Publicado:  © European Union, 2016.
CONTRIBUTORS: Blas Alberto González Navarro.
Socio fundador Blas A. Gonzalez Abogados
Abogado, ex-socio Cuatrecasas.
Magistrado en excedencia. 

Executive Summary:

Introduction

This Study is intended to provide judges, and other practitioners who are not economic experts, with practical guidance on obtaining and assessing economic evidence in relation to pass-on in the context of competition law infringements. Drawing on relevant economic theory and quantitative methods, as well as relevant legal practice and rules, it sets out a framework for evaluating the plausibility of claims, for quantifying the effects of pass-on, and, accordingly, for assessing the total extent of the harm suffered by a claimant.

EU Directive 2014/104 establishes that any person who has suffered harm caused by a competition law infringement may claim full compensation for that harm. This includes the possibility of indirect claims, which arise when those that are not directly affected by such an infringement (notably, indirect purchasers) are nevertheless harmed as a result of changes in the behaviour of directly affected firms (the direct purchasers) as well as, potentially, other intermediate firms.

There are three distinct elements that make up the recoverable harm potentially suffered by a claimant. First, there is the increase in the claimant’s costs (“the overcharge”) that may be brought about by the infringement: in legal terms, actual harm or direct loss (damnum emergens). Such harm may arise directly or because of “upstream” pass-on by a direct or indirect purchaser that supplies the claimant.

Second, the adverse impact of the overcharge on the claimant may be reduced if it passes on some or all of that overcharge to its own customers, by means of a price increase. This is the “passing-on” effect. Whilst such “downstream” passon reduces the actual harm suffered by the claimant in question, it will do so at the expense of causing harm further downstream. Indeed, the pass-on effect at one level of the supply chain implies an overcharge of the same magnitude at the next level downstream; they are two sides of the same coin.

In litigation, pass-on can, therefore, serve as a “sword”, where an indirect purchaser alleges that an overcharge has caused it harm because of upstream pass-on. It can also be used as a “shield”, where a defendant alleges that downstream pass-on by a claimant has reduced the actual harm the latter has suffered.

Third, to the extent that a claimant suffers a loss of sales volumes as a consequence of pass-on, it will lose the profit margins associated with those sales. This so-called “volume effect” constitutes recoverable loss of profit (lucrum cessans) in legal terms and forms part of the overall damage calculation. Whenever a firm increases its prices, it will almost invariably suffer such a loss of sales volumes. It is the extent of this prospective loss, which hinges on the sensitivity (or elasticity) of a firm’s demand to price increases, that tempers the extent of passing-on in the first place.

Case-law and legal framework

Directive 2014/104 establishes the new legal framework for pass-on. The Directive, and its national implementing legislation, will become the principal legal basis for adjudication of pass-on issues by national courts in the EU as this new regime becomes effective. The Directive, notably, confirms the availability of the pass-on defence (with defendants carrying the burden of proving that pass-on has occurred) and establishes a legal presumption of pass-on for indirect purchasers (provided certain conditions are met). It further provides that national courts should be able to estimate pass-on.

Prior to the Directive, the Court of Justice of the European Union had developed case-law on pass-on, principally in the area of reimbursement of taxes or charges unlawfully levied in breach of EU law, which forms part of the acquis communautaire. This case-law has inter alia stressed the importance of an adequate case-by-case economic analysis to prove pass-on. For their part, national courts have, to date, had limited experience of pass-on questions. Notably, they (and experts) have relied heavily on certain basic parameters (such as the number of firms affected by an overcharge) to assess pass-on when it has been alleged and have tended to adopt rather simple or theoretical approaches to quantification. They have not typically considered the volume effect in their quantification of damages. Experience in the US is far more developed; albeit borne from a very different legal context (for instance, the existence of opt-out class actions). All of these experiences will be of (varying) relevance for national courts: in some circumstances, providing legal parameters for their work and, in others, offering useful practical insight into what may be reasonable means of going about their task and what difficulties they may face.

The Study provides an overview of this past case-law and experience, as well as setting out the key provisions of the Directive.

