Loyalty rebates

A discount is said to be individualized when different rebate schemes are offered to each buyer. Under such schemes, a dominant firm offers different kinds of rebates to clients that demand a certain quantity or overcome another threshold. Typically, there will be a series of thresholds, and each threshold will go hand in hand with a discount on all units. The period during which the discount applies to all units may also vary. Conversely, a discount is considered standardized if the same scheme is offered to all buyers.

Another way to distinguish between discounts is to classify them as incremental (or marginal) versus retroactive. An incremental volume discount is a type of rebate in which lower rates apply only to units purchased above a specific threshold, rather than to the entire order. Thus, the discount applies to marginal quantities. It can be applied in brackets, for example: a 1% discount for units 1–10; 2% for units 11–20; 3% for units 21–30; and so on.

If, instead, once the threshold is met, the buyer receives the rebate on all units purchased within an agreed period, it is referred to as a retroactive discount (or all-units discount). This is the most common form of loyalty rebate, also referred to as conditional rebates.

Quantity-based discounts are rebate schemes that condition a discount on the customer’s purchases when these exceed a specific volume threshold within a defined period. For example, the price for the first 100 units may be set at $1,000 each, but if more than 100 units are purchased, all units will be priced at $800.

Market-share-based discounts represent another recurring type of conditional discount. In this case, the discount is based on the supplier’s share in the buyer’s total purchases. The discount is thus conditional upon purchasing the majority of the required inputs from the same provider. Notably, this scheme indirectly ties the discount to the amount the customer purchases from the supplier’s competitors.

Conditional discount practices clearly parallel exclusivity agreements. Both, obviously, share conditionality at their core. More specifically, some scholars have argued that loyalty discounts create incentives that may lead to de facto exclusivity agreements (Kaplow & Shapiro, 2007). In other words, while rebates are not identical to exclusivity agreements, they can produce equivalent economic effects, making it prima facie reasonable to consider them within the same category.

Nonetheless, there also are reasons to keep discounts distinct from exclusivity agreements (O’Donoghue & Padilla, 2020). The two differ in an important dimension: exclusivity agreements are long-term bilateral contracts involving the buyer’s commitment not to purchase from alternative suppliers during a specified reference period. This element of buyer commitment is not present in loyalty discounts, which are unilateral offers where only the supplier commits to offer different terms of trade depending on how much the buyer purchases. In effect, a buyer entering into an exclusivity agreement cannot purchase from another supplier, whereas in the case of loyalty rebates, a buyer may switch at any time to an alternative supplier—though they will, of course, lose the discount.

Economists have presented several procompetitive (efficiency) arguments explaining why companies, including dominant ones, offer loyalty discounts to clients, distributors, and other agents in order to increase their sales efforts (O’Donoghue & Padilla, 2020; Zenger, 2012).

First, some argue that this practice facilitates more efficient fixed-cost recovery. When production involves high fixed costs, prices are set above marginal costs to be high enough to recover those fixed costs; otherwise, production is not sustainable in the long term. The problem is that higher prices mean lower volume, which has adverse effects on consumer welfare. Price discrimination offers a solution by allowing relatively high prices for units with inelastic demand (the “secured portion” of sales) while simultaneously charging lower prices for units where demand is more elastic.

Compared to a uniform pricing scenario, applying discounts lowers prices for end consumers, as lower marginal prices are passed on by distributors. Thus, the manufacturer can benefit from a higher margin on inframarginal units without losing volume, achieving efficient price discrimination (Zenger, 2012).

Second, some argue that loyalty rebate schemes between manufacturers and retailers can benefit consumers by improving the incentives retailers face. These can solve various moral hazard issues in vertical relationships — that is, conflicts of interest between manufacturers and retailers concerning advertising, investment in professional salesforces, etc. For example, if retailers face a low marginal cost for a product, they are incentivized to promote or expand their sales through competitive pricing. These incentives may be weaker when the retailer’s additional gain is small.

Generally, it is difficult for a supplier to draft an efficient contract specifying the required effort from the retailer. Furthermore, even if the former were possible, it would still come with substantial costs to monitor and enforce the required effort, especially in the event of a dispute. The simpler solution is for the supplier to set an incentive scheme closely aligned with its interests.

Third, conditional discounts reduce the adverse effects of so-called double marginalization, a central problem in vertical relationships. When a supplier has market power, their wholesale price includes a markup. If the retailer also has market power, they treat the wholesale price as part of its cost structure and adds its own monopoly margin. This leads to double marginalization and higher final prices.

Both the supplier and the retailer could jointly increase their profits if retail prices were reduced. By charging nonlinear prices, the supplier can cover fixed costs and earn profits while mitigating or eliminating the supplier’s margin issue. This also improves consumer welfare by making retail prices lower than what they would be in a scenario of linear pricing. Retroactive discounts have been shown to be most effective at eliminating double marginalization (Kolay, Shaffer & Ordover, 2004).

