Of all the conduct policed by competition law, none is treated with greater suspicion than the horizontal cartel — an agreement between rivals who, on paper, are supposed to be fighting each other for the customer's rupee. Section 3(3) of the Competition Act, 2002 singles out four species of such collusion — price-fixing, output restriction, market sharing and bid rigging — and attaches to them a statutory presumption that they cause an appreciable adverse effect on competition (AAEC). This presumption is the closest Indian law comes to the American "per se" rule, though, as the courts have repeatedly clarified, it is rebuttable rather than absolute. This chapter unpacks the architecture of Section 3(3), the meaning and limits of the presumption, the proof problems that surround tacit collusion, and the body of CCI and appellate jurisprudence — from the cement and tyre cartels to the Supreme Court's decisions in Excel Crop Care and Rajasthan Cylinders — that now defines this area. For the statutory groundwork, read this alongside our chapter on anti-competitive agreements and the Competition Act hub.
What Makes an Agreement "Horizontal"
The Competition Act draws a fundamental structural distinction between agreements among enterprises operating at the same level of the production or distribution chain — horizontal agreements — and those between enterprises at different levels, such as a manufacturer and its distributor — vertical agreements. The former are governed by Section 3(3); the latter by Section 3(4), which we examine in detail in our chapter on vertical agreements and the rule of reason. The distinction is not merely taxonomic: it dictates the burden of proof, the analytical standard and the very presumption that attaches to the conduct.
Section 3(3) applies to agreements "entered into between enterprises or associations of enterprises or persons or associations of persons or between any person and enterprise" engaged in identical or similar trade of goods or provision of services, including cartels. The phrase "identical or similar trade" is the textual hinge that identifies competitors. Two enterprises selling cement, two breweries selling beer, two tyre manufacturers supplying the replacement market — each pair stands in a horizontal relationship, and any coordination between them on the parameters Section 3(3) enumerates triggers the statutory presumption. The rationale is intuitive: when those who ought to be rivals agree instead to act in concert, the competitive process that is supposed to discipline price and stimulate output is short-circuited, and consumers bear the cost.
The Statutory Scheme: Section 3(1), 3(2) and 3(3)
Section 3 opens with a general prohibition. Section 3(1) provides that no enterprise or person shall enter into any agreement in respect of production, supply, distribution, storage, acquisition or control of goods or provision of services which causes or is likely to cause an appreciable adverse effect on competition within India. Section 3(2) renders any agreement in contravention of sub-section (1) void. These two provisions establish the overarching standard; every agreement, horizontal or vertical, must ultimately be tested against the AAEC yardstick.
Section 3(3) then carves out the four most pernicious categories of horizontal conduct and provides that any agreement of these kinds entered into by competing enterprises "shall be presumed to have an appreciable adverse effect on competition." The four limbs are: (a) directly or indirectly determining purchase or sale prices; (b) limiting or controlling production, supply, markets, technical development, investment or provision of services; (c) sharing the market or source of production or provision of services by way of allocation of geographical area, type of goods or services, or number of customers in the market or any other similar way; and (d) directly or indirectly resulting in bid rigging or collusive bidding. The word "presumed" is doing enormous work here, and understanding its precise force is the key to the whole subject — a matter we take up in the next section. For the building-block concepts of "enterprise," "relevant market" and "cartel," see our chapter on definitions.
The Presumption of AAEC and the "Per Se" Question
Indian commentators routinely describe Section 3(3) as embodying a "per se" rule, borrowing the vocabulary of United States antitrust law. The borrowing is illuminating but imprecise. In American jurisprudence the per se rule is conclusive: certain restraints, notably naked price-fixing, are deemed unlawful without any inquiry into their actual or likely effects. The doctrine traces to United States v. Trenton Potteries Co. (1927), where the Supreme Court held that uniform price-fixing by those controlling a substantial part of an industry is unlawful regardless of the reasonableness of the prices set, and was emphatically reaffirmed in United States v. Socony-Vacuum Oil Co., 310 U.S. 150 (1940), which declared that price-fixing agreements are unlawful per se under the Sherman Act, with no defence based on the competitive abuses the agreement was designed to cure.
Section 3(3), by contrast, creates a rebuttable presumption, not a conclusive one. Once the existence of one of the four enumerated agreements is established, the law presumes AAEC and shifts the evidentiary burden onto the parties to demonstrate that the agreement does not, in fact, cause an appreciable adverse effect. This was put beyond doubt by the Supreme Court in Rajasthan Cylinders and Containers Ltd. v. Union of India (2018), which held that the agreements mentioned in Section 3(3) raise a presumption of AAEC and that, by its nature, the presumption is rebuttable. The practical effect is a hybrid: India enjoys the prosecutorial economy of the per se approach — the regulator need not prove anti-competitive effect once the agreement is shown — while preserving a defensive opening for the accused that pure per se illegality would deny. The label "per se" is therefore best treated as shorthand for "presumption-with-reversed-burden" rather than as a literal transplant of American doctrine.
