another set of selective distribution agreements, the OEMs restricted the sale and supply of the genuine spare parts, diagnostic tools/ equipment, technical information required to maintain automobiles to only the authorized dealers in the downstream market (Fig. 2). As a result, independent repair workshops could not compete with the authorized dealers in the aftersales automobile maintenance services.
There is a detailed analysis of relevant markets in this case. The CCI determined that spare parts formed a separate secondary market from the 'manufacture and sale of cars' market, which was the primary product. Contrary to the argument by the car manufacturers that cars and spare parts were together 'systems market', the CCI provided lengthy reasoning to justify the existence of separate 'aftermarkets' for 'sale

Vertical restraints under Indian Competition Law 11
of spare parts' and 'repair and maintenance services' in the context of this case. 56 Not only did the CCI rely upon the EU and the US jurisprudence, it also cited academic literature to form its opinion. 57 Furthermore, the CCI held that due to technical specifications and the complex engineering used in each brand, spare parts of one brand are not compatible with another brand. Neither are these spare parts interchangeable with the products in the primary market. Thus, spare parts of each brand constituted a distinct relevant product market. 58 Dexterously, the CCI also refused the contention that the primary market for 'manufacture and sale of cars' needed to be further divided in car segments, arguing that the focus of enquiry in the facts of the case was the 'aftermarket' and thus a further division of the primary market bore no relevance. 59 Similarly, the CCI also provided cogent reasons as to why spare parts constitute a 'cluster market'. 60 In the spare parts aftermarket, each OEM is the sole supplier for the supply of spare parts and diagnostic tools for their own brand of automobiles, because the spare parts of one OEM are not substitutable with that of others, 61 prohibition on OESs from supplying spare parts directly in the aftermarket, 62 and the requirement that authorized dealers should source all the spare parts from OEMs themselves. 63 The CCI gave much consideration to determine AAEC while analysing both the agreements. This meticulous exercise to determine AAEC had been hitherto absent in the previous cases. The CCI comprehensively looked at how these two sets of agreements led to complete foreclosure in the secondary markets and were not justified by the claimed efficiencies.
Conspicuously, however, any mention of market share both in the upstream and in the downstream markets is missing. With respect to both sets of agreements, even without determining foreclosure, the CCI dealt with the proffered justifications for exclusivity. 64 This, however, can be justified in the specific context of the case. It is easy to imagine that in the 'aftermarket' each OEM is a monopoly. Furthermore, this arrangement deprived independent repairers of the opportunity to compete with the authorized dealers in the car repair market.
In Re: Fx Enterprise Solutions India Pvt. Ltd. and Hyundai Motor India Limited (2017) 65 In this case the CCI was invited to scrutinize Resale Price Maintenance (RPM) and vertical tying. The Informants that were the car dealers for Hyundai brand of cars (OP) alleged the contravention of section 3(4) inasmuch as the OP had imposed exclusive supply agreement/refusal to deal, resale price maintenance and tie-in arrangement for Compressed Natural Gas (CNG) kits, lubricants and car insurance policies.
To assess the effects of the vertical restraints imposed by the OP, the CCI defined the upstream market as the 'market for all passenger cars'. 66 As the upstream market had several car manufacturers, and the subject matter of the inquiry was vertical restraints practised by one of them, the correct market definition should have been the 'market for Hyundai passenger cars'. As a result, the CCI did not identify the market position of the OP. The CCI only observed that the OP was a 'significant player'. 67 The following discussion would show that it was a vital omission on the part of the CCI, as the economic justifications for RPM cannot be accounted for if the market position of the firm is not identified.
The CCI rejected the existence of an exclusive supply agreement based on the evidence that the OP did not restrict any dealer from taking a competitor's dealership. In the absence of an exclusive supply agreement, the extent of economic analysis that the CCI would have undertaken is not clear. However, any balancing of positive effects of an exclusive supply agreement that the CCI cursorily mentioned in paragraph 75 against negative effects depended upon the extent of foreclosure, which would have been difficult to identify in the absence of market share calculation on both the upstream and the downstream levels.
