Estimating Cartel Damages: a Stuttgart Blueprint

The Stuttgart Court of Appeal has added a new chapter to German cartel damages case law with its recent bathroom fittings judgment. Relying on Courts’ statutory power to estimate damages the court develops a multi‑step model to estimate cartel overcharges and pass‑on, anchored in economic meta‑studies and a transparent catalogue of damage‑relevant factors.

Stuttgart as a repeat player in cartel damages

Stuttgart’s Court of Appeal has been shaping German cartel damages law for several years, not just in its latest bathroom fittings judgment. In earlier cartel damages cases the court already combined careful engagement with economic arguments and to deal with information asymmetries and storytelling by defendants.

In its earlier judgment on the truck cartel, the Stuttgart Court of Appeal was confronted with the now familiar defence narrative that more price transparency between competitors must inevitably intensify competition and push prices down. The defendants illustrated this with the image of neighbouring market stalls that, once they can see each other’s prices, will undercut one another until only the lowest price survives.

Stuttgart carefully unpacked this story and held that a long‑running exchange of detailed list prices and configurators is at least as consistent with stable or higher prices as with any price‑decreasing effect, as it removes uncertainty and allows each participant to align its pricing strategy with the others. In the court’s view, the ‘market stalls’ narrative therefore proves little beyond the obvious fact that information affects pricing, and it certainly does not rebut the economic and legal presumption that a 14‑year coordination of list prices is capable of generating an overcharge.

The bathroom fittings judgment: from narratives to numbers

In its bathroom fittings judgment of 20 November 2025 (2 U 263/21), the Stuttgart Court of Appeal confronts head‑on the practical problem that detailed econometric expert reports often fail to provide robust, convergent answers on cartel overcharges. After criticising the claimant’s regression analysis as methodologically unsuitable on the available data, the court explicitly refuses to let this led to a denial of justice and instead turns to its own estimation powers.

A conscious turn away from regression

Stuttgart begins its model by explicitly distancing itself from the idea that a full econometric analysis is the ‘gold standard’ for quantifying cartel damages. The court notes that regression‑based expert reports in this and other cases suffer from fragile data foundations, heavy data‑cleaning needs, and high sensitivity to modelling choices such as the selection of variables, comparison markets and cost proxies. In the bathroom fittings case, these general concerns are reinforced by data problems on the second market level (retail), which make the specific regression submitted by the claimant unsuitable as a basis for exact quantification.

On that basis, the court concludes that insisting on a full regression in every case would make the exercise of the right to damages under Article 17(1) of the Damages Directive unreasonably difficult. Instead, it leans into the methodological discretion granted to German Courts: once the court is convinced that a cartel has caused harm, the provision allows the court to choose a fit‑for‑purpose estimation method that comes sufficiently close to the hypothetical competitive scenario without pretending to deliver mathematical precision.

A multi‑step model

Within this framework the court consciously designs its own estimation model. It emphasises that entrusts the trial court with a wide methodological margin to approximate the ‘but for’ price through probability reasoning and experience‑based inferences, and that this margin includes the option to estimate at least a minimum damage where the case supports that conclusion. Drawing on recent case law of the Federal Court of Justice, the court stresses that the difficulty of reconstructing the competitive counterfactual is itself a consequence of the unlawful cartel, and any residual uncertainty should therefore not be loaded onto the claimant.

The court positions its model as a deliberately independent alternative to fully fledged comparative market analyses. It proposes to estimate the overcharge on the basis of a statistically informed overall assessment of the specific case, supported by economic meta‑studies and empirically grounded experience‑based rules. The model is described as flexible and evolving rather than rigid: it can and should be refined over time in light of new empirical findings and judicial experience, and it aims to structure the decision‑maker’s reasoning rather than to function as a mechanical formula.

The building blocks of the model

The court sketches the building blocks of this model. First, it confirms that, if the court is convinced on the usual standard that a cartel caused a significant price effect, this conviction can also serve as the basis for estimating the magnitude of the effect. Second, it states that a sufficiently solid basis for estimation can make a separate comparative market analysis unnecessary, because the aim of the court proceedure is not exact replication of the counterfactual but a plausible, economically realistic approximation.

