The Tondela Judgment: More Questions than Answers?

The Court’s burgeoning case law applying competition law in the field of sports seems to me to create so many unnecessary problems. After setting out the background to the judgment in Liga Portuguesa de Futebol Profissional and others v Portuguese Competition Authority (among the competition law community, usually referred to as Tondela, the name of the first football club listed), this post discusses the superfluity of the reference and the legal puzzles the ECJ has created. [Square brackets signal the paragraph cited, citations based on a version downloaded on 4 May 2026.]

As the post is particularly critical and raises questions, a quick takeaway is provided here: agreements by rival employers that limit their capacity to compete for workers are restrictive by object. However, the specific circumstances of a case (e.g. the business model of football competitions and the Covid-19 pandemic) are relevant elements of economic context which may allow one to find that such agreements do not restrict competition by object. These circumstances are not a get out of jail free card: we need evidence to prove that they cast doubt on the anticompetitive object of the agreement. It is doubtful many cases will succeed with these claims. Even if these claims are successful a competition authority may establish anticompetitive effects; but this seems a hard path to take, in particular if the agreement is found to have certain pro-competitive features when its object is analysed.

Facts and Questions from the Portuguese Tribunal

Portugal declared a state of emergency in March 2020 owing to the Covid-19 pandemic. Football matches could not be played and the 2019/2020 sporting season would have to be extended. The Portuguese Professional Football League and First Division clubs concluded an agreement by which no club would hire a player who unilaterally terminated their employment contract citing the Covid-19 pandemic. The same agreement was entered into by some of the clubs in the Second Division. The point of this was to preserve clubs’ finances and their ability to keep playing football without their top players leaving to the richer clubs. The Portuguese Competition Authority intervened quickly to impose interim measures and in 2022 issued a decision finding that the agreement among clubs not to recruit players was a restriction of competition by object. This was appealed to the Tribunal da Concorrência, Regulação e Supervisão.

The Tribunal was clear that in ordinary circumstances an agreement among competitors not to poach or hire workers limits professional mobility and weakens the bargaining power of workers such that it should be treated as a restriction by object. However, the Tribunal was not certain about how to handle the relevant economic and legal context: sports being a special economic sector, and the Covid-19 pandemic affecting the sporting season. It sent three questions to the ECJ: Do these circumstances suggest that this is a purely sporting agreement such that competition law does not apply? Do these circumstances matter when we consider if the agreement is harmful by object? And if there is a restriction based on effects, is the agreement saved by the Wouters line of case law?

Dialogue Between the Courts

It is clear from the reference that the judges at the Tribunal know the law, so why make a reference? Indeed, many of the passages in the judgment refer to earlier case-law (notably Superleague and FIFA) which have provided legal standards to address each of the three questions. The tone of the judgment also suggests that the ECJ did not think this reference was needed. After explaining the legal framework and even suggesting how to apply it, the ECJ recalls that “it is for the referring court alone to determine, in a precise and reasoned manner,” whether the agreement is restrictive by object. [88] And, restating the proportionality standard in Wouters, the ECJ states that it is for the national court “to carry out an in-depth examination” of the conditions “in light of the arguments and evidence put forward by the parties, and also all of the relevant elements of fact and law.” [99] This seems to be a message to the national court to avoid sending references where all that is needed is to apply well-established law to the facts.

One final element of the ECJ’s response is notable. The first question is about what falls in the category of purely sporting rules, like those establishing ranking criteria to select athletes for certain competitions. This question is answered in the ECJ’s preliminary observations [29-34] and not, as the ECJ writes, in providing a joint answer the questions 1 and 2 [90-100]. This poor structuring is unhelpful when the Court is asked to provide clarity in a setting where the main difficulty is finding out in which legal box certain pieces of evidence fit.

Two Types of Object Restrictions?

