This is the opening piece of a short series examining how three European jurisdictions seek to resolve large competition claims efficiently.
Private enforcement of competition law in Europe has grown into an established field of litigation, where mass proceedings are routinely brought and resolved. Claims initiated by or on behalf of sometimes thousands of victims, including terabytes of transaction data, and economic evidence of a particular complexity challenge the ordinary rules of civil procedure.
Once the numbers of injured parties and claims reach a certain scale, the design of the procedural system has a decisive impact on how those claims can be enforced efficiently. An efficient system resolves well-founded claims through adequate compensation and rejects unfounded claims, while minimising the overall cost in time, money and institutional resources.
The efficiency of a system can be measured from various perspectives. It has a temporal dimension (the duration from claim to final resolution), a financial dimension (the cost of running the proceedings measured against recoverable damages, viewed from the perspective of claimants and defendants alike), an institutional dimension (the concentration of cases before specialised courts) and a structural dimension (the procedural mechanisms available for aggregating claims and for staging proceedings into phases). These dimensions are interdependent, as the same procedural measure can advance efficiency in one dimension while impairing another.
With claimants usually having a choice of various forums, Germany, the Netherlands and the United Kingdom have emerged as the leading jurisdictions for (mass) competition litigation. In fact, many of the same cartels or abusive behaviours have given rise to proceedings across more than one of them. Each of these jurisdictions has developed a distinct procedural approach to resolve large-scale competition claims. And each has its advantages and drawbacks, with none superior to the others across the board.
With that frame in place, the first question is: how, and to what extent, does the aggregation of claims create efficiency?
1.1 The efficiency case for aggregation and its limits
The aggregation of several claims from the same or different victims of a competition infringement is considered efficient in various ways.
The first efficiency driver is cost efficiency. The aggregation of claims reduces the overall costs of litigation, as it spreads the fixed costs (e.g., understanding the infringement and the markets affected, crafting legal arguments, developing the economic theories) across the aggregate claim value. These costs arise normally only once, whether the proceedings serve one claimant or a multitude of claimants. And even defendants against competition claims can save costs when faced with one instead of multiple sets of proceedings. Bundling can also reduce defendants’ costs by replacing multiple proceedings with a single action, avoiding duplicative court appearances, repeat expert evidence or inconsistent rulings. Where fees do not rise proportionally with the value in dispute, or are capped above a certain threshold, aggregation compounds these savings. The economies of scale therefore speak in favour of aggregation.
However, collective enforcement also creates costs of its own, which individual claims would not incur. For example, there are transaction costs for its constitution, be it gathering claims by assignment, organizing a claimant group or defining a represented class. The mechanism must then survive judicial scrutiny through certification, an admissibility test or challenges to the validity of assignments. This stage – rather than the merits – risks becoming the first battleground of the proceedings. The net efficiency gain therefore depends on the ratio of common to individual points of dispute – the more common points, the greater the saving on fixed costs.
Aggregation often cannot reduce the costs specific to each injured party, be it their transactions in dispute, pass-on data, or knowledge relevant for limitation, either (save, in some instances, in opt-out class actions where liability and quantum can be determined in aggregate). For data-driven exercises of collecting and structuring the relevant data and identifiers and validating it with documentation, digitalization at all levels (including the courts) has allowed massive savings and paved the way for AI-supported and automated processes, which are now the new standard.
An argument heard in some jurisdictions is that aggregation delivers a more accurate quantification of harm. A single claimant’s data may not be sufficient to robustly determine damages by way of a regression analysis. Pooling the transaction data of various claimants can be what makes a robust quantification exercise possible, and a larger, more representative sample can produce a more precise estimate than any individual claimant could establish alone. In a case of pooled data, however, it must be ensured that the underlying transactions are sufficiently comparable, creating homogeneity in the data.
Efficiency gains appear also in a temporal sense, in the time taken until the resolution of the case – though only on average. A single claimant litigating alone might in principle reach judgment faster than as part of a large bundle. Still, the acceleration that bundling delivers stems from the fact that without bundling, many parallel proceedings would compete for the same judicial and expert resources.
Litigating the common issues of a cartel or market power abuse case once, rather than in a (large) number of parallel proceedings, also protects consistency. A single decision on the same infringement with identical economic evidence avoids divergent judgments which parallel proceedings inevitably risk. This, in turn, reduces the risk of the same questions being appealed, again with potentially divergent outcomes.
