chevron-down Created with Sketch Beta.

Antitrust Magazine

Volume 35, Issue 1 | Fall 2020

Assessing Umbrella Pricing Incentives

Mary Beth Savio

Summary

  • The unilateral "umbrella" pricing incentives that a non-cartel member may experience as a result of a cartel’s elevated pricing may be useful in understanding the link between the pricing of cartel and non-cartel members. 
  • In combination with the broader set of facts specific to a given matter, this approach may contribute to an understanding of the factors, such as indeterminacy and remoteness, that courts consider when evaluating whether such purchasers have standing to sue the cartel for damages.
Assessing Umbrella Pricing Incentives
Klaus Vedfelt via Getty Images

Jump to:

When collusive agreements involve a subset of firms in an industry, they may create the incentive and ability for firms that are not participants in the cartel to raise their prices––so-called umbrella pricing. In the European Union, Canada, and in some cases in the United States, consumers who purchase a product from firms that engage in umbrella pricing (umbrella purchasers) have standing to sue the cartel members for damages resulting from “overcharges” from umbrella pricing. The defendants in these matters are the cartel members, not the non-cartel members from which the consumers purchased the good. In jurisdictions that allow umbrella purchaser claims, plaintiffs must demonstrate that the losses they incurred stem from collusive actions, which predictably undermined competition. This highlights the importance of identifying and characterizing the link between the pricing of cartel members and non-cartel members.

This article provides an overview of the requirements for standing for umbrella purchasers in the U.S., EU, and Canada, and demonstrates how in a marketplace with differentiated products, an analysis of the upward pricing pressure experienced by non-cartel members may contribute to an assessment of whether the requirements for standing are met.

Standing for Umbrella Purchasers Varies by Jurisdiction

The ability of umbrella purchasers to sue the participants in a cartel for damages associated with price increases implemented by non-cartel members varies across countries and jurisdictions. In the EU in 2014, the Court of Justice of the European Union ruled that EU Member States cannot categorically exclude umbrella pricing claims against cartel members, and determined that umbrella purchasers have standing in matters where the cartel’s actions created an incentive for umbrella pricing and where that effect was foreseeable.

In the United States, courts are split on the question of standing for umbrella purchasers. U.S. courts have tended to reject umbrella claims, citing the difficulty of proving injury and calculating damages; for example, in In re Vitamins Antitrust Litigation the court noted that “‘[t]he causal connection between plaintiffs’ injury and the alleged conspiracy is necessarily attenuated by significant intervening factors, such as independent pricing decisions of the non-conspiring [parties]’ . . . [and the] intervening factors . . . make it impossible to apportion damages between the overcharges and the other factors with any reasonable degree of certainty.”

But in some cases, U.S. courts have recognized the standing of umbrella purchasers, noting that plaintiffs would have to demonstrate that non-cartel members’ prices were elevated as a result of the cartel’s actions, and assessing the difficulty of estimating damages for umbrella purchasers to be no more complex than estimating damages for cartel purchasers. For example, in Costco Wholesale Corp. v. AU Optronics Corp., the court found that if “Costco has standing to pursue its version of the umbrella theory . . . Costco would have to present evidence showing that the ‘minor players’ charged their higher panel prices because of the conspiracy, not for independent reasons . . . .” and “Costco’s claims [regarding defendants] will require the court and the jury to tackle . . . evidentiary complexities . . . regardless of Costco’s umbrella damages. Eliminating the umbrella theory would not make the economic inquiries less complex.”

In Canada, courts were split on standing until September 2019, when the Supreme Court of Canada determined that plaintiffs who purchase from umbrella pricers have standing to sue cartel members for damages. The decision was made in the context of an appeal to Godfrey v. Toshiba, a class action filed in British Columbia that alleged that Pioneer, Toshiba, and other manufacturers of optical disk drives (ODDs) and ODD products had conspired to fix prices between 2004 and 2010. The proposed plaintiff class included (among others) umbrella purchasers who had purchased an ODD or ODD products from a non-defendant.

The Supreme Court of Canada was asked to address whether umbrella purchasers have the right to sue the members of a cartel for their economic losses. The majority ruled that umbrella purchasers have standing to sue if they can prove that their losses foreseeably resulted from unlawful behavior by the cartel members––that is, whether plaintiffs can show that non-cartel members increased their prices as a predictable result of the cartel members’ price-fixing agreement. The Court determined that these conditions were met in Godfrey, but noted that they might not be met in other cases, noting “I am of the view that indeterminate liability would not arise in this case in any event” but “whether indeterminate liability might properly be considered at all in the context of a claim under s. 36(1)(a) of the Competition Act––I am content to leave for another day.”