Fourth, fidelity rebates can solve hold-up problems. For example, a manufacturer may be reluctant to invest in training a retailer’s salesforce if part of the knowledge transferred could be used to promote rival products instead. This underinvestment problem can be solved if retailers commit to concentrate their acquiring effort on the products of the manufacturer that trains their staff. Since such a commitment is difficult to guarantee ex ante, one option for the manufacturer is to offer a loyalty rebate, thus ensuring the retailer has an incentive to focus purchases on its products. This provides incentives for complementary investments from both suppliers and retailers, thereby increasing efficiency.

Under certain conditions, loyalty rebates can have anticompetitive effects (Zenger, 2012). Exclusionary effects may rise when a dominant company’s discount scheme is so aggressive that the buyer cannot afford to purchase from anyone else. In practice, the customer may choose not to hire a competitor to avoid the financial penalty that the dominant supplier would apply to the rest of their orders.

By applying retroactive conditional discounts, a dominant firm leverages the non-contestable portion of a client’s demand (i.e., the amount the client would purchase anyway from the dominant firm) to effectively subsidize the contestable portion (i.e., the portion the client could shift to a rival). As a result, market power is transferred from the non-contestable portion of demand to the contestable portion. Given economies of scale, this may prevent competitors from entering the market or competing on equal terms for the entire demand of an individual client with a competitive price.

For this exclusionary effect to occur, the Leverage Theory (Whinston, 1990) must apply. This theory establishes the conditions under which market power can be transferred from one market to another. Although, in this case, the transfer is from one segment of demand (non-contestable) to another (contestable), not to a different market, the Leverage Theory still applies.

In summary, it has been suggested that loyalty rebates can create significant switching cost problems for a dominant firm’s rivals. They may also be a cheaper form of exclusion than strategies like predatory pricing, as the dominant firm does not need to invest in loss-generating activities when engaging in loyalty rebates.

In real-world markets, the use of loyalty rebates is likely to have anticompetitive effects only when several cumulative features are present (OECD, 2003). The analysis in previous sections helps formulate a theory of harm.

First, the seller must be able to rely on a secure base of sales (also called the non-contestable portion of demand) to offer unbeatable prices for the purchase of the incremental units customers wish to buy from rivals. Normally, this means that the prices for the secured base are higher than the prices charged for incremental units which are above the relevant threshold. The discount then enjoys a base of secured sales — sales the customer would be reluctant to shift to a rival unless offered substantial price incentives. Without a secured base of sales with a particular customer, the pricing structure offered by a dominant firm would be irrelevant since the demand would be fully contestable.

This exercise involves determining whether the discounts aim to prevent a rival from reaching minimum efficient scale in the market or a market subset. The existence of a secured sales base may result from customer switching costs (learning, transaction costs, etc.), long-term contractual commitments, or capacity constraints. In practice, loyalty and target discounts will only be effective for firms with strong brands or those that are essential business partners for some reason.

Second, there must be a strong incumbent advantage, represented by a substantial asymmetry in access to a crucial infrastructure or input, or a preexisting customer base — and more generally, by a high and persistent market share. This is reinforced by switching costs, infrequent purchases, and demand externalities. Accordingly, establishing dominance is a prerequisite to identifying and eventually sanction exclusionary discounts, and the stronger the dominance, the greater the concern for anticompetitive effects. Conversely, where dominance is weak due to the presence of relatively strong rivals, there is greater reason to be skeptical that a discount scheme would have exclusionary effects.

Third, exclusionary price discrimination is more likely when demand is fragmented — that is, when there are many buyers whose individual demand is insufficient to support a new entrant and who cannot coordinate their purchasing decisions. Conversely, buyer concentration can significantly alleviate anticompetitive concerns. In this regard, central purchasing agencies or institutions that facilitate communication among buyers can help mitigate coordination failures. A careful analysis of buyer structure and bargaining power is therefore important, particularly to assess the extent to which buyers depend on the dominant firm for part or most of their sales.

Fourth, exclusionary price discrimination is more likely when downstream competition is weak. When downstream competition is intense, however, a single buyer’s demand may be large enough for an upstream rival to cover entry costs or, more generally, to reach efficient scale. The incumbent, then, cannot exploit coordination failures to exclude a more efficient rival.

Fifth, discount schemes that reference rivals — i.e., those that condition trade terms on how much the buyer purchases from a rival of the incumbent — such as market share discounts, raise greater anticompetitive risks because they limit how much benefit the incumbent must forgo to exclude rivals. Finally, it has also been found that the exclusionary effects of individualized discounts are stronger than those of standardized discounts, suggesting they should be treated with greater suspicion when used by a dominant firm.

In the United States, loyalty rebates are presumptively lawful, unless such discounts are predatory. In those cases, courts follow the predatory pricing standard set in Brooke Group: the plaintiff must demonstrate that the defendant charged prices below cost and that there is a high likelihood the alleged predator will recoup the profits sacrificed.

Quantity discounts that lead to prices above cost are generally considered legal (Kobayashi, 2005). This pro-discount presumption is reflected in the Virgin Atlantic case, where the Court of Appeals (Second Circuit) upheld a lower court’s ruling that British Airways’ travel agency discounts were within the bounds of the law.