The economics of pass-on

Pass-on and the associated volume effects arise because of the incentives that a firm may have to respond to increases in its costs by raising prices. Economics can play an important role in identifying such incentives and the sensitivities of pass-on effects to the specific features of the case at hand. As such, it provides a structure for establishing the coherence and plausibility of claims about passon, and a framework within which qualitative and quantitative evidence can be evaluated.

Relevant cost effects

Economics suggests that it is changes in a firm’s variable (or, more precisely, marginal) costs which will usually have the most immediate influence on pricing decisions. At the same time, fixed costs (i.e. costs that do not vary with the level of output) are predicted to affect those decisions in some specific situations and, notably, over the longer term. On the other hand, small cost changes may have no influence on prices – at least not immediately – if firms incur “menu” costs in adjusting prices, if there are rigidities affecting output adjustment, or if relevant change in costs is not identified as such.

The effects of competition

The extent to which a firm can increase price profitably will depend on the volume of sales it will lose as a result. This, in turn, will depend on the extent to which competitors are themselves affected by the overcharge (pass-on of “industrywide” overcharges is generally predicted to be greater than for “firm-specific” overcharges of the same magnitude), and the intensity of competition on the market, including the way those rivals will respond to any passing-on by affected firms.

Under textbook conditions of perfect competition, a firm which is the only supplier to experience a cost increase due to an infringement (i.e. the overcharge is firmspecific) will be unable to pass on that overcharge at all. On the other hand, if all competitors are affected, i.e. the impact of the overcharge is industry-wide, the extent of pass-on in this perfectly competitive environment is predicted to depend on the relative price elasticities of supply and demand. Notably, 100% pass-on of industry-wide overcharges is predicted when industry supply is perfectly elastic (i.e. the supply curve is flat).

Where competition is imperfect – and few markets, if any, are likely to resemble the textbook model of “perfect” competition in practice – the pass-on of industrywide overcharges is generally predicted to increase as competition intensifies, whereas the pass-on of firm-specific overcharges will tend to decrease. However, even under monopoly, a firm can be expected to pass on some of the effect of an overcharge affecting marginal costs. In these imperfectly competitive settings (involving monopoly and oligopoly markets), economic theory indicates that the curvature of demand, i.e. the way that the slope or elasticity of demand changes as price changes, can have an important influence on the magnitude of pass-on. Pass-on rates that exceed 100% are possible, as are much lower rates, as a result. Nevertheless, the influence of demand curvature is predicted to diminish as competition intensifies.

In general, predictions regarding the extent of pass-on will depend on the precise nature of competition, inter alia. The more complete the information that is available on relevant facts of the case at hand, the better the guidance that economic theory is likely to be able to provide to courts on the magnitude of passon effects.

The impact of buyer power

In some settings, buyer power acts as a constraint on the pricing behaviour of suppliers. It might be supposed that this would automatically enable such buyers to resist the pass-on of overcharges too. However, this may or may not be the case. Instead, a detailed analysis of the characteristics of specific negotiations and the context in which they take place is required to establish pass-on implications.

The relationship between pass-on and volume effects

In general, the pass-on and volume effects will have opposite effects on the harm suffered by a claimant. When the affected purchaser is a monopoly on the downstream market, the volume effect will generally exceed the pass-on effect (though the difference will shrink as the overcharge diminishes). As a result, a measure of the overcharge alone will understate that harm. Outside monopoly, however, the relationship between the pass-on and volume effects in imperfectly competitive settings will also depend on the strategic responses of competitors. If competitors would respond to an increase in the affected firm’s prices by also raising their own prices, this will tend to reduce the harm, all else being equal. Conversely, if rivals would expand their output in response to a reduction in the affected firm’s sales, then this will tend to magnify the harm.  

Quantifying the impact of pass-on

Sound economic analysis will take into account all available evidence and, if possible, be supported by robust empirical analysis. This Study presents a number of approaches to estimating the passing-on and volume effects. The focus is on quantitative approaches. Where the availability of suitable data is limited, however, experts may have to rely on more qualitative evidence (e.g. documents and statements on how firms set prices).