The next allegation, in this case, was that the OP had practised RPM by prescribing the maximum discount that the dealers could give on a prescribed maximum selling price. Thus, effectively, a minimum resale price was mandated. The OP had also engaged the service of an independent firm to monitor these discounts.
The CCI in its abridged economic analysis looked mostly at the adverse effects of RPM on inter-and intra-brand competition. 68 Furthermore, the OP argued that the discount control mechanism was put in place at the behest of the dealers. The CCI found that, in effect, this could mean a collusive outcome at the level of the distributors. 69 The CCI summarily rejected positive effects of RPM without discussing any evidence in the specific context of the case. 70 Considering that the Indian car manufacturing market is highly competitive, 71 the OP might have intended to stifle intra-brand price competition to strengthen its position in the market. In the peculiar context of the automobile sector, pre-sale services effectively influence consumer choice. To this end, RPM can be an effective means to avoid the 'free-rider' problem. 72  the short run would have increased inter-brand competition, thereby increasing welfare. Hence, pro-competitive reasons to practise RPM in the specific context of the case could not have been brushed aside without a thorough investigation, especially if the OP did not have a strong market position and intended to strengthen the same by practising RPM. A final determination of AAEC, however, could have been done only by assessing the market position of the OP and its competitors-a critical analytical step that the CCI did not take. It is true that in the EU, RPM is treated as a 'hardcore' restraint that does not have the protection of VBER. At the same time, however, a presumption against RPM is rebuttable under Article 101(3). Importantly the US Supreme Court, since 2007, prefers a rule of reason analysis for minimum RPM. 73 There are two reasons for a more convincing fuller economic analysis of RPM in the Indian context. First, in a departure from the EU approach, economic theory treats RPM more leniently, if not as totally pro-competitive. 74 The EU formalism vis-à-vis RPM has not restricted the Member states from adopting a more lenient approach. In an Austrian case, the Vienna Regional Court and the Supreme Court accepted that RPM imposed by two German publishers in the retail channel passed the test in Article 101(3) in the specific context of the case. 75 In this case, the RPM was found to be necessary to ensure that the retail outlets could return unsold papers or magazines without any charges, which in turn ensured the diversity and availability of titles.
The second reason to adopt a fuller economic analysis vis-à-vis RPM is that the legislative framework of sections 3(4) and 19(3) makes no presumption about the adverse effects of RPM. Additionally, there are no 'per se' cases mentioned in the legislation. The CCI itself has noted in another case that all vertical agreements in India are to be analysed through a rule of reason framework. 76 The CCI is required to assess foreclosure as per section 19(3)(a-c). To that end, factors mentioned in paragraph 111 of the EU Vertical agreement Guidelines should illuminate the scrutiny. In the EU, Article 101 TFEU is set in motion only if an agreement is likely to have an appreciable effect on competition. 77 'Appreciability' features twice in the context of anti-competitive agreements: first, as a filter in the 'de-minimis' notice 78 and block exemption requirement, 79 where if the market position of the parties is below a particular threshold they are presumed to fall outside Article 101; second, at the stage of 101(1) where the agreement should 'contribute significantly to a foreclosure of the market' 80 in order to be violative of Article 101(1).
In the absence of any 'de-minimis' or 'bloc exemption' filter, appreciability in India needs to be determined only at the stage of section 19(3) (a-c) analysis. For this reason, therefore, assessment of the market position of the undertakings (ie market share, entry barriers), including that of the competitors is critical. If the impugned agreement significantly forecloses competition then the scrutiny should advance to the stage of section 19(3)(d-f), to assess AAEC on aggregate. The finding of 'significant foreclosure' in this way would inform the 'appreciability' requirement of the test mentioned in the scheme of sections 3(4) and 19(3).