From there, Stuttgart announces a structured procedure that it will apply in the case and, by implication, in future cases as well. This procedure comprises several successive steps: establishing the existence of a price effect with clearly preponderant probability; locating the overcharge within a statistically derived ‘regular corridor’ for typical cartels; narrowing that corridor through a comprehensive assessment of agreement‑related, organisational, market and demand‑side factors; and finally dealing with pass‑on and possible downstream passing‑on as separate, later stages. Each of these stages is then unpacked in the following sections of the judgment, which collectively occupy dozens of pages and form the heart of the ‘Stuttgart blueprint’ for estimating cartel damages.

Step 1 – establishing a cartel‑related price effect

In the first step, Stuttgart asks whether the established infringement has, with clearly preponderant probability, led to a non‑trivial price increase on the first market level between cartelists and their direct customers. The court anchors this in two elements: the general experience, reflected in the Damages Directive, that cartels typically raise prices on average and the specific features of the case (type of conduct, duration, coverage, discipline, and market context).

Applied to the bathroom fittings cartel, the court first recalls the findings of the European Commission’s 2010 decision: a long‑lasting, multi‑country cartel covering a broad range of bathroom fittings, centred on coordinated list price increases, discount structures and exchanges of sensitive commercial information across six Member States. On this basis, and taking into account the duration of the cartel (over a decade), the breadth of market coverage and the organisational features documented in the Commission’s decision, Stuttgart concludes that it is clearly more likely than not that the cartel led to higher transaction prices on the first market level, even if the claimant’s specific regression analysis cannot be used. The failed regression is therefore downgraded to one weak, non‑decisive piece of evidence within a broader matrix that, in the court’s view, robustly supports the existence of a cartel‑related price effect.

Step 2 – locating the case within the 5–25% corridor

In the second step, Stuttgart moves from the existence of a price effect to its rough magnitude by placing the cartel within a statistically informed ‘regular corridor’ of 5 to 25% overcharge on the first market level. To construct this corridor, the court synthesises several economic meta‑studies (Connor, Oxera, Smuda, Boyer/Kotchoni) and notes that, once outliers and publication biases are controlled for, European cartels typically exhibit median overcharges in the area of 15 to 18% of the transaction price. This range is then cross‑checked against the Federal Court of Justice’s acceptance of a 15% overcharge as a standard‑terms lump sum (Schienenkartell VI), which Stuttgart treats as a normatively endorsed reference point for an ‘ordinary’ cartel damage.

On that basis, the court defines a cautious regular corridor from 5 to 25%: 5 % as a conservative lower bound for a pre‑selected set of pre‑effective cartels, 25% as an upper bound that still covers the bulk of empirically observed European overcharges without being driven by extreme outliers. Stuttgart then classifies the bathroom fittings cartel as a ‘normal’ cartel within this corridor, rather than an exceptional case that would justify going below 5% or above 25%: the cartel is long‑lasting, covers a significant share of the relevant markets, involves coordinated list price increases and discounts and shows no structural features that would systematically neutralise its price‑raising potential.

Step 3 – refining the overcharge within the corridor

In the third step, Stuttgart refines the overcharge estimate within the 5–25% corridor by carrying out a structured overall assessment of four ‘impact areas’: the content and duration of the agreement, the cartel’s internal organisation and discipline, the market conditions, and the reaction possibilities on the demand side. For each area the judgment assembles recognised determinants from case law and economic literature, such as type of coordination (prices, quantities, customer allocation, information exchange), duration and stability, market coverage, degree of product homogeneity, number and relative size of participants, role of trade associations, entry barriers, and demand elasticity. These determinants are then evaluated in terms of whether they are price‑increasing or price‑dampening in the specific case, and how strongly they ought to weigh on the final percentage.

When this framework is applied to the bathroom fittings cartel, most of the key factors point clearly towards the upper part of the corridor. The cartel combined coordinated list price increases and discount structures with intensive information exchange, operated over more than a decade, and involved a high and stable market coverage in several Member States, supported by sector associations. The products at issue (standardised sanitary fittings) exhibit a significant degree of homogeneity and are sold in a relatively concentrated market structure, which, together with limited outside options for wholesalers in the relevant period, speaks in favour of a strong price effect; at the same time, there is no convincing evidence of weak cartel discipline, destabilising market entry or highly elastic demand that would systematically have restrained price increases. Weighing these elements, Stuttgart concludes that the realistic overcharge on the first market level lies in a relatively narrow band near the top of the corridor, around 21–23% of the purchase price, and adopts a value within this range as the basis for calculating the claimant’s gross damage before pass‑on.