In contrast, the dissenting opinion argued that standing for umbrella purchasers exposes defendants to indeterminate liability because cartel members cannot predict or control the actions of firms that are not members of the cartel (indeterminacy), and that standing for umbrella purchasers exposes defendants to potential claims by customers of firms whose prices are only tangentially influenced by the cartel members (remoteness): “[Claimants should not be allowed] to recover from defendants for any losses that in some way flowed from the alleged price‑fixing conspiracy as it would expose defendants to liability that is potentially limitless in scope for loss and damage that are too remote from any price-fixing that occurred.”

The Court’s decision opens the door to expanded class actions with regard to the size of the class and the amount of potential damages in price-fixing matters in Canada. But it also holds that umbrella purchasers must demonstrate that their losses resulted from collusive actions that predictably undermined competition––a set of requirements similar to those in the EU and in some U.S. courts. These standards require assessing the link (or lack thereof) between the pricing of cartel members and non-cartel members. For example, when members of a cartel increase their prices, do non-cartel members have an incentive to increase their prices? Can the magnitude of the umbrella price increase be approximated (and thus be foreseen)?

An Analysis of Upward Pricing Pressure May Assist in Determining Whether the Requirements for Standing for Umbrella Purchasers Are Met

The logic underlying umbrella purchaser claims is that the prices charged by non-cartel participants are set in response to the supracompetitive prices charged by the cartel members: the elevated prices maintained by the cartel reduce the competitive restraints on non-cartel members and, as a result, the non-cartel members set their prices higher than they would have in the absence of the cartel.

Whether such claims characterize a particular matter depend on the facts specific to the matter. A non-cartel member might not increase its price for a number of reasons. For example, a non-cartel member might be unable to change prices for customers who purchase under contractual pricing agreements. A non-cartel member also might not have sufficient bargaining power to negotiate a price increase with customers. Furthermore, if a non-­cartel member’s product is a somewhat distant substitute for the products produced by the cartel, it may only be able to raise prices by a de minimis amount. And a non-cartel member may increase its price for reasons unrelated to the cartel, such as an increase in marginal cost. The point is that an analysis of umbrella pricing requires a fact-specific assessment of the particular matter.

One approach that likely will contribute to understanding umbrella pricing incentives is to perform an upward pricing pressure analysis for the non-cartel members. When cartel members increase their prices, some of the cartel members’ customers may shift their purchases to non-cartel members, which may create an incentive for non-cartel members to increase their prices. Modeling these interactions can likely help to characterize and quantify the link between the pricing of cartel- and non-cartel members.

As an example of one such model, consider a market with three firms which are all at the same level in the chain of production. Each firm sells a single differentiated product, and the products are (imperfect) substitutes. The firms initially compete on price, with each firm determining the price that maximizes its profit, given the prices of the other two firms. In equilibrium, each firm’s price will be set such that any increase in price beyond that point would be unprofitable for the firm––the additional revenue it would earn by setting a higher price would be more than offset by the profit it would lose when some of the firm’s customers shift their purchases to other firms.

Now assume Firms 1 and 2 form a cartel in which they each agree to increase their price by a certain amount. (Assume Firm 3 does not join the cartel, perhaps to avoid violating antitrust laws.) Collusion between Firms 1 and 2 may change the competitive incentives for Firm 3: the price that Firm 3 set before the formation of the cartel may no longer be the price that will maximize Firm 3’s profit. That is because as a result of the price increase of Firms 1 and 2, some consumers may shift their purchases to Firm 3, and if the demand for Firm 3’s product increases, the price which maximizes Firm 3’s profits also typically increases. Firm 3’s incentive to increase price will be counterweighted, as before, by the profit Firm 3 would expect to lose when some of the Firm 3’s customers shifted their purchases away from Firm 3. Thus, in the context of this model, whether and how much Firm 3 will increase its price in response to the cartelization of Firms 1 and 2 depends on two things: (1) the extent to which the customers of Firms 1 and 2 will respond to a price increase by Firms 1 and 2 by shifting their purchases to Firm 3, and (2) the extent to which customers of Firm 3 will respond to a price increase by Firm 3 by shifting their purchases away from Firm 3.

