On 1 December 1986, the so-called “Balladur” Ordinance on freedom of prices and competition[1] introduced, in Article 8, paragraph 2, a new legal concept (the abusive exploitation of a state of economic dependence) intended to protect weaker parties against the abusive use of their partners’ economic power. The legislature was freeing prices, but at the same time intended to equip the competitive order with a safeguard against the structural imbalances that market forces alone could not correct.
Six months before the fortieth anniversary of this founding text, the verdict is harsh: Article L. 420-2, paragraph 2, of the French Commercial Code, which took over its substance, has never really moved beyond the status of an unkept promise. The set of four cumulative criteria inherited from the 1993 Cora decision[2] had gradually rendered it inoperative, as the French Competition Authority (Autorité de la concurrence) itself acknowledged in its opinion of 31 March 2015 on the consolidation of purchasing alliances[3], without any legislative follow-up. Academic commentators have since regularly remarked on this, whether resigned or critical depending on the author, but with no illusions about the practical prospects of an action brought on this basis.
Two concurrent developments, a few days apart in spring 2026, have significantly altered this balance:
- First, on 13 May 2026, the Commercial Chamber of the French Supreme Court (Cour de cassation), following on from the Apple case, handed down a judgment which, without abandoning the traditional criteria, substantially rewrote how they are applied[4].
- Next, the Senate report of 19 May 2026 on the margins of manufacturers and large-scale retailers[5], adopted six days later, invited the legislature to recast the very definition of the concept (proposal no. 12, which the committee expects to feed into a future legislative vehicle).
“The combination of these two developments reopens, for suppliers to large-scale retailers and their lawyers, a ground for litigation that had come to be regarded as barren. The irony of history is worth noting: just as we commemorate the text that laid down its principle, it appears, for the first time in forty years, capable of producing its intended effects.”
This article sets out to draw the practical consequences. It first revisits the previous framework, an understanding of which remains necessary to analyse the recent shifts (I); it then examines what the Commercial Chamber actually changed in its judgment of 13 May 2026 (II); finally, it proposes a set of arguments that can be deployed in written submissions now, without waiting for the legislative overhaul envisaged by the Senate report (III).
I. The 1986 concept and the case-law lock of the four criteria
Abuse of economic dependence derives from Article 8 of Ordinance No. 86-1243 of 1 December 1986 on freedom of prices and competition, the adoption of which was one of the most formative acts of contemporary French economic law.
By breaking with the system of administrative price control inherited from the Ordinance of 30 June 1945[6], the legislature established the principle of freedom of pricing; but at the same time it introduced two corrective mechanisms designed to prevent restored competition from turning into predation: the prohibition of abuse of a dominant position, on the one hand, and the prohibition of abusive exploitation of economic dependence, on the other.
The first was based on market power; the second was based on the contractual relationship and was aimed precisely at situations in which, without any operator being dominant in the EU law sense, a partner’s economic subjection to its counterparty enabled the latter to exploit that situation abusively. It is this second limb that came to be known, in the language of the Ordinance, as “abuse of economic dependence”.
Under paragraph 2 of Article L. 420-2 of the French Commercial Code, which preserves its wording in the text currently in force, the abusive exploitation by an undertaking or group of undertakings of the state of economic dependence in which a customer or supplier undertaking finds itself in relation to it is prohibited, where it is liable to affect the functioning or structure of competition.
The text expressly refers, by way of example, to refusal to sell, tied sales, the discriminatory practices referred to in Articles L. 442-1 to L. 442-3 and range agreements, a non-exhaustive list that the courts have supplemented case by case.
A conceptual clarification is called for, which sheds light on its distinctive nature. Unlike abuse of a dominant position, abuse of economic dependence is assessed not by reference to an undertaking’s position on a given market, but in the light of the specific features of the commercial relationship it has with its upstream or downstream partners. It is this particular feature (which makes it possible to address situations in which no operator could be described as dominant within the meaning of “mainstream” competition law but where contractual imbalance nonetheless leads to abusive exploitation) that makes the prohibition useful for the retail sector, where buying power can produce effects equivalent to those of a dominant position without formally displaying its characteristics.