In ZF Meritor v. Eaton Corp., the Third Circuit ruled in 2012 that loyalty rebates may be anticompetitive even if the discounted prices are not below the seller’s costs. This was primarily due to Eaton’s large market share in the truck transmission market and the entrenched demand from truck manufacturers —representing the non-contestable portion of demand— making Eaton’s products “must-have” goods. Thus, no truck manufacturer could meet client demand without at least some of Eaton’s products, meaning that “no [manufacturer] could afford to lose Eaton as a supplier” (ZF Meritor LLC v. Eaton Corp., 2012). Under these circumstances, the Third Circuit held that loyalty discounts had exclusionary potential regardless of whether they met the below-cost pricing test.

The European approach to discount practices differs significantly from that of the United States, as both jurisdictions often reach different conclusions. The Court of Justice of the European Union has held that “It can be deduced from that case-law generally that any fidelity-building rebate system applied by an undertaking in a dominant position tends to prevent customers from obtaining supplies from competitors, in breach of Article 82 EC, irrespective of whether the rebate system is discriminatory. The same applies to a fidelity-building performance reward scheme practised by a purchaser in a dominant position in relation to its suppliers of services” (British Airways v. European Commission, 2003).

However, over the years the European Commission has gradually moved away from this formalistic approach —where certain discount schemes are condemned per se (regardless of their effects)— towards incorporating economic considerations into cases. This shift was most clearly expressed with the 2009 publication of the Guidance on the Commission’s enforcement priorities in applying Article 82 [now 102] of the EC Treaty to abusive exclusionary conduct by dominant undertakings.

Despite this evolution, many observers argue that the “more economic approach” has not been fully embraced in EU case law and that European courts have sent mixed signals on the necessity to analyze competitive effects (Fumagalli, Motta & Calcagno, 2018). On the more formalistic side of the spectrum is the Tomra case (Tomra Systems ASA and Others v. European Commission), where the Commission found that the company —which marketed machines for collecting used beverage containers— had adopted a policy to maintain its dominant position by blocking entry and limiting rivals through loyalty rebates. In particular, the company used individualized retroactive rebate schemes with thresholds tied to the client’s total (or near-total) demand (Fumagalli, Motta & Calcagno, 2018). The Commission concluded that combining retroactive rebates with (almost) full-requirement thresholds created a significant incentive to buy nearly all equipment from Tomra and artificially raised switching costs. This was because surpassing the bonus threshold affected the customers’ entire purchase volume for the period, not just the purchases exceeding the threshold. The General Court upheld the Commission’s findings but clarified that the Commission was not obliged to perform a competitive effects analysis. The Court of Justice later fully upheld the General Court’s judgment, confirming the formalistic approach to rebate schemes offered by dominant firms.

The 2017 CJEU ruling in the Intel case is arguably the most important decision in this area and marked a potential shift away from the formalistic stance on loyalty discounts.

In 2009, the Commission accused Intel of abusing its dominant position in the CPU market by granting loyalty rebates and engaging in exclusionary practices targeting major computer manufacturers (Dell, HP, NEC, Acer, and Lenovo). To support its case, the Commission applied the so-called As-Efficient-Competitor Test (AEC Test). However, in June 2014, although the General Court upheld the Commission’s decision in full, it held that Intel’s rebates were exclusive in nature and, as such, inherently exclusionary. Thus, it did not consider the arguments regarding the AEC Test in determining the legality of the discounts.

In 2017, the CJEU overturned the General Court’s ruling and remanded the case for further consideration. This decision has been interpreted by some as a victory for the more economic approach in European jurisprudence. The CJEU ruled that “balancing of the favourable and unfavourable effects of the practice in question on competition can be carried out in the Commission’s decision only after an analysis of the intrinsic capacity of that practice to foreclose competitors which are at least as efficient as the dominant undertaking” (para. 140).

Broadly speaking, Chile’s Competition Tribunal (TDLC) has rejected loyalty rebates when applied by dominant players on a widespread basis without valid economic justification (Rencoret, 2020).

The Fósforos case illustrates this reasoning. On June 20, 2008, the Fiscalía Nacional Económica (FNE) filed a complaint against Compañía Chilena de Fósforos S.A. (CCF) for, among other things, offering a system of incentives and volume discounts that lacked economic justification and were structured with an exclusionary intent. These discounts were indivisible, retroactive, and required meeting sales targets at least equal to those of previous years. The TDLC found that these discounts had the same exclusionary effect as exclusivity contracts, since they incentivized buyers to concentrate all purchases with CCF, thereby preventing market entry by rivals (Domper, 2015). The absence of economies of scale or other efficiencies to justify the scheme, combined with CCF’s dominant position, confirmed that the discounts were designed to erect artificial barriers to entry.

Loyalty discounts have also been assessed in several subsequent cases brought by the FNE or private plaintiffs, including the Beer, Soft Drinks, Ice Cream, and Detergents cases. All of these concluded with settlement agreements in which the defendants committed not to enter into retroactive incentive arrangements tied to performance targets, at least while holding a market share above 50% or 60% (Rencoret, 2020).