Defining a Cartel: Section 2(c)
The cartel is the archetypal horizontal agreement, and the Act defines it expressly. Section 2(c) provides that a "cartel" includes an association of producers, sellers, distributors, traders or service providers who, by agreement amongst themselves, limit, control or attempt to control the production, distribution, sale or price of, or trade in, goods or provision of services. Three features of this definition repay attention. First, the word "includes" makes the definition illustrative rather than exhaustive, allowing the CCI to capture novel forms of collusion. Second, the definition reaches an "attempt to control," so an unsuccessful cartel — one that fails to move prices — is no less a cartel in law. Third, the definition is medium-agnostic: it does not require a written contract, a formal association or any particular institutional form.
That last point matters because cartels are, by their nature, clandestine. As the cement cartel litigation made plain, the existence of written material is not necessary to establish a common understanding; it is enough if the activities of the enterprises imply the existence of an agreement. The Act reinforces this by defining "agreement" in Section 2(b) to include any arrangement or understanding or action in concert, "whether or not" such arrangement is in writing or intended to be enforceable by legal proceedings. The cartel is thus proved by inference far more often than by document.
Price-Fixing and Output Restriction: Clauses (a) and (b)
Clause (a) reaches agreements that "directly or indirectly" determine purchase or sale prices. The breadth of "indirectly" is deliberate: it captures not only naked agreements on a common price but also agreements on discounts, rebates, credit terms, surcharges, freight equalisation or any other component that feeds into the price a customer ultimately pays. The CCI's order in the zinc-carbon dry cell batteries matter illustrates how price coordination operates in practice — the manufacturers' top management met regularly to agree on price increases, typically led by one player and followed by others under the cover of "following the market leader," while simultaneously agreeing not to push sales aggressively so as to avoid a price war.
Clause (b) addresses agreements that limit or control production, supply, markets, technical development, investment or provision of services. Output restriction is the economic mirror-image of price-fixing: by curtailing supply, colluding firms achieve the same supra-competitive prices they could fix directly. The tyre cartel order against the members of the Automotive Tyre Manufacturers Association (ATMA) — Apollo Tyres, MRF, CEAT, JK Tyre and Birla Tyres — found contravention of both Section 3(3)(a) and 3(3)(b) read with Section 3(1): the manufacturers acted in concert both to raise prices of cross-ply/bias tyres in the replacement market and to limit and control production and supply, exchanging price-sensitive information through their association. The case is a textbook example of an industry association functioning as the conduit for collusion — a recurring pattern the CCI scrutinises closely.
Market Sharing: Clause (c)
Clause (c) prohibits agreements that share the market or source of production or provision of services, whether by allocation of geographical area, by type of goods or services, by number of customers, or in any other similar way. Market allocation is in some respects the most efficient form of cartel from the colluders' standpoint: instead of monitoring a common price, each firm is simply granted a protected sphere within which it faces no rivalry. The customer in an allocated territory loses the benefit of competition entirely, paying whatever the local "owner" of that territory chooses to charge.
Market-sharing frequently accompanies the other limbs. In the zinc-carbon battery cartel the CCI found not only price coordination but also an agreement to divide the geographical market and consumers among the participants — a finding that engaged clause (c) alongside clauses (a) and (b). Because the four limbs are not mutually exclusive, a single sophisticated cartel will typically contravene several at once, and the CCI's orders routinely record multiple simultaneous breaches read with Section 3(1).
The closing words of clause (c) — "or any other similar way" — are a deliberate residuary phrase that prevents colluders from escaping liability merely by inventing a novel basis of allocation. Whether rivals carve up the market by region, by product line, by customer category, by tender or by any functional equivalent, the conduct falls within the clause so long as its effect is to substitute a protected sphere for genuine rivalry. This drafting technique — pairing specific illustrations with a catch-all — recurs throughout Section 3 and reflects the legislature's awareness that anti-competitive ingenuity will always outpace any exhaustive list.