On this basis, since the Cora decision and the subsequent judgments of the Commercial Chamber handed down between 1993 and 2004, a finding of economic dependence rested on four cumulative criteria:
- The reputation of the dominant supplier’s brand;
- The size of its market share;
- The size of its share in the dependent party’s turnover;
- The absence, for the dependent party, of an equivalent solution that could be implemented within a reasonable time.
This last criterion, regarded as decisive by both the Authority and the courts, was given such a restrictive reading that, in practice, it amounted to requiring proof that switching was almost technically impossible, a requirement that almost all complainants failed to meet.
The outcome of this construction barely changed over the following two decades. With the exception of a few sector-specific cases, the prohibition remained largely unapplied, as the French Competition Authority acknowledged in its opinion of 31 March 2015, noting that the need for the victim to establish the absence of an equivalent solution led, in the vast majority of cases, to the failure of the action.
Several possible developments had been suggested at the time: assessing the criteria in a more case-specific manner, giving central importance to the actual ability of the operator allegedly in a situation of dependence to implement an alternative solution within a reasonable time, and broadening the time horizon of effects to include the consequences expected in the medium term.
None of these proposals had been taken up by the legislature, however, and the 1986 text had continued on its static course until the Apple sequence revived its examination.
II. The courts’ reinterpretation of the concept of abuse of economic dependence
The judgment handed down by the Commercial Chamber on 13 May 2026, and published in the Bulletin and the Annual Report, follows directly on from the decision of the French Competition Authority of 16 March 2020[7] and the judgment of the Paris Court of Appeal of 6 October 2022[8] which, without overturning the finding of abuse of economic dependence against Apple vis-à-vis its Apple Premium Resellers, had substantially reduced the penalty.
The dismissal of the appeals confirms a finding of abuse of economic dependence in circumstances that go well beyond the case at hand, and that is why the judgment deserves careful reading by lawyers acting for suppliers to large-scale retailers.
The traditional four-criteria test is not abandoned (the Court expressly restates it), but its application is reconsidered around three shifts which significantly alter its practical scope.
The most immediately noticeable shift concerns the equivalent solution test, which now expressly takes into account the cost and time of switching: the absence of such a solution is established where switching would be particularly costly financially and could not be implemented within a reasonable time. The criterion thus ceases to be reduced to the theoretical existence of an alternative partner and shifts towards the concrete possibility, economically and operationally, of making use of one.
For a supplier whose production line has been calibrated for a retailer’s own brand, whose logistics flows have been configured on a proprietary EDI system, whose sales force has been sized to serve a single outlet and whose production equipment has not been depreciated, proving dependence becomes achievable, whereas until now it remained out of reach.
In addition, the alternative market criterion has been recast. The existence of other suppliers that lead in terms of market share is no longer sufficient to rule out dependence, where none of those suppliers offers a comparable distribution model. The test becomes qualitative where previous practice was content with a quantitative comparison, and it covers all the characteristics that make, or do not make, the alternative partner an economically viable option: product range, reputation, territorial coverage, logistics requirements, capacity to absorb volumes.
In the large-scale retail sector, where the main retailers, through buying alliances, apply structurally equivalent terms, this shift in analysis changes the equation for suppliers, who can now demonstrate without excessive difficulty that competing retailers are not, in reality, substitute solutions.
It is, however, in the very definition of abuse that the most operative shift lies. The Court reiterates that the dominant undertaking must, failing which it commits an abuse, take care not to impose on its dependent partner measures that any rational undertaking would refuse and which the partner accepts precisely because of its state of dependence. The approach is no longer merely case-based (refusal to sell, tied sales, range agreements) but becomes behavioural, making it possible to include any practice whose economic rationale can only be conceived on the basis of an entrenched asymmetry.
The Court also confirms, in line with the 2022 appeal judgment, that the effect on the functioning or structure of competition need not be appreciable: it is sufficient that it is liable to weaken the distribution channel concerned or the competitive dynamics that sustain it. In this way, the 1986 text regains its original ambition: that of an effective safeguard against the abusive exploitation of economic subjection, to which the case law of the first three decades had unwittingly denied any useful expression.
