In the USA, where case law on antitrust damages is more developed than in other jurisdictions, relatively few cases have dealt with the issue of quantifying pass-on. The reason is that the passing-on defence has been ruled out by the federal courts, and only direct purchasers can claim cartel damages (although a number of US states do allow indirect purchasers to claim damages). Economic incentives have played a role in this policy decision—it is perceived that direct purchasers are best placed, and hence most likely, to file a claim.
In contrast, under EU legal principles, anyone who has suffered a loss has a right to claim compensatory damages, whether they are a direct or indirect customer. To ensure that only purchasers who actually suffered overcharge harm can effectively claim compensation, the European Commission’s proposal for a Directive on competition law damages—issued in summer 2013—explicitly recognises the possibility for the defendant to invoke the passing-on defence. The situation in the various national courts in the EU is not clear-cut and case law is limited, although in most jurisdictions the relevance of the passing-on defence has been recognised.
Policy considerations in the proposed Directive
The Commission’s proposed Directive also contains some elements regarding pass-on that are driven by policy considerations beyond the compensation principle.
First, the Commission proposes that, in situations where the overcharge was passed on to parties for whom it is ‘legally impossible’ to claim compensation, the passing-on defence cannot be invoked. The aim is to avoid a situation where no one can effectively claim damages from a cartel. While this objective might be achieved, disallowing the pass-on defence in certain circumstances would also provide for the possibility of some claimants being compensated beyond the loss that they actually suffered.
In the case of an action for damages brought by an indirect purchaser, the Commission further proposes the introduction of a rebuttable presumption, subject to certain conditions, that direct purchasers did pass on (at least some) cartel-related price increases to their own downstream customers. This proposal recognises practical difficulties that indirect purchasers might experience with regard to proving that direct purchasers did indeed pass on part or all of the cartel-related price increases to them.
From an economic perspective, in certain circumstances it is plausible for a direct purchaser to have passed on all of a cartel-related price increase to their own downstream customers. However, there are many other circumstances in which a direct purchaser has absorbed the cartel overcharge in part or full. The extent of pass-on will always depend on the facts of the case. A rebuttable pass-on presumption is therefore not grounded in economic principles as such, but rather driven by policy considerations.
Rebutting this presumption can also be challenging for the defendant. This is because the defendant carries the burden of proof for the passing-on defence, while the commercial decision—and internal evidence thereof—concerning whether or not to pass on price increases lies with the purchaser.
Finally, while the Commission refers to the rebuttable pass-on presumption only in the context of an action brought by indirect purchasers, it is unclear how this presumption could be ignored in actions brought by direct purchasers. Introducing a rebuttable pass-on presumption in all damages claims would weaken the position of direct purchasers in many private damages actions, which is unlikely to be consistent with the EU policy objective of facilitating such actions.
Independent of the current policy discussion, allowing the passing-on defence means that the degree of pass-on needs to be quantified. How can economists help?
Economic theory provides a number of useful benchmarks for assessing pass-on. At one end of the spectrum is complete pass-on, where the full value of the input cost change caused by the cartel is translated into a change in the retail price of the product. Economic theory predicts that this outcome will be found when the market resembles perfect competition—a situation where costs and prices are very closely linked. While most markets are not perfectly competitive, empirical evidence suggests that a number are sufficiently competitive as to result in high pass-on rates—see, for example, the recent UK Office of Fair Trading (OFT) inquiry into petrol retailing, which found pass-on rates close to 100%.
At the other end of the spectrum is complete absorption, where the full value of the input cost change is absorbed by the claimant (and nothing is passed on). This may occur if the claimant competes with suppliers that buy non-cartelised inputs and is hence unable to pass on cartel-related cost increases. For example, in a damages action against a sugar cartel in Spain, the court considered that biscuit producers in the country were affected because they compete in European markets with foreign producers that did not purchase their sugar from the cartel. By implication, the Spanish producers had to absorb the overcharge on sugar or else lose market share.
In between these two outcomes lie situations where the level of pass-on is driven by the degree of market power held by the retailer. The greater the market power—and hence the lower the degree of competition—the further the pass-on rate falls below 100%. A theoretical finding in economics is that a standard textbook monopolist would seek to pass on around 50% of a cost increase. In oligopolistic markets the theoretical pass-on rate would be between 50% and 100%.
