§5.1.3 Challenges

(112) When estimating the passing-on related price effect, the court may particularly consider techniques which, to the largest extent possible, control for factors other than the one stemming from the infringement. The difference-in-differences method is such a technique. It requires information or data from a comparator market (for instance another geographical market) and time-series data from the market affected by the passing-on. However, the court should be aware that there are potential challenges that may affect how valid the comparator-based methods may be. (113) Ideally, the comparator market is similar to the infringement market, but itself not affected by the infringement. However, the purchasers on each of the markets often use the same input. In such a case, it might be difficult to find an unaffected comparator. In particular, if the scope of the infringement covers a broad geographic area, it is likely that products, similar to the product in question and incorporating the same input, have potentially been affected, as well. This can make it difficult to find a suitable comparator market. (114) In other circumstances, the comparator market may be indirectly affected by the initial overcharge. In the stylised example of the copper cartel in Box 6, the wire harnesses supplier B1 purchases copper from the infringer A1. Even though the wire harnesses supplier B2 in the comparator market does not purchase from the infringer A1, the wire harnesses suppliers B2 and B1 may be competitors on the same geographic downstream markets. This implies that, if the wire harnesses supplier B1 increases its prices in response to the initial infringement, its competitors may raise their prices, as well. In this case, the price that the wire harnesses supplier B2 offers may have been indirectly affected by the infringement, and as a result may not provide a suitable comparator (97). (115) As regards comparison over time, it may be challenging to identify with sufficient precision the period when the market was affected by a certain infringement. The parties may present a decision issued by a competition authority which mentions an infringement period, i.e. sets out dates at which the infringement started and ended. However, this period may not correspond to the period in which a market was actually affected by the infringement. It is also important to note that determining the dates of either the infringement period or the period in which the market was affected, can have a strong impact on the outcome of the analysis. In reality, the effect of the infringement may not be limited to the period provided in such decision (98). On the one hand, the start date identified by the competition authority may post-date the actual start of the infringement, for instance due to lack of reliable evidence (99). On the other hand, the end date provided in an infringement decision may pre-date the end of the actual infringement. (116) The effects of an infringement may also not be limited to the duration of the infringement. It is possible that the infringement will affect the market concerned even after the conduct prohibited under EU competition law has ceased. In particular, this may be the case in oligopolistic markets, if the information gathered during the infringement allows the suppliers of a certain product to adopt, on a sustainable basis after the infringement has ended, a course of action aimed at selling at a higher price than the competitive price, i.e. that would have been charged in the absence of the infringement, without engaging in practices prohibited by EU competition law (100). (117) The possibility that purchasers at different levels of the supply chain may delay the passing-on of an overcharge can also affect the comparison significantly (101). The example in Box 1 above may illustrate this. Suppose the car manufacturer C negotiates prices with the wire harnesses supplier B on an annual basis. The wire harnesses supplier B only adjusts prices once a year after the negotiations with the car manufacturer C have been finalised. If a price fixing cartel in the copper market is established just after the negotiations between the wire harness supplier and the car manufacturer have ended, it is only when the next year's annual negotiations take place that the wire harnesses supplier may have had an opportunity to pass on the copper-price increase in their own price. (118) Hence, the delay of passing-on down the supply chain may result in difficulties when deciding on the relevant period for comparing prices during and before or during and after the infringement (or both). The court may adjust the analysis by considering the nature of each case, for instance by analysing the pricing policy of the parties and, based on that, introducing a certain time lag when analysing the pricing patterns at different levels of the supply chain. (119) The use of regression analyses when employing the comparator-based approach may increase the accuracy of the estimated passing-on effect. A judge should be aware that, while these types of regression models offer a direct estimate of the passing-on effect, they do not quantify the extent to which each of the theoretical factors affecting the passing-on, discussed in section 3 above (e.g. nature of input costs and nature of the product demand), contributed to the effect. A strength of using regression analyses is that it is less likely to make assumptions related to economic theory that are wrong. While such a factor-by-factor quantitative breakdown of the estimated passing-on effects is usually not necessary for the estimation of the passing-on effect, it can be noted that the techniques described in section 5.2.3 below might provide a quantitative breakdown of the factors.

← 5.1.2 · All articles · 5.2 →

Source: EUR-Lex (Cellar) · retrieved 2026-09-07