One of the few certainties of European administrative law is that, when a decision entails complex technical, economic, or scientific assessments, and its author enjoys broad discretion accordingly, judicial review is limited to manifest errors of appraisal. Yet, this does not apply to the review carried out by the Boards of Appeal (BoAs) of EU agencies.

It should be recalled from the outset that all EU agencies that adopt binding decisions are equipped with BoAs. These are internal bodies competent to hear “appeals” brought by natural and legal persons against such decisions. Following their review, BoAs adopt a new decision which, depending on the case, may confirm the contested one, replace it or refer the case back to the respective agency. The decisions taken by the BoAs can then be challenged before the EU General Court (GC).

In its Aquind judgment of March 2023, the EU Court of Justice (ECJ) held that, when examining EU agencies’ decisions entailing complex assessments, BoAs must not confine themselves to a limited review, but must proceed to a “full review” (paras. 63, 73).

According to the ECJ, while BoAs “perform quasi-judicial functions through adversarial proceedings”, they are also composed of technical experts, being therefore equipped with the necessary expertise to conduct assessments of complex facts (paras. 59, 64). The intensity of the review of EU agencies’ decisions entailing complex assessments carried out by the BoAs must therefore be greater than the limited review by the EU courts. Otherwise, the GC would be performing a limited review of a decision resulting from a limited review, thereby failing to ensure the effective judicial protection of the rights of the parties concerned (paras. 65-67).

While the message delivered by the ECJ in Aquind was loud and clear, it was less clear what it implied in practice and to whom it was addressed. More specifically, the judgment did not clarify what such full review consisted of, nor whether it applied to all BoAs (the Aquind judgment being binding only for the BoA of the European Union Agency for the Cooperation of Energy Regulators—ACER). It remained to be seen also to what extent all BoAs were capable of performing such a review.

The special issue of the Rivista del contenzioso europeo “Quo vadis, Boards of Appeal?”, edited by Jacopo Alberti, has, among others, the merit of trying to answer these questions, and of doing so by directly involving the members of all BoAs, thus complementing, from an insider’s perspective, the research already conducted before the Aquind judgment (Chamon–Volpato–Eliantonio, Greco, and Marchetti).

I was therefore particularly happy to accept the invitation of the REALaw blog to discuss the findings of this special issue. This is all the more timely since the ECJ is expected to rule again soon on the matter, in the joined cases Polskie sieci elektroenergetyczne. In the remainder of this contribution, I will try to reflect, starting from these findings, on the past and present of BoAs’ review, and then draw some tentative conclusions about its future.

1. The search for a common (judicial?) standard

The establishing regulations of EU agencies equipped with BoAs do not define the standard of the review that these bodies must carry out (Oosterhuis, p. 109). In the absence of specific provisions and guidance from the EU courts, the intensity of BoAs’ scrutiny was therefore initially defined by the BoAs themselves, in two different ways.

The BoAs of EU agencies operating in the field of industrial property (i.e. the European Union Intellectual Property Office (EUIPO) and the Community Plant Variety Office (CPVO)) exercised an in-depth review of the contested decisions, thus proceeding with a new assessment of the case, independent of, or in any case not limited to, the arguments and evidence raised by the appellant (Humphreys, pp. 186-7). On the contrary, all the other active BoAs (i.e., those of the European Chemicals Agency (ECHA), the European Union Aviation Safety Agency (EASA), the ACER, the Single Resolution Board (SRB), and that of the European Banking Authority (EBA), the European Securities and Markets Authority (ESMA) and the European Insurance and Occupation Pensions Authority (EIOPA)) essentially confined themselves to verifying whether the respective agency’s decision was vitiated by a manifest error of assessment.

It is to be noted that this dichotomy did not depend on the number of appeals brought before the two categories of BoAs. Paradoxically, this number was (and still is) in fact proportionally (i.e. per BoA member) much higher for those BoAs that proceeded to a full review of the contested decisions. What emerges from the special issue is rather that this difference in approach is attributable to other ‘institutional’ factors.

These factors are, first, the availability of sufficient resources to enter into the technical merits of the contested decisions within the tight deadlines dictated by the applicable procedural rules. Only BoAs that had full-time members and were supported by registry and research services were able to perform an in-depth review (Navin-Jones, pp. 198-202). Second, the ‘judicial attitude’ of the BoAs to their function. The BoAs tended to apply the same standard of review applied by the EU courts in their respective fields, which was indeed full, in industrial property matters, and limited, in the other sectors (Oosterhuis, p. 109).