The rule of reason analysis in the US too requires extensive scrutiny. Vertical restraints under Indian Competition Law 15 assumed significant foreclosure and looked at the objective justifications offered, rather than first looking at the quantum of foreclosure in the tied market. 82 In this case, the theory of harm would have been the same as one resulting from exclusive supply agreement, ie a possibility of 'single branding'. 83 Once again, this assessment required determining the overall market condition such as market shares of the parties involved in tying, market share of the competitors, the cumulative effect of any such similar agreements, entry barriers etc. If the parties had low market shares, a significant part of the market would still have been available to the competitors in the tied product market.
In a subsequent case decided in 2019, the CCI observed that even though some other jurisdictions treat RPM as a 'hardcore' restraint, in India RPM is assessed under the rule of reason analysis. 84 Remarkably, the CCI observed that the rule of reason analysis of vertical agreements is 'in sync with the existing economic literature', 85 and went on to set out both positive and negative effects of RPM. Even though, based on the factual enquiry no RPM was found, the DG in this case had also concluded that the OP did not possess sufficient market power to cause AAEC under section 19(3) as while in the upstream markets the OP had 19.7 per cent in Chimney and 25.2 per cent in Hob markets, the downstream market had several dealers. 86 As no RPM was established on facts, the CCI did not have the occasion to discuss the market position of the parties. Therefore, it is not clear if the CCI would look at the market power aspect while dealing with RPM in future.

V. OVERRELIANCE ON THE EU JURISPRUDENCE
Aside from the legal and economic infirmities in CCI's decisions discussed above, a remarkable infirmity is its overreliance on the EU jurisprudence. It is true that the CCI has benefited a lot by looking at the experience of more mature jurisdictions, an overreliance, however, has led to mechanical transplantations of certain EU concepts that fall outside the legislative mandate. For example, in the Shamsher Kataria 87 case, whereas on the one hand the CCI demonstrated its ability to engage in complex legal and economic analysis, on the other, this case inadvertently borrowed some EU concepts that the Indian Competition Act does not permit. In this case, the CCI while discussing efficiency defence under section 19(3) of the Act observed.
Article 101(3) (analogous to section 19(3) of the Act) provides that an agreement, containing restrictive clauses which 'contributes to improving the production or distribution of goods or to promoting technical or economic progress, while allowing consumers a fair share of the resulting benefit', will cause AAEC if such restrictive clauses 'afford such undertakings the possibility of eliminating competition in respect of a substantial part of the products in question.' Therefore, the agreement as a whole must not lead to the elimination of competition. The criterion of attempting to balance the efficiency gains and the foreclosure effects of vertical agreements is to reflect the view that short term efficiency gains must not be outweighed by longer-term losses stemming from the elimination of competition. 88 The CCI went on to completely follow the EU jurisprudence on 'eliminating competition in respect of a substantial part of the products in question' by observing the following.
Therefore, the Commission is of the opinion that in instances where an agreement, irrespective of the fact that it may contain certain efficiency enhancing provisions, allows an enterprise to completely eliminate competition in the market, and thereby become a dominant enterprise and indulge in abusive exclusionary behavior, the factors listed in section 19(3)(a)-(c) should be prioritized over the factors listed in section 19(3)(d)-(f). 89 undistorted competition on the common market'. 94 The CCI noted that if a dominant firm enters into a vertical agreement . . . then the Commission should give more priority to factors laid down under section 19(3) (a) to (c) than the pro-competitive factors stated under section 19(3) (d) to (f) of the Act, given the special responsibility of such firms not to impair genuine competition in the applicable market. 95 The 'special responsibility' of the dominant firm arises from the assumption that the competitive structure of the market has already been harmed due to 'dominance'. 96 Thus, once again, the focus is on the 'structure' of the market. This theory has been criticized as a reduced competitive constraint from the dominant firm's rivals is assumed, rather than analysed. Thus, in a way, competitive harm is 'taken for granted'. 97 Instead of relying upon the value-laden concept of 'special responsibility', it would have been better if the CCI had approached the factors mentioned in section 19(3) from the standpoint of economics. Any agreement that is entered into by an enterprise with significant market power may have more negative effects as compared to positive ones. The Article 102 Guidance paper also notes that '[i]n general, the stronger the dominant position, the higher the likelihood that conduct protecting that position leads to anticompetitive foreclosure'. 98 Another case where the CCI unjustifiably followed the EU jurisprudence is In Re: Fx Enterprise Solutions India Pvt. Ltd. and Hyundai Motor India Limited. 99 As discussed before, this case exhibits formalism where the CCI focused more on the negative aspects of RPM and chose to condemn the same without contextual economic analysis. This formalism is perhaps due to overreliance on the EU jurisprudence on RPM. The inclusion of RPM as a 'hardcore restraint' in the EU is a 'policy choice' that may not find support in economics. 100

VI. PROPOSED SOLUTIONS
The above analysis demonstrates certain infirmities in CCI's cases on vertical agreements. Based on the analysis, the following solutions are proposed.