Bid Rigging and Collusive Bidding: Clause (d)
Clause (d) targets agreements that "directly or indirectly" result in bid rigging or collusive bidding. The Act supplies its own gloss: the Explanation to Section 3(3) defines "bid rigging" as any agreement between enterprises engaged in identical or similar production or trading of goods or provision of services which has the effect of eliminating or reducing competition for bids or adversely affecting or manipulating the process for bidding. Bid rigging is the cartel form most often encountered in public procurement, where it directly injures the public exchequer, and it takes familiar shapes: cover bidding (rivals submit deliberately high "complementary" bids to make a designated winner look competitive), bid suppression (a rival refrains from bidding), and bid rotation (the winner is pre-allocated in turn).
The leading authority is Excel Crop Care Ltd. v. Competition Commission of India, AIR 2017 SC 2734, reported as (2017) 8 SCC 47, decided on 8 May 2017 by Sikri and Ramana JJ. The case arose from a complaint by the Food Corporation of India that suppliers of aluminium phosphide tablets had rigged tenders between 2007 and 2009 by quoting identical prices. The Supreme Court upheld the finding of bid rigging and gave the Explanation a purposive reading, holding that the "process for bidding" covers every stage from the notice inviting tender right through to the award of the contract, including intermediate stages such as pre-bid clarifications. It also confirmed that where, during an investigation into one tender, the Director General uncovers evidence that the same parties rigged other tenders, those additional contraventions may properly be brought within the report — the investigation is not artificially confined to the four corners of the original reference.
Proving a Cartel: Parallelism, "Plus Factors" and Circumstantial Evidence
The central evidentiary difficulty in cartel enforcement is that direct proof — a signed agreement, a minute of a price-fixing meeting — is rare. Cartels are deliberately covert, and the regulator is usually left to reason from circumstantial evidence: parallel pricing, simultaneous and identical price changes, exchange of sensitive information, common agents, records of meetings and the like. Indian competition jurisprudence has, accordingly, accepted that a cartel may be inferred from conduct, provided the inference is the only reasonable one. The cement and battery orders both proceeded substantially on circumstantial proof.
But parallel conduct alone is a treacherous foundation, because it is also the natural behaviour of firms in an oligopoly. In a market of a few large sellers, each rationally watches and matches its rivals' prices without any agreement at all — a phenomenon economists call "conscious parallelism" or interdependent oligopolistic behaviour. To distinguish unlawful collusion from lawful interdependence, enforcers look for "plus factors" — additional circumstances that are inconsistent with independent action, such as artificially identical bids in a market where costs differ, evidence of inter-firm communication, a plausible motive to collude, or conduct against the participants' individual self-interest. The plus-factors framework is what allows the regulator to convert suspicious parallelism into a sustainable finding of agreement.
The standard of proof in cartel cases is the civil standard of the balance of probabilities rather than the criminal standard of proof beyond reasonable doubt, but the inference of agreement must nonetheless be the only reasonable one open on the evidence. Where the proven facts are equally consistent with independent oligopolistic behaviour as with collusion, the presumption under Section 3(3) is never triggered, because the threshold question — whether an "agreement" within Section 2(b) existed at all — has not been answered in the regulator's favour. The plus-factors inquiry therefore operates at the very front of the analysis: it is not a defence to a presumed cartel but the means by which the existence of the cartel is first established. Only once that hurdle is cleared does the burden shift to the enterprises to rebut AAEC.
The Limits of Inference: Rajasthan Cylinders
The most important judicial check on inference-based cartel findings is Rajasthan Cylinders and Containers Ltd. v. Union of India (2018), decided by the Supreme Court on 1 October 2018. The CCI and COMPAT had found dozens of manufacturers of 14.2 kg LPG cylinders guilty of bid rigging under Section 3(3)(d) in tenders floated by Indian Oil Corporation, relying chiefly on the near-identity of their quoted prices. The Supreme Court allowed the appeals and set aside the cartel finding. Its reasoning is essential to a balanced understanding of Section 3(3): the market was a highly concentrated oligopsony, with only three public-sector buyers dictating terms to a large number of sellers; in such a structure, parallel and even identical bids are exactly what one would expect from rational, independent suppliers responding to a dominant buyer's tender conditions. Identical pricing, the Court held, would not by itself lead to the conclusion that there was a concerted practice in the absence of other credible and corroborative evidence — that is, in the absence of plus factors.
The case is doubly significant. First, it vindicates the rebuttable nature of the Section 3(3) presumption: the suppliers were held to have discharged the burden of showing that the structural features of the market, rather than any agreement, explained the parallel bids. Second, it imports into Indian law the same caution that animates the American "conscious parallelism" doctrine and the European treatment of oligopolistic interdependence — namely, that competition law punishes agreements, not the mere economic fact of similar behaviour in a concentrated market. Read alongside Excel Crop Care, which upheld a bid-rigging finding on strong circumstantial proof, Rajasthan Cylinders marks the boundary at which inference must stop.