Injured parties deciding to bundle their claims also often do so to increase their settlement leverage. By combining numerous claims into a single action, claimants present a more credible litigation risk and thereby improve their bargaining position in settlement negotiations.
The notion of rational disinterest is often mentioned as an argument in favour of the aggregation of claims. Rational disinterest means that cartel overcharges and damages are often spread across a large number of victims, and for many of them the individual harm is too small to justify the cost, duration and risk of a claim. Aggregation – often paired with litigation funding – is the natural answer to make those claims viable. However, enabling claims which would otherwise not be brought is not an efficiency aspect strictly-speaking. Rather, it serves the core objectives of private competition law enforcement: securing full compensation for injured parties and preserving the deterrent effect of competition law. The CJEU in ASG 2, where it held that a national prohibition of the German assignment model may be incompatible with EU law, did not reason with regards to procedural economy, but rather the European principle of effectiveness and the access to justice.
The three jurisdictions examined have divergent approaches. The contrasting approaches are not solely the product of competition-law policy. They also reflect deeper procedural traditions. Civil law systems have traditionally favoured a predominantly written process in which judges actively structure and sequence the litigation, whereas common law systems evolved around oral hearings, adversarial presentation of evidence and greater procedural flexibility. The respective aggregation mechanisms developed in Germany, the Netherlands and the United Kingdom exhibit traces of those traditions, even if modern competition litigation has increasingly blurred the distinction.
1.2 Germany – aggregation without a class action
In Germany a competition damages claim can be pursued collectively via different mechanisms.
Collective actions. The statutory collective redress regimes, the Musterfeststellungsklage (model declaratory action with erga omnes fact-finding) as well as the opt-in Abhilfeklage (collective redress action), which was introduced in 2023 implementing the Representative Actions Directive, are however not suited to bring competition claims. They are confined to consumer claims with small enterprises treated as consumers. The collective redress action is additionally subject to a strict third-party funding bar, as funders may not receive more than 10% of the proceeds of the case. Both restrictions render the instruments effectively unattractive for competition damages cases. An opt-out class action regime is unknown to German law.
Joint action as co-claimants. German procedural law provides for the option of bringing a action by several injured parties jointly. Each claimant keeps its own claim and remains a party to the proceedings.
The efficiency driver here are the positive effects created on statutory court costs and lawyers’ fees by joining the claims. Under German law, both court fees and the statutory lawyers’ fees are calculated based on the value in dispute on a so-called degressive scale: the fee rises as the amount in dispute rises, but it rises ever more slowly. A claim ten times larger does attract less than ten times the fees; the marginal cost of each additional euro in dispute falls as the value climbs. The value is also capped at €30 million, so values in dispute higher than that do not increase court costs or statutory lawyers’ fees. In consequence, the more claims are packed into a single proceeding, the cheaper each becomes to run.
Assignment model. The dominant route in practice to compensate for the absence of collective redress is the assignment model. Injured parties assign their claims to a specialised provider registered for debt collection, which then sues in its own name as holder of the bundled claims. The attraction for the injured party is that it becomes neither a claimant nor – in case the model is combined with litigation funding – a cost debtor. This non-statutory aggregation through the substantive law of assignments has been recognized by the CJEU in ASG 2, which held that a Member State may not categorically prohibit it where it is the only effective means of enforcing the right to compensation.
The same cost incentives as for the joint action by several claimants are available for the assignment model. However, the assignments themselves can create efficiency issues: Only if they are valid, the specialised provider can claim on its own behalf and can have standing. There are various arguments with which assignments are being attacked in a lawsuit. The validity of the assignments has shown to be the first battleground and to make up a large part of the legal dispute on the merits of a bundled case.
The Federal Court of Justice has recently brought more clarity to this question. It confirmed that (a) the debt collection registration covers cartel damages claims, (b) that claims assigned by foreign injured parties can be pursued and that German law is applicable to those and (c) that the involvement of a litigation funder does not generally affect the validity of the assignments (see the previous blog post on The Competition Law Blog). While this decision has provided long-awaited validation of the model, it would be naïve to think that it puts further disputes about the validity of assignments to a halt.