The attention that the French Competition Authority is paying to this ground is not, moreover, an isolated development. The decisions handed down between the end of 2024 and spring 2025 in the context of merger control in the large-scale retail sector (on the acquisitions of Casino group stores by Intermarché, Carrefour and Auchan[9]) all examine, in now-standardised terms, the risks of increased buying power liable to place certain suppliers in a situation of economic dependence vis-à-vis the new entities.
The Apple sequence, the Commercial Chamber’s renewed test and the regulator’s increased attention thus converge towards an effective revival of a ground that suppliers had learnt to stop invoking.
III. Practical brief: the arguments available in support of abuse of dependence
Proposal no. 12 of the Senate report of 19 May 2026 invites the legislature to recast the definition of abuse of economic dependence along the lines of the suggestions made in 2015. This would be useful, insofar as it would enshrine in statute a development already under way in the case law; it is not, however, a prerequisite for litigation. The judgment of 13 May 2026 opens up, without any change in the law, a space that suppliers’ lawyers can occupy without waiting for legislative implementation, provided they deploy it methodically. In our view, seven lines of argument make up the useful toolkit.
The first is to present the four traditional criteria to the court as components of a body of evidence, rather than as a cumulative validation test. By rewriting how the criteria are applied, the Commercial Chamber allows a combined and weighted reading, in which each indicator contributes to the overall demonstration without any one of them being an obstacle on its own. The supplier’s lawyer must build the case around the share of turnover, of course, but also around the length of the relationship, the specific investments made, the logistical and informational dependence created by the architecture of the partnership, and the absence of an alternative outlet with a comparable model; the weighting of the whole matters more than the mechanical addition of the elements.
The second is to document in advance the specific investments made to serve the dominant retailer, which the Commercial Chamber’s judgment has made a central matter of evidence. Dedicated production lines, equipment depreciated over the life of the relationship, proprietary EDI configurations, specific sales training, custom packaging: every expense incurred on the retailer’s behalf supports the demonstration of the cost and time of switching.
Without contemporaneous and precisely quantified documentation, dependence becomes theoretical again and the benefit of the shift in case law is lost; conversely, a bundle of evidence prepared in advance (which requires suppliers to adopt an internal traceability policy, the introduction of which should be recommended) changes the impact of the written submissions.
The third is to use the qualitative test on alternative retailers. The task is no longer to prove that no other retailer exists, a demonstration the courts rightly considered unrealistic, but to establish that competing models are unsuited to absorbing, within a useful timeframe, the volume needed for industrial viability.
No alternative retailer offers the same national coverage, the same stock turnover and the same range logic; the buying alliances of the main retailers apply structurally equivalent terms; alternative channels (farm shops, specialist e-commerce, convenience retail) are not, as the market stands, able to absorb the redeployment of an ousted supplier.
On this point, the Senate report provides a public factual benchmark whose evidential value is set to grow as it is relied on in the case law.
The fourth concerns characterising the abuse by reference to the behavioural standard laid down by the Commercial Chamber. In written submissions, each of the practices complained of should be presented as a measure that any rational undertaking would refuse, and which the partner accepts only because of its state of dependence.
The practices documented in the Senate report lend themselves particularly well to this characterisation: mass order suspensions in the middle of negotiations, international services whose disproportion has no commercial justification, so-called “top to top” meetings invoiced at several hundred thousand euros, multi-retailer alignments imposed without consideration, indirect challenges to agricultural raw material costs by way of discussions on industrial raw materials. Each of these practices can be mapped, without artifice, onto the standard spelt out by the Court.
The fifth is to combine the legal grounds. The Paris Court of Appeal, in its judgment of 6 October 2022, expressly endorsed the idea that the scope of Article L. 420-2 may include practices also falling under Article L. 442-1[10].
The practical consequence of this combination is significant: the same conduct may be characterised simultaneously as abuse of economic dependence and as a restrictive trade practice, which reduces the risk of recharacterisation, broadens the range of penalties incurred (civil and competition law) and makes it possible to combine an order to cease the practices with compensation for loss and a civil fine.
The sixth concerns the lowering of the threshold relating to the effect on competition. The practice need not erode market share on a national scale: it is sufficient that it is liable to weaken the distribution channel concerned or the competitive dynamics that sustain it. For an SME supplier whose delisting jeopardises its industrial viability, or for a sector whose profitability is being eroded by pressure from a European purchasing alliance, this threshold becomes attainable.