The likelihood of these occurrences can be better understood by breaking them down into their component parts. For example, the tendency of customers of Firms 1 and 2 to respond to a price increase by shifting their purchases to Firm 3 depends on the size of Firm 1 and 2’s price increase; the sensitivity of Firm 1 and 2’s customers to price; and the extent to which Firms 1 and 2’s customers consider Firm 3’s product to be a good substitute for the products of Firms 1 and 2. Similarly, the tendency of Firm 3’s customers to respond to a price increase by shifting their purchases away from Firm 3 depends on the size of Firm 3’s price increase and the sensitivity of Firm 3’s customers to price.

Firm 3’s price increase will depend on the strength of these tendencies, as well as the respective volumes involved. For example, if Firm 3’s customer base is small relative to the customer bases of Firms 1 and 2, then all else equal, the number of customers Firm 3 expects to lose due to a price increase may be small relative to the number of customers Firm 3 gains due to the price increases of Firms 1 and 2.

In summary, Firm 3’s price increase will be larger:

  • the greater is the price increase of Firms 1 and 2
  • the more price sensitive are the consumers of Firms 1 and/or 2
  • the greater is the substitutability of Firm 3’s product for the products of Firms 1 and/or 2;
  • the less price sensitive are the original consumers of Firm 3
  • the smaller is Firm 3’s quantity relative to Firms 1 and/or 2.

Firm 3’s pricing incentives can be further characterized and quantified by the following formula, which estimates the amount by which Firm 3 will increase its price in response to an percent price increase by Firms 1 and 2.

Equation 1

Equation 1

Equation (1) shows that Firm 3’s percentage price increase will be proportional to the percentage price increase, s, of the colluding firms, where the proportionality factor is given by the fraction in brackets in Equation (1). The first term in the numerator includes the own-price elasticity of demand for Firm 1 (ε1)––that is, the percent by which Firm 1’s demand decreases in response to a 1 percent increase in Firm 1’s price; the diversion ratio from Firm 1 to Firm 3 (δ13), which is the fraction of the demand lost by Firm 1 that is diverted to Firm 3; and the pre-cartel quantity (q1) of Firm 1. The magnitude ε1 δ13 q1 in Equation (1) reflects the consumer demand that is shifted from Firm 1 to Firm 3, when Firm 1 increases its price by 1 percent. The second term, ε2 δ23 q2, has a similar interpretation with respect to Firm 2’s price increase. The term in the denominator includes the own-price elasticity of demand for Firm 3 (ε3) and the pre-cartel quantity sold by Firm 3 (q3). This term reflects the consumer demand that Firm 3 will lose for each percentage point increase in its price. Equation (2) describes the same relationship using an alternative formulation; it expresses the determinants of Firm 3’s price increase as variables that often can be estimated or approximated using firms’ financial information (Q i , mi , s) and strategic plans and studies (δij ), where Q i is the pre-cartel quantity share of Firm i and mi denotes Firm i ’s percentage profit margin.

Equation 2

Equation 2

Table 1 provides a set of illustrative examples which demonstrate how, in the context of this model, Equation 2 can be used to estimate the extent to which Firm 3 will increase its price in response to the cartelization of Firms 1 and 2. The first column of the table serves as the base case. The base case assumes that Firms 1, 2, and 3 are of the same size and profitability, and that 45 percent of consumer demand diverted from Firm 1 or Firm 2 would shift to Firm 3. The base case also assumes that through collusion, Firms 1 and 2 agree to sustain a 10 percent price increase. Under these assumptions, Equation (2) implies that the price increases of Firms 1 and 2 create the incentive for Firm 3 to increase its price by 4.5 percent.

Table 1: Illustrative Examples


 
Example 1: Firm 3 is similar to Firms 1 and 2 Example 2: Firm 3’s product is a more distant substitute  Example 3: Firm 3 is small relative toFirms 1 and 2  Example 4: Firm 3’s consumers are more pricesensitive
  Firms 1 and 2 are established producers of widgets. Firm 3 is:
Variable Name Notation an established producer of widgets an established producer of gadgets  a new producerof widgets an established producer of economy widgets
% price increase by Firms 1 and 2 s 10% 10% 10% 10%

Quantity demanded from each of Firms

1 and 2

q1, q2 1,000,000 1,000,000 1,000,000 1,000,000
Quantity demanded from Firm 3 q3 1,000,000 1,000,000 1,000,000 1,000,000
% profit margin of Firms 1 and 2 m1, m2 30% 30% 30% 30%
% profit margin of Firm 3 m3 30% 30% 30% 30%

Diversion ratio from Firms 1 and 2

to Firm 3

δ13, δ23 45% 45% 45% 45%
% price increase by Firm 3 (p3’ − p3)/(p3) 4.5% 3.5% 9.0% 2.3%

Note: values of qi, mi, δij, and p3 are stated on a pre-cartel basis. The pre-cartel quantity share of Firm i is Qi ≡ qi / Σ31qj.