The Cement Cartel: The High-Water Mark of Enforcement
The single most consequential cartel proceeding in Indian competition history is the cement matter, Builders Association of India v. Cement Manufacturers' Association. Acting on the Builders Association's information, the CCI by its order dated 20 June 2012 held the Cement Manufacturers' Association and a group of leading producers — including UltraTech, ACC, Ambuja and others — guilty of contravening Section 3(3)(a) and 3(3)(b) read with Section 3(1), having coordinated prices and limited and controlled production and supply. The Commission imposed penalties of roughly Rs 6,300 crore, the largest cumulative cartel fine then levied in India, calculated at 0.5 times the net profit of the manufacturers for the relevant period, with a separate penalty on the association.
The litigation has had a long and instructive afterlife. COMPAT set aside the original 2012 order on natural-justice grounds and remanded the matter for fresh hearing; in 2016 the CCI re-decided the case, reaffirmed its findings and reimposed the penalties. The proceeding established several enduring propositions: that an industry association can itself be a cartel facilitator and a contravening party; that the existence of an agreement may be established from conduct and circumstantial material without a written document; and that the CCI is prepared to impose deterrent penalties at the upper end of its statutory powers. It also exposed the procedural fragility of CCI orders, since failures of natural justice have repeatedly led appellate fora to remand or set aside otherwise substantive findings.
Rebutting the Presumption: Section 19(3) Factors and the Joint Venture Proviso
Because the Section 3(3) presumption is rebuttable, the parties may seek to show that their agreement does not in fact cause an appreciable adverse effect on competition. The analytical lens for any AAEC inquiry is supplied by Section 19(3), which directs the Commission to have "due regard" to six factors — three that weigh toward a finding of AAEC and three that weigh against it. The harmful factors are: creation of barriers to new entrants; driving existing competitors out of the market; and foreclosure of competition by hindering entry. The beneficial factors are: accrual of benefits to consumers; improvements in production or distribution of goods or provision of services; and promotion of technical, scientific and economic development by means of production or distribution of goods or provision of services. A horizontal cartel of the naked variety will rarely satisfy the beneficial factors, which is precisely why the presumption so seldom yields in practice; but the statutory door is, formally, never closed.
Section 3(3) also contains an important carve-out. Its proviso states that nothing contained in the sub-section shall apply to any agreement entered into by way of joint ventures if such agreement increases efficiency in production, supply, distribution, storage, acquisition or control of goods or provision of services. The proviso recognises that not all cooperation between competitors is harmful — genuine joint ventures may pool resources, share risk and generate efficiencies that benefit consumers. The burden, however, lies on the parties to demonstrate the efficiency-enhancing character of the venture; the proviso is a defence to be earned, not a label to be claimed.
Consequences: Penalties under Section 27 and Leniency under Section 46
Where the CCI finds a contravention of Section 3, Section 27 empowers it to issue cease-and-desist directions, declare the agreement void, and impose monetary penalties. Section 27(b) permits a penalty of up to 10% of the average turnover of the preceding three financial years, with a heavier ceiling for cartels — up to three times the profit for each year of the continuance of the cartel, or 10% of turnover for each such year, whichever is higher. The basis on which that penalty is calculated was settled by Excel Crop Care, where the Supreme Court held that, at least for multi-product enterprises, the penalty must be computed on the relevant turnover — the turnover derived from the product or service affected by the contravention — and not on the enterprise's total turnover. The relevant-turnover principle injects proportionality into penalty-setting and has reshaped CCI practice across all subsequent cartel orders.
To crack the wall of secrecy that protects cartels, the Act offers an incentive to defectors. Section 46 empowers the CCI to impose a lesser penalty on a cartel member who makes a full, true and vital disclosure, the mechanics being governed by the Competition Commission of India (Lesser Penalty) Regulations, 2009. The first applicant to provide vital evidence enabling the CCI to form a prima facie opinion may receive up to a 100% reduction; the second applicant up to 50%; and subsequent applicants up to 30%. The regime is no longer theoretical: in the zinc-carbon dry cell batteries cartel (2018) the CCI granted Panasonic a 100% reduction as the first informant — its first ever full waiver — while Eveready and Indo National (Nippo) received reduced penalties. The beer cartel decision of 2021 against United Breweries, Carlsberg India and Anheuser-Busch InBev — where ABI received a 100% waiver as the first leniency applicant and the penalties were subsequently upheld by the NCLAT — confirms that leniency has become the CCI's principal detection tool, much as it is for the European Commission and the US Department of Justice.