Aggregation reaches its limit where the volume of claims makes adjudication within a reasonable time impossible. The Federal Court of Justice found that limit exceeded in the recent case, where some 70,000 claims from around 3,300 injured parties in 21 countries had been bundled into one action and presented to the court largely unsorted and, on the claimant’s own account, in part without prior examination. With express reference to the time required to resolve this case, the court held that the claims could not be adjudicated together and that the lower courts must now separate the proceedings along suitable criteria. In a more recent case, the Federal Court of Justice held that the bundling of more than 300,000 claims from 36 mostly German assignors, presented in a structured manner, is not abusive. These judgments show that claim aggregation by assignment is structurally reversible and the cost advantages of bundling cannot be taken entirely for granted at the outset. However, with a robust and high-quality presentation of the claims the separation risk can be bypassed.
Coordination by the court. Lastly, German courts have a set of procedural tools to avoid doing the same work repeatedly. Where several actions pending before the same court concern the same subject matter, the court may join them and conduct them as a single proceeding for a hearing or for the taking of evidence. The trucks cartel litigation before the Landgericht München I illustrates the latter: in November 2025 the specialised chamber examined its two court-appointed economists in a joint hearing covering 36 separate proceedings, so the economic evidence was tested once rather than 36 times. A court can also rely on a (court-appointed neutral) expert report obtained in other proceedings instead of commissioning a fresh one. This is relevant for an economic damages analysis which is usually expensive and time-consuming. It is therefore cost- and time-efficient for a later case (and for that only) to make use of pre-existing economic analyses.
1.3 The Netherlands – three paths to collective redress
The Netherlands has followed a different path from Germany. Rather than relying on a single mechanism, Dutch law offers claimants three separate routes for bringing collective actions: direct, individual and regular on behalf of several claimants, the assignment model, and the Wet afwikkeling massaschade in collectieve actie (WAMCA), the Dutch class action regime.
Direct, regular civil law proceedings, under which individual claimants authorise a Dutch attorney to litigate on their behalf while each claimant remains a party, are the least common of the three in collective competition cases, in particular where large claimant groups are involved. It is very often the case that these types of proceedings are too expensive and take too much time for individual claimants when considered against the background of the dispute which is centred around damages for often past behaviour of defendants.
The assignment model has therefore become the dominant and established structure in Dutch competition damages litigation: injured parties (almost always companies) assign their claims to a special purpose vehicle (SPV), which then litigates in its own behalf as the bundled claimant. Dutch courts have accepted this model, though – as in Germany – the validity of the underlying assignments is not a given, and its scrutiny may consume considerable time before the substantive dispute is even reached. In the Trucks case, the Amsterdam District Court initially found the assignment documentation to be insufficiently substantiated, but rather than dismissing the claims outright, allowed a further period to supplement the record. It later confirmed the standing of claim vehicles bundling damage claims through assignment or mandate and upheld the validity of the assignment model as a route to compensation. Defects identified in the assignment chain do not automatically cause the case to fall apart – claimants are typically given the opportunity to cure them. This pattern points to an underlying efficiency logic: Dutch courts appear reluctant to let collective redress fail purely on a technical or curable defect in the assignment documentation, provided the substance of the claim is intact.
The third route, the WAMCA opt-out regime introduced in 2020, is the newest and, in a sense, the most ambitious. It was enacted to address the shortcomings of its predecessors, the Collective Action Act (WCA) and the Collective Settlement Act (WCAM). It allows a qualified representative entity to bring a collective action on behalf of an entire class (also consumers) without requiring individual (domestic) claimants to assign their rights or to opt in. Foreigners have also access to the WAMCA regime, however, for them the proceedings are opt-in instead of opt-out. The regime is still comparatively young, and the body of case law interpreting it is slowly expanding.
Before a WAMCA case reaches the merits, it must clear the first and most crucial hurdle: the admissibility phase. Under the WAMCA, the claimant organisation is put under scrutiny, whereby ensuring that the group has standing and sufficiently represent the interests of the claimant group. The court’s assessment focusses on several key areas, including whether the affected group is sufficiently represented, whether sufficient governance and control mechanisms are in place (such as via a supervisory board and statutes), and whether there are sufficient financial means to cover the legal costs throughout the course of the proceedings. A notable development in this regard is the recent Dutch Supreme Court ruling in the TPC/Oracle and Salesforce case issued on 17 July 2026, where it confirmed that the admissibility requirements must be tested ex nunc (by reference to the situation at the time of the court’s decision, not when the proceedings were issued). Similar to the assignment model, this gives claimant organisations room to fix representativeness or governance defects rather than let admissibility challenges shut down a case on the merits.