Price-setting practices and pass-on
Basic economic insights alone may not be sufficient to explain pass-on rates in specific markets. The way in which companies actually set prices in practice matters. Examples include cost-plus pricing, price pointing, and the passing on of small-cost items versus large-cost items, as discussed below.
Cost-plus, or mark-up, pricing occurs where a seller calculates the cost of the product and then adds a fixed proportion of it as a mark-up. For example, a retailer might buy (or produce) a product for €8.00, apply a 25% mark-up, and price the product for sale at €10.00. If the input cost rose by €1.00 to €9.00, the retail price would rise to €11.25. If the €1.00 input cost rise were due solely to a cartel between manufacturers of the product, the retailers’ cost-plus pricing rule would suggest that it passed on the overcharge in full.
In this particular example, the retailer actually passed on more than 100% of the overcharge. This is because its margin has also risen. It was earning €2.00 before the input cost rise, and is now earning €2.25, implying a pass-on rate of 125% (a 100% pass-on rate would imply that the absolute margin stayed the same at €2.00). Yet one can ask—from both a legal and an economic perspective—whether the extra €0.25 represents pass-on as such, since it reflects a reconfiguration of price, not an increase in cost that is passed on.
Another important question is whether the mechanistic pricing rule—in this case, the 25% mark-up—did remain the same during the period of the cartel. If it did not—for example, if the retailer adjusted its percentage mark-up to reflect changing market conditions—the rate of pass-on might still be less than 100%, despite cost-plus pricing rules.
Another practical pricing rule is price pointing. Sellers may prefer ‘attractive’ price points, such as €9.95 rather than €10.05, or, to continue the example above, a rounded price of €11.00 or €12.00 rather than €11.25. Sellers may also prefer simplicity and continuity in pricing. The price of the print edition of The Times newspaper remained at £1 for several years (it is now £1.20), while the price of a Big Mac (an important international pricing benchmark monitored through The Economist’s Big Mac Index) in Spain has been €3.50 for years. Any changes in input costs in this period—e.g. the cost of beef or printing paper—would therefore not have been passed on. Moreover, there can be ‘menu’ costs associated with resetting and communicating the final-product price—most restaurants do not print new menus every time the prices of meat, fish or potatoes change.
Price changes in all these situations can be lumpy, and may no longer be directly related to changes in underlying costs. The rate of pass-on may be low as a result, at least in the short term.
Small price increases: the espresso machine effect
Another interesting question relates to the likelihood of relatively small price increases being passed on. Assume that pass-on is 100% and the overcharge is 20%. The final-product price increases by 10% if the input cost represents 50% of the final-product price, but only by 2% if it represents 10% of the final-product price. (In other words, if the final-product price is €100, and the initial input cost is €50 before the cartel and €60 during the cartel, the price increases to €110, which is 10% higher; if the input cost is only €10, the final price increases to €102, so by 2%.)
Even though in standard theoretical models it does not matter in practice, the magnitude of input cost may influence the rate of pass-on. A variety of factors can determine this, and will typically have to be considered on a case-by-case basis.
If the affected input cost makes up only a small proportion of the final-product price, there may be no pass-on if the affected business chooses not to reset prices. This may occur where downstream prices are set with respect to only major and more visible input costs, or where there are ‘menu’ costs associated with resetting and communicating the final-product price (as noted above).
Alternatively, small changes in the input price may well be fully passed on in some circumstances if their magnitude is sufficiently small as to avoid any significant demand reduction. This reasoning was used by a French court in the 2007 vitamins cartel case (Juva v Hoffman La Roche). The court considered vitamins to be a small part of the finished good and that a small price increase would be sufficient to offset the overcharge. It also noted that the price of the claimant’s finished good had increased by more than the prices of the vitamins, and that its sales volumes had also grown.
Economic theory is divided on this point. One economic approach is to assess the relationship between price and overall costs (or overall marginal costs), and infer from this that the pass-on rate is the same for all individual cost items, be they large or small. After all, when setting prices, companies may look at their total costs, so even small items are considered in this way (since total costs are simply the sum of individual cost items).