Sequential approaches

Experts may consider estimating the three components of damages separately or sequentially. Quantification of the overcharge that increases the claimant’s costs will typically constitute the first step. If that overcharge is positive, even if small, the next step is to consider whether and, if so, to what extent this has been passed on. Where pass-on has occurred, it is then relevant to consider the extent to which the claimant has lost sales volumes as a result.

Pass-on effects
The main challenge in estimating the pass-on effect is to obtain a measure of the relevant price increase (on the downstream market). Doing so requires a measure of the counterfactual price which would have prevailed ‘but for’ the infringement. The pass-on effect can then be computed by multiplying the estimated price increase by the volume of the firm’s sales.

Where relevant data is available, experts may adopt a direct approach to this task, i.e. using price information directly to estimate the increase brought about by passing-on of the overcharge. This will involve the use of comparator-based techniques, as can also be used to estimate the initial overcharge. Candidate benchmarks for the counterfactual price include the prices of (i) the affected product before or after the infringement period (the “before/during/after” approach), and (ii) the same or similar products in different geographies, that were not subject to the infringement (a “benchmarking” approach). A combination of these two techniques (a “difference-in-differences” approach) may yield more robust results.

Economics indicates that the change in prices resulting from an overcharge can also be calculated by multiplying the relevant part of the overcharge (i.e. the absolute amount by which marginal costs have been increased) by a measure of the relevant pass-on rate. This is the rate at which changes in costs are translated into changes in prices. A pass-on rate can be estimated where cost and price data are observable, including using multi-variable regression techniques. Evidence in respect of firm’s pricing policies will also offer relevant insight. An expert might also draw on pre-existing estimates of pass-on rates.

An important practical consideration is whether pass-on rates estimated in a different set of circumstances provide an appropriate measure of the pass-on rate in the case at hand. For example, pass-on rates might differ according to the scale of the cost changes concerned or according to the input that is affected. This may be a particular issue when such pass-on rates are used to estimate the effects of a (small) overcharge that cannot be identified directly.

Volume effects
The volume effect can be computed by multiplying the margin the purchaser would have earned ‘but for’ the infringement (i.e. the counterfactual margin) by the reduction in volume sold that results from passing on the overcharge.

If the purchaser has passed on only part of the overcharge (that is, the pass-on rate is below 100%), then the observed margin during the infringement period can be very different to the counterfactual margin. Information on the pass-on effect can be used to make an appropriate adjustment. Alternatively, the expert may consider using comparator-based techniques to recover a measure of the counterfactual margin.

To quantify the reduction in volume sold, comparator-based techniques can again be deployed. Alternatively, the expert may consider using a measure of the elasticity (or price sensitivity) of demand for the product or service in question, combined with an estimate of the pass-on effect on price. Because the elasticity varies with price changes, this will deliver an approximate estimate of the volume loss.

Moreover, the choice of the appropriate elasticity is important. When a whole market is subject to an overcharge (and all firms are affected similarly), economics suggests that an estimate of the aggregate market demand elasticity is likely to provide the best measure of the proportionate impact on any firm’s sales, since it captures the consequences of a market-wide increase in prices. However, a measure of the affected firm’s own-price elasticity of demand may provide a better starting point when an overcharge is firm-specific. Nevertheless, in this case, any volume loss estimates may also need to account for the impact of competitors’ price (alternatively, output) responses.

Further, at the cost of imposing additional structure, economic theory can simplify the volume effect calculation by replacing the need for a measure of elasticity, albeit an appropriate account of competitors’ response will still be required.

Holistic approaches

Experts may consider adopting a holistic approach which calculates total damage in an integrated way, accounting simultaneously for the pass-on and the volume effects. Discount and simulation approaches constitute two such approaches, both drawing on formal economic models. By introducing additional (assumed) structure to the calculation, these approaches may reduce the data requirements to allow an expert who already has an estimate of the initial overcharge to put a specific value on the purchaser’s total economic loss.