Legal ambiguities
The legal ambiguities are primarily a matter of interpretation and do not emanate from the legislative scheme of the Act. The article has shown that carving out a third category of agreements through case law-that are neither horizontal nor vertical-may be problematic so far as discharging the initial burden of proof is concerned. Even if such a category needs to be devised to cast the enforcement net wide, the optimal solution is through the legislative route by way of amendment. The Draft Competition (Amendment) Bill, 2020 proposes carving out such a category within section 3(4). Thus, the responsibility of discharging the initial burden of proof will fall on the CCI. This Bill is yet to become the law, however. Additionally, only the time will tell whether the initial discharge of burden of proof by the antitrust agency in such agreements is the optimal solution.
The article also showed that vertical tying (section 3(4)(a)) and unilateral tying (section 4(2)(d)) are different not only with respect to the requirement of an agreement in the former, but also with respect to the degree of market power. The CCI in one case mistakenly treated the degree of market power required for vertical agreements same as required in unilateral tying cases by a dominant undertaking. This is conceptually wrong and will leave out from scrutiny anti-competitive vertical tying agreements entered into by a non-dominant undertaking.

Fuller economic analysis
The economic deficiency in the cases is worrisome. As the discussion in Section IV demonstrates, largely the analysis in these cases is truncated and incoherent. The rigour in defining relevant market shown in Shamsher Kataria 101 is missing in other cases. In the facts of the Shamsher Kataria, the CCI did not have to toil hard to show foreclosure in the downstream markets. However, in other cases where the CCI was required to show foreclosure, it did not do so convincingly.
The AAEC standard elaborated in section 19(3) suggests an effects-based analysis, as foreclosure (section 19(3) (a-c)) in itself is not anti-competitive if it can benefit consumers (section 19(3) d-f) on aggregate. Additionally, the rule of reason inquiry, which is the standard in vertical agreements, merits a full-fledged economic analysis. The CCI, therefore, needs a clear analytical framework in vertical agreement cases. As a solution to this, following an effects-based analysis through the rule-ofreason framework, the CCI should follow these steps-a clear determination of the market position (first define the relevant market and then assess market shares, entry barriers and any countervailing buyer power) of the firm and its competitors, a convincing theory of harm supported by evidence, and a proper balance between pro-and-anti-competitive effects to determine the net effect on welfare in the specific context of the particular relevant market.
As the discussion in Section IV has shown, the very first analytical step in this framework, ie determining the market position of parties that requires drawing a relevant market, determination of market shares and entry barriers is largely missing. In the absence of this, the determination of the degree of foreclosure becomes a presumptive exercise. In its case law, the CCI has proceeded ahead with the enquiry by merely stating that the parties had a 'substantial' market position. 102 Not much stress was laid on the requirement for delineating market definition in Hiranandani and Ghanshyam Das Vij. In Hiranandani the CCI was of the view that there was no need to define the relevant market while assessing anti-competitive agreements. This is not to say that the CCI has not evolved as a regulator. In contrast to these cases, the CCI delineated the relevant market in Shamsher Kataria and Hyundai cases based on the substitutability of products. The above discussion, however, shows that there is still scope for improvement in the economic analysis.