Where multiple representative foundations compete to bring the same claim, the court resolves this by appointing a single exclusive representative, similarly to a carriage dispute in the UK. Only once the exclusive representative has been selected, will the case be able to proceed.
It must be noted that this layered admissibility and certification process can become a source of delay: practitioner feedback cited in the government’s ongoing evaluation of the regime points specifically to processing times caused by the longer admissibility stage and to uncertainty surrounding appeals in WAMCA proceedings.
Finally, a defining feature of Dutch competition damages litigation more generally, is the strict phasing of proceedings and active case management of the courts. This is especially important in collective actions, as a larger claimant group adds a layer of complexity to a case that can warrant a phased approach, for instance through discussions on procedural issues such as jurisdiction and applicable law, in particular where not all claimants are Dutch. This phased approached will be discussed at length in the second edition of this competition litigation blog series. For the purpose of this article, this form of active case management acts as a double-edged sword: on the one hand, it creates efficiency in the proceedings as parties are not required to invest resources upfront into elements of a case that are dependent on the outcomes of the preceding phases. On the other hand, it can also be a source of inefficiency, as these individual phases can themselves run into delays. The obligation of Dutch courts under Article 20 of the Dutch Code of Civil Procedure to conduct proceedings in an efficient manner cannot always avoid unnecessary delays.
1.4 The United Kingdom – specialist collective proceedings
The United Kingdom has adopted a different model. Unlike Germany, it has a statutory collective proceedings regime specifically designed for competition claims. Unlike the Netherlands, collective competition claims can only be brought in the specialist Competition Appeal Tribunal (CAT). The regime is available on both an opt-in and opt-out basis.
Under section 47B of the Competition Act 1998, a proposed class representative may bring claims on behalf of a class if the Tribunal grants a Collective Proceedings Order (CPO). The regime applies to both follow-on and standalone claims, and permits the Tribunal to award aggregate damages without requiring each class member’s loss to be individually assessed.
This structure captures many of the efficiencies associated with collective proceedings. Common questions concerning infringement, the relevant counterfactual, causation and economic loss can be determined once for the class, avoiding the need for claimant-by-claimant litigation. In opt-out proceedings, class members need not become parties or take active steps to participate. The proceedings brought by Dr Rachael Kent on behalf of approximately 19.6 million UK consumers provides a useful illustration: although each class member’s loss was relatively modest, the opt-out regime enabled those claims to be pursued collectively. As the Supreme Court recognised in Merricks v Mastercard, the CPO regime facilitates claims that might otherwise never realistically be brought.
While grouping claims in principle brings efficiencies, such proceedings have added procedural steps. Before a collective claim can proceed, the CAT must be satisfied that the proposed proceedings are suitable for collective determination. This preliminary stage can reduce inefficiencies later in the proceedings by testing whether common issues can realistically be resolved on a class-wide basis. At the same time, certification itself adds a procedural layer that can become a significant source of cost and delay, meaning that part of the efficiency debate in UK collective proceedings concerns whether the benefits of aggregation justify the complexity required to establish them.
It is worth noting that the formal CPO procedure is not the only route to aggregation available in the UK. Claimants can also achieve many of the same efficiencies informally, by bringing individual claims that are then case managed together and listed for trial in batches, without any certification process. The Trucks litigation illustrates both routes operating side by side: alongside the CPO ultimately granted to the Road Haulage Association, a separate set of individual claims – including the Veolia, Suez and Wolseley proceedings, amongst others were bundled through case management directions and progressed without any certification process at all.
Whether collective proceedings are likely to deliver efficiency gains is tested at the certification stage. The CAT will not grant a CPO simply because claims can be grouped together, it must consider the overall costs and benefits of doing so. In Waterside Class Limited v Mowi ASA, the CAT refused to certify proposed opt-out proceedings after a detailed costs-and-benefits assessment. It found that the anticipated costs of litigation, the modest sums likely to be returned to individual class members and the proposed distribution arrangements were unlikely to deliver sufficient value to the class. The decision illustrates that certification is not automatic.