Step 4 – estimating pass‑on to the claimant

In the fourth step, Stuttgart turns from the first‑level overcharge to the question of how much of that overcharge was passed on along the supply chain to the claimant. Conceptually, the court models pass‑on on the basis of a Cournot‑type framework: the extent to which an upstream cost increase (here, the cartel overcharge) is shifted to downstream buyers depends on the intensity of competition and the price elasticity of demand on the downstream market. High downstream competition and relatively inelastic demand tend to result in substantial pass‑on, whereas weak competition or very elastic demand may lead intermediaries to absorb a larger share of the overcharge in their own margins.

Applied to the bathroom fittings case, the court examines the wholesale market on which the claimant purchased the cartelised products. It finds that sanitary wholesalers supplying the claimant operated in a competitive environment and faced significant pressure to maintain volumes, while end‑consumer demand for standard bathroom fittings was not highly price‑sensitive at the relevant margins. On that basis, and drawing on the Cournot logic, Stuttgart concludes that a large portion of the 21 to 23% upstream overcharge was passed on to the claimant and quantifies the pass‑on rate at around 80%, meaning that the claimant is treated as having borne roughly four‑fifths of the first‑level overcharge in its own purchase prices.

Step 5 – downstream passing‑on and benefits

In the fifth step, Stuttgart addresses the mirror question on the claimant’s side: whether the claimant has itself passed on part of the overcharge to its own customers so that a corresponding benefit must be deducted. The court sets out a cautious framework: defendants bear the burden of substantiating downstream passing‑on, and in markets with many small end‑customers and value‑added services, such as retail, any alleged benefits are often diffuse and hard to trace in a way that would justify concrete deductions.

Applied to the case at hand, the court emphasises that the bathroom fittings were used as inputs into a broader DIY retail offering, with no evidence that the claimant systematically adjusted its consumer prices in line with the overcharge. Given the fragmented end‑customer base and the low risk that those customers will individually sue the cartelists, Stuttgart finds no sufficient basis to quantify downstream passing‑on and therefore refuses to reduce the claimant’s damages on this ground, leaving the 80% passed‑on overcharge as the relevant loss to be compensated.

Methodological underpinnings: flexible systems and empirical corridors

The Stuttgart model does not emerge in a methodological vacuum. It sits squarely in the ‘flexible system’ tradition developed in recent German scholarship on cartel damages, which builds on Wilburg’s idea of a basic valuation, structured elements and empirically informed corridors instead of a search for a single ‘true’ overcharge. At its core, the judgment adopts the same basic valuation that has been proposed in this debate: the cartel‑induced widening of the price‑setting range and its exploitation in the form of higher prices as the relevant benchmark, with market coverage, substitutability, duration, customer sensitivity and cartel discipline as central elements.

Where the literature had set out these elements in explicitly methodological terms and suggested overcharge corridors and tables as a way to translate economic knowledge into the practice, Stuttgart effectively ‘judicialises’ this intuition. The court reorganises the same determinants into four impact areas (agreement, organisation, market conditions, demand reaction), builds empirically grounded corridors (5 to 25% on the first level, with room for higher values in exceptional cases), and uses them to structure its balancing within the ‘estimation window’ that late Heike Schweitzer and others have identified as inherent in German law. In doing so, it comes remarkably close to, for instance, Isikay’s and others work on presumptive overcharge tables: reducing the role of case‑specific mega‑expert reports, relying more openly on robust meta‑evidence and experience‑based ranges and requiring courts to make their weighing of damage‑relevant elements transparent instead of hiding it behind the rhetoric of exact economic measurement.

Link to Federal Court of Justice and EU law

Stuttgart also makes clear that its flexible, empirically informed model is not judicial free‑styling, but an implementation of the Federal Court of Justice’s recent guidance as well as Article 17 of the Damages Directive. The court repeatedly stresses that residual uncertainty about the exact ‘but for’ price is an inevitable consequence of the infringement and must not be shifted to claimants if German and the EU law effectiveness principle are to have any bite.

Practical significance for future cases

Finally, the judgment implicitly offers a template for other courts: instead of commissioning ever more elaborate regressions, they can adopt a similar multi‑step structure, rely on shared empirical sources for corridors, and then explain their weighing of case‑specific factors within that frame. For practitioners, this makes damages outcomes more predictable: once the key determinants of a given cartel are mapped to Stuttgart’s four impact areas, the likely position within the corridor  becomes much easier to anticipate.