The cases also lack a 'counter-factual' enquiry. Drawing a counter-factual is a crucial step in determining anti-competitive effects. 103 It is apparent that the CCI has shown its willingness to take efficiencies (section 19(3)(d)-(f)) in determining the net AAEC. The efficiency defence, however, will be meaningful if the degree of foreclosure is assessed correctly.
As all vertical agreements need to be assessed through the rule of reason framework of section 19(3), the CCI would require 'safe harbour' rules in order to conserve its resources and facilitate ease of business. 104 At present, a clear discussion on the market share threshold is missing. 105 This will make devising any guidelines in the future a tough task.

Do not transplant the EU jurisprudence mechanically
The CCI should also be careful against overreliance on the EU case law. In Shamsher Kataria 106 and In Re: Fx Enterprise Solutions India Pvt. Ltd. and Hyundai Motor India Limited, 107 the CCI borrowed concepts that the Indian Act does not mandate. In the former case, the CCI equated section 19(3) of the Act with Article 101(3) of the TFEU. As shown in Section V, these two provisions have different mandates. Also, in this case the CCI borrowed the value-laden EU concept of 'special responsibility' of the dominant undertaking. Concepts such as 'special responsibility' and ensuring that an agreement does not eliminate competition 'in respect of a substantial part of the products in question' while carrying out efficiency defence under Article 101(3) TFEU, can be justified based on the objectives of EU competition law, which includes market integration and maintaining the market structure (and not just consumer welfare). 108 Arguably, the strictness with which the CCI treated RPM in In Re: Fx Enterprise Solutions India Pvt. Ltd. and Hyundai Motor India Limited also comes from the EU's harsh treatment of RPM. As has been shown above, mainstream economics (and now also the US) treats RPM more leniently.
As a solution, while the CCI should look at the international jurisprudence for guidance, reliance should be placed on mainstream economics to develop concepts that are in harmony with the underlying legislative mandate.

VII. CONCLUSION
The correct welfare assessment of vertical agreements in competition law is a difficult craft from the standpoint of law and economics. Intending to assess the level of analysis by the CCI in vertical restraint cases, this article scrutinized some leading cases on vertical agreements in India. The article looked at the EU and the US jurisprudence on the concerned issues, in view of their higher experience. The lessons drawn in this article are aimed at assisting the CCI in strengthening its analysis in these formative years.
The article began with a discussion on the legislative scheme of sections 3(4) and 19(3) of the Indian Competition Act that set out the legal framework for assessing vertical agreements. The legislative scheme is an improvement over the EU. The legislature has not divided the cases in the 'by object' and the 'by effect' categories. Neither is there a bifurcation between the assessment of restriction of competition and efficiency defence that could offset the harm. This way, the Indian legislature has learnt from the difficulties of interpretation in Article 101 TFEU. The subsequent inquiry into the legal and economic analysis of the cases revealed that the legislative 'leapfrogging' is not matched by the strength of analysis that the EU case law and soft law instruments have to offer.
There are some legal ambiguities in the cases. The CCI has suggested that there may be a third category of agreements that are neither horizontal nor vertical. Such an interpretation is doubtful in view of the legislative scheme of the Act. This interpretation also distorts the enforcement, as it is not clear who will discharge the burden of proof with respect to these agreements. Also, the CCI has mistakenly treated the market power requirement in vertical tying and unilateral tying as same. The CCI should amend its interpretation in future cases.
The article also analysed the strength of the economic analysis. The scrutiny revealed that there is no coherent analytical framework. Whereas the CCI claims to follow the rule of reason standard vis-à-vis vertical agreements, most cases display truncated economic analysis. A clear delineation of the relevant market, assessment 108  Vertical restraints under Indian Competition Law 21 of market power, quantum of foreclosure, and a proper balance between negative and positive effects of the restraint is missing. Finally, the scrutiny revealed that an overreliance on the EU jurisprudence has led to the borrowing of concepts that do not fit well with the legislative scheme of the Indian Act. Towards the end, the article suggested lessons to remove these infirmities in CCI's analysis.