The Senate report expressly states that current practices are weakening the entire agri-food industrial fabric, and the evidential demonstration, which until now had to be carried out alone, can henceforth rely on this parliamentary finding.
The seventh and last concerns the procedural combination, which in practice determines the effectiveness of the strategy. A complaint to the Authority remains the reference route for large-scale practices, as confirmed by the competitive assessment of the Concordis and Aura buying alliances opened on 9 January 2026 on the basis of Article L. 462-10, II, of the French Commercial Code[11]. In parallel, a civil action before the Commercial Court (tribunal de commerce) makes it possible to combine the grounds under Articles L. 420-2 and L. 442-1, to obtain an order to cease the practices and compensation for loss and, where appropriate, to have the Minister for the Economy join the proceedings as an intervener.
Finally, joint action with the DGCCRF (the French consumer protection and fraud control authority) remains useful for practices that may be characterised administratively, subject, as the Senate report points out, to the availability of evidence and to documentation from the injured suppliers, the quality of which ultimately determines the success of the action brought.
***
While the Senate report of 19 May 2026 clearly sought to mark a political milestone by proposing to recast the definition of abuse of economic dependence, the judgment handed down six days earlier by the Commercial Chamber had already, more discreetly, covered part of the ground. The legislature remains free to enshrine in statute the relaxation brought about by the judges, or to leave it to the case law to roll out the test case by case. The question remains open, and the timetable for its resolution will depend as much on the parliamentary momentum that follows the report as on the robustness of the decisions handed down in its wake.
For suppliers and their lawyers, on the other hand, there is no such need to wait. Since 13 May 2026, abuse of economic dependence has once again become an actionable ground, provided that its test is applied with the rigour that the case law now requires. Documentation of specific investments, the qualitative test on alternative retailers, the behavioural standard, combination with Article L. 442-1 and careful procedural coordination together form the new toolkit available to injured manufacturers.
***
As the fortieth anniversary of the Ordinance that laid down its principle approaches, the 1986 text appears, for the first time since its adoption, on the verge of producing the useful effects its drafter expected of it; and it is the suppliers who take the initiative in the coming months who, through their written submissions, will have helped to put it to the test.
[1] Ord. No. 86-1243 of 1 December 1986 on freedom of prices and competition, Art. 8, para. 2
[2] Cons. conc., dec. no. 93-D-21 of 8 June 1993, Groupe Cora.
For the leading judgment establishing the four cumulative criteria test: Cass. com., 12 Oct. 1993, nos. 91-17.090 and 91-16.988;
For the confirmation of the Cora decision on appeal: Cass. com., 10 Dec. 1996, no. 94-16.192.
[3] Aut. conc., opinion no. 15-A-06 of 31 March 2015 on the consolidation of purchasing and listing alliances in the large-scale retail sector, paras. 261 et seq.
[4] Cass. com., 13 May 2026, nos. 22-22.623, 22-22.647, 22-22.677 and 22-22.727 (joined).
[5] French Senate, Report on behalf of the committee of inquiry into the margins of manufacturers and large-scale retailers, no. 632 (2025-2026), Volume I, 19 May 2026.
[6] Ord. No. 45-1483 of 30 June 1945 on prices, repealed by Article 56 of the aforementioned Ordinance of 1 December 1986.
[7] Aut. conc., dec. no. 20-D-04, 16 March 2020.
[8] CA Paris, pôle 5, ch. 7, 6 Oct. 2022, No. 20/08582, Société Apple Inc. et al.
[9] In particular, Aut. conc., Decision No. 24-DCC-255, 28 Nov. 2024; No. 24-DCC-288, 13 Dec. 2024; No. 25-DCC-56, 13 Mar. 2025; No. 25-DCC-65, 21 Mar. 2025.
[10] C. com., art. L. 442-1, I, new 1° and 2.
[11] Aut. conc., press release of 9 January 2026 opening, for the first time since the mechanism was introduced, a competitive assessment procedure (procédure de bilan concurrentiel) in respect of the AURA and CONCORDIS purchasing alliances.