Carriage disputes can add a further procedural stage where rival applicants seek to represent overlapping classes. The CAT determines which representative is best placed to conduct the proceedings weighing economic methodology, litigation strategy, governance, funding and legal representation. Although this creates additional cost, it ultimately promotes efficiency by ensuring that only one collective action proceeds in respect of substantially the same conduct. In BIRA Trading Ltd v Amazon.com Inc and Professor Andreas Stephan v Amazon.com Inc, the CAT selected Prof Stephan as the representative for overlapping Amazon Marketplace claims.
Aggregate damages provide one of the regime’s clearest advantages. The CAT may calculate class-wide loss without determining every individual member’s loss. As the Supreme Court explained in Merricks v Mastercard, “Where aggregate damages are to be awarded, section 47C of the Act removes the ordinary requirement for the separate assessment of each claimant’s loss in the plainest terms” (paragraph 58). Aggregation does not, however, eliminate heterogeneity. Material differences in purchasing behaviour, contractual terms or pass-on may still require subclasses or differentiated methodologies.
Distribution of the aggregate damages fund through a Tribunal-approved process presents a related challenge. Identifying class members, verifying eligibility and designing a proportionate distribution methodology are costly, administratively complex, and closely scrutinised. To date, there are few examples of distributions in CAT class actions. Experience from opt-out collective redress regimes in other jurisdictions suggests relatively modest take-up rates, that is, the proportion of eligible class members who actually come forward to claim compensation, often in the region of 5–20%. Against that background, the less than 1% take-up achieved in the Boundary Fares settlement attracted particular attention and illustrates the practical difficulties of securing meaningful participation even after a successful collective claim. The effectiveness of distribution is therefore an important measure of whether the procedural efficiencies achieved through aggregation ultimately translate into actual redress for class members.
1.5 Comparative takeaways
While all three jurisdictions have developed pathways for the bundling of competition damages claims, each system has developed distinct pathways to handle efficiency-related considerations. Each model carries advantages claimants can draw on and disadvantages they must weigh, whether in the certainty of the mechanism, the cost of testing its validity, or the resources required to run it. Both potential claimants and defendants are advised to carefully monitor the fast-paced developments around efficiency-related issues in each jurisdiction.
| Germany | Netherlands | UK | |
| Route to aggregation | Assignment to an SPV with legal services licence or joinder of parties. | Power of attorney; assignment to SPV (dominant); WAMCA opt-out. | Statutory opt-in/opt-out mechanism for competition claims only before the CAT. Other procedural mechanisms for bundling individual claims are available. |
| Nature of device | Substantive law of assignments only. | Substantive law + purpose-built regime in parallel. | Purpose-built procedural collective-redress mechanism for follow-on and standalone competition claims, including aggregate damages |
| Validity/ certification risks | FCJ has approved assignment model for cartel claims, foreign assignors and funded claims. | Assignment validity scrutinized but curable defects are not automatically fatal; WAMCA has an admissibility gate. | Certification tests representative suitability and claim eligibility; Certification acts as a filter but can itself generate substantial cost and delay. |
| Funding | FCJ has confirmed admissibility of litigation funding, limiting influence of funders. | Third-party funding is common in both opt-out and opt-in claims. Funding arrangements are scrutinised in WAMCA proceedings. | Third-party funding central to opt-out claims; funding arrangements scrutinised at certification. |
| Cost mechanics | Losing party’s costs are awarded on a fixed, statutory scale based on the procedural steps taken and the claims value band capped at €30m; bundling lowers cost per claim. | Losing party’s costs are awarded on a fixed, court-set scale based on the procedural steps taken and the claims value band. | Loser pays principle. No costs cap. Insurance is commonly used to protect against the risk of having to pay the other sides’ costs. |
| Stability of the bundle | Court may order separation mid-stream but can also combine cases for joint hearings. | Curable defects – substance over form. | Relatively stable after certification, but not irreversible: the CAT may vary or revoke a CPO, amend the class definition or alter the opt-in/opt-out basis where appropriate. |
| Central current issues | Disclosure of litigation funding agreement. | WAMCA immaturity; long admissibility phase. | Opt-in v opt-out; cost/benefits analysis of opt-out claims; funding reform; distribution of damages. |
Leave a Reply