Ranking and reliability

The reliability of the damage estimate will depend primarily on the quality of the information (or data) used and the nature of the assumptions adopted.

In general, the holistic approaches rely on economic models that make relatively strong assumptions about firm and consumer behaviour. These models can only provide a reliable basis for predicting market outcomes, therefore, if these assumptions reflect the realities of the market. Hence, motivation for relevant modelling choices will be especially important in these cases.

In contrast, the sequential approach does not rely to the same extent on particular behavioural assumptions. Instead, comparator-based techniques, which can be used to derive estimates of counterfactual prices, margins, and volumes, depend mostly on finding suitable benchmarks that are uncontaminated by the infringement, as well as gathering sufficient data to control for confounding influences, notably using multi-variable regression analysis. Other, less sophisticated, techniques may also be considered where the available data are limited or when using more sophisticated analyses would be disproportionate. The costs and benefits of adopting alternative approaches should be identified.

The accuracy with which the pass-on rate can be estimated depends largely on the amount of information (notably, data) available. For instance, when there is not sufficient relevant data available, experts may have to rely on pass-on rates derived in different circumstances, assuming thereby that the relevant pass-on rate is the same. Where such assumptions prove critical to the estimates of harm, they should be tested against relevant facts.

The assumptions which underpin a particular quantification exercise should be set out transparently, and their plausibility explored. This includes a description of data sources and any data manipulations that have been undertaken. Where possible, the sensitivity of estimates to the specific assumptions adopted and the quality of the data used should be explored. For instance, the expert may evaluate how results vary if plausible adjustments to key assumptions are made. Moreover, where detailed data analysis has been conducted, the expert can obtain statistical measures of the potential margins of error associated with the estimated parameters of interest.

The judge’s role of assessing economic evidence of pass-on

Where pass-on is invoked, national courts are charged with assessing economic evidence of pass-on as part of their task of evaluating its existence and extent in accordance with national legal rules of evidence, causation and standard of proof (within the framework of EU rules in relation inter alia to burden of proof and presumptions, as well as the principles of effectiveness and full compensation). This judicial task will necessarily involve consideration of all relevant evidence adduced by the parties, including evidence of a factual nature (such as a firm’s pricing policies). These, and other issues of particular relevance to courts’ assessment of economic evidence of pass-on, are also addressed in this Study.

Disclosure

The Study considers a number of ways for courts to manage effectively the access of parties to documentation and information held by other parties to litigation or other non-litigating firms, in order to prove pass-on. Judges should take adequate control of such disclosure processes, ensuring fulfilment of the principles of reasonableness and proportionality. Economic experts can assist judges in this task; e.g. by assessing whether pass-on is sufficiently plausible in the circumstances of the case, or explaining why access to certain data can be expected to result in better estimates of pass-on effects. Mechanisms for ensuring the adequate confidentiality of business secrets which have to be shared are also relevant.  

Use of economic experts

The rules on expert evidence vary significantly between the EU Member States. Nevertheless, judges may employ a number of aids and mechanisms to maximize the usefulness of that evidence; including: (i) the use of qualified court-appointed experts and/or party experts; (ii) the exchange of information, and discussion of approaches, by experts (whether appointed by the parties or by the court) with a view to narrowing issues; and (iii) the testing of conflicting evidence. The Study provides some insight into these types of procedural tools.

Parallel Proceedings

Courts may be called to consider the existence of separate judicial proceedings where pass-on in the same or similar circumstances is (or has been) an issue. Courts have a duty in such cases to try to avoid inconsistent rulings and, accordingly, to ensure results which are coherent from an economic (as well as a legal) perspective; albeit such duties are by no means absolute. More generally, the economic treatment of similar issues in other proceedings may offer useful insight or serve as a point of contrast. The Study considers what mechanisms may be available to courts in these circumstances.

39 steps: a checklist for judges

Finally, to assist national courts in assessing economic evidence in relation to the quantification of pass-on and volume effects, this Study contains a practical checklist of issues which may arise, organised around a set of key questions. This is designed to help courts understand and evaluate the economic evidence adduced.