Comparative Perspective and Exam Pointers
Indian cartel law sits in a recognisable comparative tradition. The American per se rule for price-fixing, born in Trenton Potteries and hardened in Socony-Vacuum, supplied the conceptual template for treating certain horizontal restraints as inherently suspect; European Union law under Article 101 TFEU treats price-fixing and market-sharing as restrictions "by object," dispensing with proof of effect in much the same spirit. India's Section 3(3) blends these influences but stops short of conclusive illegality, retaining the rebuttable presumption that Rajasthan Cylinders confirms. The distinction between this presumption and the genuine rule-of-reason analysis applied to vertical agreements under Section 3(4) is a perennial examination favourite — be ready to contrast the two and to explain why the burden of proof falls where it does.
For revision, anchor the topic on four pillars: the statutory text of Section 3(3)(a)–(d) and the Explanation defining bid rigging; the nature of the presumption as rebuttable rather than absolute; the proof problem and the parallelism/plus-factors framework illustrated by Excel Crop Care and Rajasthan Cylinders; and the enforcement architecture of Sections 27 and 46. Carry the comparison forward by reading our chapter on abuse of dominant position, which addresses the other principal head of substantive contravention under the Act, and revisit the overview of anti-competitive agreements to see how horizontal and vertical restraints fit within the single architecture of Section 3.
Frequently asked questions
Is Section 3(3) of the Competition Act a true "per se" rule like American antitrust law?
Not quite. Section 3(3) creates a rebuttable presumption that the four enumerated horizontal restraints cause an appreciable adverse effect on competition, shifting the burden onto the accused enterprises to prove otherwise. The American per se rule, by contrast, is conclusive — it admits no rebuttal once the restraint is shown. The Supreme Court in Rajasthan Cylinders and Containers Ltd. v. Union of India (2018) expressly held the Section 3(3) presumption to be rebuttable, so "per se" is best read as shorthand for a presumption-with-reversed-burden rather than a literal transplant of the doctrine in Socony-Vacuum.
What are the four types of horizontal agreement presumed to cause AAEC?
Section 3(3) lists: (a) directly or indirectly determining purchase or sale prices (price-fixing); (b) limiting or controlling production, supply, markets, technical development, investment or services (output restriction); (c) sharing the market or source of supply by allocation of geographical area, type of goods or services, or number of customers (market allocation); and (d) directly or indirectly resulting in bid rigging or collusive bidding. The Explanation to the sub-section separately defines bid rigging.
Does a cartel have to be proved by a written agreement?
No. Section 2(b) defines "agreement" to include any arrangement, understanding or action in concert, whether or not it is in writing or intended to be legally enforceable. The cement cartel litigation confirmed that written material is unnecessary — it is enough if the enterprises' activities imply a common understanding. Cartels are therefore usually proved by circumstantial evidence such as parallel pricing, exchange of sensitive information and records of meetings, supported by "plus factors" that exclude innocent explanations.
What did the Supreme Court decide in Excel Crop Care v. CCI?
Excel Crop Care Ltd. v. CCI, AIR 2017 SC 2734 / (2017) 8 SCC 47, decided on 8 May 2017, upheld a finding of bid rigging by suppliers of aluminium phosphide tablets to the Food Corporation of India. The Court held that the "process for bidding" in the Explanation to Section 3(3) covers every stage from the tender notice to the award of the contract, and that the Director General may bring additional contraventions discovered during an investigation within the report. Crucially, it also held that penalties for multi-product firms must be computed on relevant turnover, not total turnover.
Why was the cartel finding set aside in the LPG cylinder (Rajasthan Cylinders) case?
In Rajasthan Cylinders and Containers Ltd. v. Union of India (2018) the Supreme Court allowed the manufacturers' appeals and set aside the bid-rigging finding because the market was an oligopsony dominated by a few public-sector buyers. In such a structure, parallel and even identical bids are the natural outcome of rational, independent responses to a dominant buyer's tender, so identical pricing alone — without credible corroborative "plus factor" evidence of a concerted practice — could not establish a cartel under Section 3(3)(d). The case marks the outer limit of inference-based enforcement.
How does the leniency programme under Section 46 work?
Section 46, read with the CCI (Lesser Penalty) Regulations, 2009, lets a cartel member that makes a full and vital disclosure obtain a reduced penalty: up to 100% for the first applicant, up to 50% for the second, and up to 30% for subsequent applicants. The CCI granted its first ever 100% waiver to Panasonic in the zinc-carbon dry cell batteries cartel (2018), and ABI obtained a 100% waiver as first applicant in the beer cartel (2021), penalties in which were later upheld by the NCLAT. Leniency is now the CCI's primary tool for detecting otherwise covert cartels.