However, there may be differences between small-cost items and large-cost items, both in commercial reality and in theory (perhaps in a similar way to physics, where quantum mechanics, dealing with phenomena at microscopic scale, is famously inconsistent with relativity theory, which deals with large-scale phenomena).
In practice, companies may change prices with regard to only major or more visible input costs, while small-cost items are ignored. Can there still be pass-on in such circumstances?
Take the example of a major law firm that has acquired a fancy espresso machine for its staff area, at a price of €299. It turns out that the manufacturers of the machine had formed a cartel such that the law firm paid an overcharge of 10%, so €29.90. Has the law firm passed this additional cost on?
Let’s say this successful law firm increases its hourly rates by 10% at the start of the new year. When setting the new rates, the firm’s management committee did not pay much heed to the cost of the espresso machine. However, it did look at the overall expenditure of the business—in particular, lawyer salaries and office costs, which include the cost of the machine. The resulting extra income rapidly swamps the overcharge, so the law firm manages to recover all its costs overall. But does this mean that it has actually passed on the costs of the espresso machine to its clients? Or did it absorb these costs? (Another way of asking this is: would its price rise have been any different if the espresso machine had been cheaper?) There are no clear legal, economic or philosophical answers to this question.
Competition authorities: mind your language!
Competition authorities often consider it important to demonstrate the consumer benefits of their interventions. This is one reason why, in infringement decisions, a competition authority may include a statement that the effects of the anticompetitive conduct (a cartel or abuse of dominance) were ultimately felt by end-consumers.
Such statements are not helpful from the perspective of claimants in damages actions who operate at intermediate layers of the supply chain, since these statements imply that all costs have been passed on to end-consumers. For example, in another damages action in France against the vitamins cartel, the court held that the earlier European Commission decision and press release had stated that the cartel affected end-consumers and therefore that direct purchasers were able to pass on their cost increase. The judgment highlights that, logically, direct purchasers cannot have been harmed by the overcharge if it is established that end-consumers have faced 100% of the overcharge—i.e. the overcharge harm must not be double-counted. It is open to question whether the Commission’s statement about end-consumers being harmed was simply a general statement, and whether it was actually meant to imply that consumers suffered 100% of the overcharge.
Pass-on and volume effects
Where a purchaser has passed on (a proportion of) the overcharge, they may still have suffered a volume harm resulting from that pass-on, which is legally a separate type of harm. This harm is more likely to arise where the cartelised input makes up a greater proportion of the final-product price.
In the USA, this harm is rarely claimed for in cartel damages actions. In Europe, the Court of Justice acknowledged the possibility of such a loss-of-volume harm in the presence of complete pass-on in a 1997 judgment on port fees illegally levied in the French territory (this was not a competition law ruling).
Perhaps as a result of the passing-on defence being allowed, and used, in Europe, there is an increasing number of cartel damages claims that are not restricted to the overcharge but also cover harm from lost volumes.
In the EU, both direct and indirect purchasers can claim antitrust damages. This raises the question of pass-on—i.e. at what stage in the supply chain did the damage from a cartel materialise? If purchasers passed on all or some of a cartel overcharge to their own downstream customers, they cannot claim compensation for that proportion of the overcharge. For this reason, pass-on is a key issue in many cartel damages cases in the EU.
Economic theory provides useful insights into the likely degree of pass-on. Higher levels of competition tend to result in higher levels of pass-on. This is because intense competition results in prices being closely related to costs, so if costs change—regardless of whether this is due to a cartel or other factors—so do prices. A competition assessment therefore provides a useful starting point for any pass-on assessment.
There are, however, instances where economic insights may not be sufficient to explain the likely pass-on outcomes. Price-setting practices, such as cost-plus pricing, usually suggest higher rates of pass-on; price pointing suggests lower rates. Cost increases that are small relative to the price of the final product might be associated with full or zero pass-on, depending on the facts of the case.
Ultimately, one would aim to estimate empirically the actual relevant pass-on rates. This would require access to data on actual prices and costs at the relevant layers of the supply chain, and would usually involve econometric techniques. Obtaining reliable results will not always be possible due to difficulties with data. The number of cases where a full empirical pass-on analysis has been carried out by economists and then reviewed by the court is, as yet, limited. However, approaches are being developed, and at some point the case law will develop too.