The Court has now settled on two subsets of object restrictions, as seen already in Superleague. The first are “certain types of coordination” which are “harmful by their very nature.” [39] This includes “primarily, certain forms of collusive conduct which are particularly harmful to competition, such as horizontal cartels leading to price-fixing, limitations on production capacity or allocation of customers.” [40] The second are other types of conduct which are restrictive by object “without necessarily being equally harmful to competition.” [41] For this unspecified class “it is necessary to demonstrate that such conduct carries a sufficient degree of harm to competition, meaning that they must have a manifest anticompetitive rationale.” [41] I have no idea what is to be gained by this bifurcation.

Indeed, this taxonomy is rendered irrelevant when the Court the explains that to determine object “it is necessary to examine, first, the content of the agreement, decision or practice in question; secondly, the economic and legal context of which it forms a part; and, thirdly, its objectives.” [43] It would be better if the Court were just to state this as the essence of an object inquiry.

Perhaps the reason for the bifurcation is that it allows the Court to say that when the agreement is “particularly harmful to competition” then the analysis of the relevant economic and legal context “may be limited to what is strictly necessary to establish a restriction by object” [47] while for “other types of conduct” one needs “a more in-depth examination” including of the relevant regulatory and institutional framework. [48] This approach is flawed. The whole point of the economic and legal context analysis is precisely to decide if the agreement is anticompetitive by object because the content of the agreement is always insufficient to make that determination. The amount of economic and legal context analysis required is best seen as a sliding scale: with more evidence required the more welfare-ambiguous the conduct is. And recall that the defendant is entitled to bring forward evidence to cast doubt on a finding of a restriction by object which the competition authority may not have considered and which the authority will have to engage with. In other words, the determination of an object restriction should be seen as an inquiry where the amount of economic context required is a function of how much evidence is needed to have a firm conviction that the agreement restricts competition by object. This approach may be seen in the judgment itself where the Court explains that the examination of the economic and legal context makes it possible to ensure whether the agreement is sufficiently harmful to justify a finding that it restricts competition by object [49], a passage which sits oddly with the preceding analysis.

Moreover, the ECJ is even confused as what types of agreement fall in which of the two categories. Agreements “leading to competing undertakings being excluded from the market” [41] are in the second category but cartels leading to “the exclusion of potentially competing undertakings” [47] are in the first category. It is not clear why excluding potential competition is more harmful than excluding actual competition and even if the ECJ has this distinction in mind. Perhaps the distinction is between cartels and other agreements, but the definition of cartels in the Damages Directive (Article 2(14)) is indeterminate.

Agreements Harming Competition for Workers

The ECJ agrees with the referring court that an agreement not to recruit players, like a no-poach agreement, may restrict competition by object. It would have been helpful to understand better what economic context allows for this finding. For example, the ECJ refers to the fact that the agreement affects “two high-level national competitions in a Member State,” [53] In my view, this is important because it indicates that market power matters (nearly all clubs entered into the agreement and football players have limited exit options other than plying their trade in another club). What if there is a competition with 20 teams and two teams agree not to hire each other’s players? Is this an object restriction? Or suppose the relevant geographical market is Brussels and two law firms (among the hundreds that may be found there) agree not to hire each other’s secretarial staff. Is this an object restriction when the staff in question have many exit options even outside of the law firms?

Some of the Court’s case law (e.g. Expedia) would suggest an affirmative answer to the two questions posed above based on the terms of the agreement. This is why the economic and legal context is vital: to reveal that however bad an agreement sounds (no-poaching), what matters is its impact on allocative efficiency. But the case-law provides no such clarity and suffice it to say that absent any guidance by the ECJ, one may infer that all agreements between employers not to poach or hire each other’s workers are restrictive by object. This is an overly inclusive approach. See on this Friedrich Preetz’s post suggesting more nuance is needed.

Agreements Harming Competition for Recruiting Athletes During a Pandemic

While sports teams compete to recruit the best players they can afford, their economic success depends on the “proper functioning, sustainability and success” of the competitions they participate in and this requires that there is “a certain equality of opportunity as between the participating professional football clubs, given the interdependence that binds them together.” [60] This may be news to followers of the first football division in Portugal where three teams have won all but two championships since 1936. But the legally salient point is the following: an agreement which harms workers might well be saved if it is shown that the same agreement ensures the proper functioning of the football ecosystem.