The second column considers an alternative scenario; it is identical to the base case except it assumes that Firm 3’s product is a more distant substitute for the products of Firms 1 and 2. Intuitively, that would suggest that Firm 3’s price increase would be lower than it was in the base case, and that is what Equation 2 shows: if 35 percent of consumer demand diverted from Firm 1 or Firm 2 would shift to Firm 3 (instead of 45 percent, as assumed in the base case), then a 10 percent price increase by Firms 1 and 2 creates the incentive for Firm 3 to increase its price by 3.5 percent (instead of 4.5 percent , as implied by the base case).

The third column considers a scenario that is identical to the base case except it assumes that Firm 3 is half the size of Firms 1 and 2. Consistent with intuition, under this assumption Equation 2 shows that Firm 3’s price increase is higher than it was in the base case: a 10 percent price increase by Firms 1 and 2 creates the incentive for Firm 3 to increase its price by 9 percent (instead of 4.5 percent , as implied by the base case).

The fourth column considers a scenario that is identical to the base case except it assumes that the price sensitivity of Firm 3’s customers is doubled relative to the base case (as measured by a halving of its percentage profit margin). As one would expect, Equation 2 shows that under this assumption Firm 3’s price increase is lower than it was under the base case: a 10% price increase by Firms 1 and 2 creates the incentive for Firm 3 to increase its price by 2.3 percent (instead of 4.5 percent , as implied by the base case).

Several additional comments are worth noting. First, Firm 3’s price increase does not involve collusion with Firms 1 and 2. In fact, as noted by Inderst et al., Firm 3 need not be aware of the collusive agreement between Firms 1 and 2; Firm 3 merely observes Firm 1 and 2’s price increases and responds unilaterally, in accordance with its own economic incentives.

Second, as explained by Inderst et al., if the price increases of Firm 1 and 2 are large enough, it is possible that Firm 3 would not be in the same relevant market that Firms 1 and 2 are in when they set their prices competitively. For example, in the context of this model, suppose Firms 1 and 2 are the only two firms in their relevant market when they set their prices competitively. If Firms 1 and 2 then collude to increase their prices significantly, some consumers will shift their purchases from Firms 1 and 2 to a firm that provides an imperfect but affordable substitute (which in this example would be Firm 3). In this situation, the elevation of Firm 1 and 2’s prices gives Firm 3 the incentive to increase its price, even though Firm 3 is not included in the relevant market of Firms 1 and 2 when they price competitively.

Third, Equation 2 does not provide a complete analysis of umbrella pricing incentives. Consistent with the literature on unilateral incentives created by upward and downward pricing pressure, Equation 2 analyzes the pricing incentives of a given player (Firm 3) holding the actions of other players fixed. It also does not account for the possibility of repositioning of existing firms or entry by new firms in response to Firm 3’s price increase. As noted in other studies, upward and downward pricing pressure analyses characterize and quantify potential outcomes based on firms’ economic incentives without involving the development of a more complex simulation model.

Fourth, under the assumptions of the model, Equation 2 implies that Firm 3 will increase its price by some amount, since the terms in Equation 2 are all positive. However, the model accounts for the fact that Firm 3’s price increase could be inconsequentially small. For example, if Firm 3’s product is a more distant substitute for the products of Firms 1 and 2, then the model indicates that the price increases of Firms 1 and 2 may have very little impact on Firm 3’s price. More generally, this observation highlights the importance of considering the implications of the model in the context of the broader fact base of a particular matter. As noted above, a non-cartel member might not change its price if, for example, it is unable to change prices for customers who purchase under contractual pricing agreements or if it does not have sufficient bargaining power to negotiate a price increase with customers. Or a non-cartel member may increase its price for reasons wholly or partly related to increases in its marginal cost. And so on. These examples illustrate the importance of considering the results of the model in the context of a more holistic assessment of the economic factors influencing the firm.

In summary, an analysis of the unilateral pricing incentives that a non-cartel member may experience as a result of a cartel’s elevated pricing is likely to be a useful tool to understand the link between the pricing of cartel and non-­cartel members. In combination with the broader set of facts specific to a given matter, this approach may contribute to an understanding of the factors, such as indeterminacy and remoteness, that courts consider when evaluating whether umbrella purchasers have standing.

The author thanks Michelle Burtis, Serge Moresi, Margaret Sanderson, and Timothy Snail for insightful comments.

    Authors