To help the national court make sense of this point, the Court clarifies that neither Covid-19 nor Article 165 TFEU justify excluding the application of competition law to the facts at hand. [74-75] However, the impact of the Covid-19 pandemic created a risk that players would quit a club which chose to reduce wages and be “freely hired by another club, which would have inevitably and significantly altered the composition of the various teams involved, thereby undermining the integrity of competition.” [73] And this would have particularly affected poorer clubs who may not be able to stop an exodus of players and be unable to sign new ones. This is a factor which must be taken into account as part of the context to understand if there is a restriction of competition by object.

The Court insists that this is a “competition-based analysis” [78] and that the agreement may pursue “in parallel, an objectively anticompetitive aim, consisting in the restriction of competition on the player recruitment market and an objectively procompetitive aim, consisting in ensuring stability of player rosters playing in the First and Second Divisions.” [86, my translation here, the text online reads ‘pro-competition’. Pro-competitive aligns with the French version and with Super Bock [36], even if there the Court speaks of procompetitive effects.] What would have been useful is that the Court explained what the consequences would be of such a finding. Is the referring court expected to say that if this is so then the agreement should be assessed having regard to its effects? But what if there could be a less restrictive means of achieving the procompetitive aim? Could this still be treated as an effects case and proportionality be left for that stage of analysis, or should it be considered as part of the object assessment?

There is one final point which may well require further clarification. The Court explains that the conduct does not result from rules issued by the football association [70] but that the agreement “was adopted in the presence of the President of the association and may have “been supported, and even endorsed” by them. [72] The national court is advised to investigate this, but it is not clear why this should matter. If the Court is going to say that context matters, then it should also go further and explain the legal salience of this element. In what way does the support of the regulator make a difference? Is it a matter of attribution? Does it serve to support a finding that there is no object restriction? Recall that if we were dealing with a decision of the association then the governance criteria found in Superleague would have to be considered to determine if there was a restriction of competition.

Pro-competitive Effects Versus Legitimate Objectives in the Public Interest

As we know, the Court is of the view that if there is a restriction by effect, then the defendant may either plead a legitimate objective or that the conditions for the application of Article 101(3) TFEU are met. Conversely, if there is a restriction by object, the defendant may only make arguments to try and meet the conditions of Article 101(3) but cannot claim that the agreement is justified by a legitimate objective because in an object restriction the degree of harm “is too great for it to be regarded as justified and proportionate.” [92] As I argued, this position is not convincing but is now settled.

And we also know that if an agreement pursues a pro-competitive aim, that this is evidence that can lead to a finding that conduct is not restrictive of competition by object. The ECJ is clear that there is a vital difference between finding that an agreement pursues a pro-competitive aim (relevant to determine the object of the agreement) and a finding that the conduct pursues a legitimate objective (relevant to trigger the Wouters defence in case of agreements restrictive by effects). [87] But: what is the difference between a pro-competitive aim and an objective in the public interest? According to this judgment “ensuring the regularity of sporting competitions” which may justify “rules intended to ensure the maintenance of a certain degree of stability in club’s player rosters” [96] is a public interest objective, but I can see no difference between this and the pro-competitive aim of  “ensuring the stability of player rosters.” [88] Some work is needed to determine the precise contours of the legitimate objective defence.

And I am left to wonder what is left of Article 101(3) TFEU when pro-competitive criteria and public interest considerations are relevant for an assessment of Article 101(1).

In sum, this judgment reveals that some tidying up of certain elements of the Article 101 TFEU analysis is warranted. I doubt whether better questions from national courts would help the process, however. Dare we ask for Guidelines?

Giorgio Monti, Professor of Competition Law, Tilburg University and Research Fellow, CERRE. According to the ASCOLA Transparency and Disclosure Declaration I have no conflicts to disclose.