Interestingly enough, the case of the ECHA BoA, which went from full review to the so-called limited review (Oosterhuis, p. 114-5), despite being composed of full-time members, seems to show that the second factor weighed more than the first and that the BoAs were essentially united in their search for a common (yet different) judicial standard of review.

2. “Full review”: what’s in a name?

While the Aquind judgment has disavowed those BoAs that carried out a limited review, the ECJ did not explicitly clarify what the requested “full review” entailed, nor how it differed in practice from the limited review carried out by EU judges.

In this respect, two considerations emerging from the special issue seem particularly relevant. First, it is noted that “full review” within the meaning of the Aquind judgment is a “middle ground” between, on the one hand, the limited review exercised by EU judges and, on the other hand, a fully-fledged re-examination of the contested decision, based on a de novo assessment of all the technical considerations on which it is based (Oosterhuis, pp. 107, 117-8; Buchet, pp. 230, 235; Sánchez Rydelski, p. 216; see also Alberti, pp. 327-8). This is essentially because, as underlined by the ECJ itself in Aquind (para. 59), the appeal procedure before the BoAs is and remains an adversarial procedure, limited as such to the examination of the arguments and evidence put forward by the appellants.

As the GC already held in BASF (para. 89), and recently reaffirmed in TenneT TSO (para. 176), full review therefore seems to mean that BoAs must examine whether these arguments and evidence are capable of demonstrating that the considerations on which the contested decision is based are vitiated by error (Oosterhuis, p. 121; Buchet, p. 23). An implicit confirmation of the position expressed by the GC can be found in the Aquind judgment (paras. 69-70). It remains to be seen whether an express one will arrive in the forthcoming ECJ judgment in the joined cases Polskie sieci elektroenergetyczne, even though the Advocate General’s (AG) opinion did not delve into this issue.

Second, the special issue also highlights that it is not easy to distinguish a full review so defined from the “process-oriented review” (Vos, Lenaerts) that EU courts are now called upon to perform when reviewing decisions entailing complex scientific, technical, or economic assessments.

In such a context, EU courts are in fact called upon to verify whether there has been a manifest error of assessment or a misuse of powers, or whether the author of the contested decision has manifestly exceeded the limits of its discretion (cfr. Crédit lyonnais, para 55 and PlasticsEurope, para. 46). While they cannot substitute their assessment of scientific and technical facts for that of the institutions that adopted the act, EU courts must nonetheless establish whether the evidence relied on is factually accurate, reliable, consistent, comprehensive and conclusive (see Crédit lyonnais, para. 56).

Admittedly, if the only difference between judicial and administrative review resides in the gravity of the error of assessment they can sanction (manifest for the EU courts, “simple” for BoAs), there is indeed the risk that the distinction between full and limited review becomes formalistic (Sánchez Rydelski, p. 216). What can really distinguish these two reviews, and make the second “full”, is rather the possibility of substituting the BoA’s assessment of the scientific and technical facts for that of the agency, of course to the extent that this assessment is contested by the appellants before the BoA (see also Tovo, pp. 7-9). It is precisely this, however, that several BoAs do not yet seem willing to do.

3. “Full review” is the new black?

The BoAs of ACER and ECHA appear to have embraced the new standard of review established by the ECJ in Aquind, at least in principle (Buchet, p. 230 ff.; Prek, pp. 241, 244; see also Biondi, p. 64). After all, this has already happened since the GC judgments in the BASF and Aquind cases, respectively, delivered in 2019 and 2020 (Oosterhuis, pp. 123-4). Signs of an intensification of the review exercised by the ACER BoA could be seen even before this second ruling (Tovo, pp. 53-4) and can be found in other BoAs, starting with that of EASA (Sánchez Rydelski, p. 211 ff.), which have thus come closer to the “deep scrutiny” that those of EUIPO and CPVO continue to carry out (Humphreys, p. 194).

However, other BoAs, in particular those of SRB and ESAs, have instead shown a certain reticence in adopting the Aquind standard of review so far. On the one hand, they object that they are not in functional continuity with the respective agency. On the other, they argue that their scrutiny, essentially equivalent to the “process-oriented review” carried out by the EU courts, already ensures a full review of law and facts, with the sole limited exception of those considerations that fall within the agency’s margin of appreciation (Lamandini-Ramos Muñoz, pp. 263-273). While the first argument has recently been refuted by the GC (Gaz-System, paras. 34-37), the second instead testifies to what was said previously on the difficulty (and the need) of drawing a dividing line between limited and full review.

Even assuming that all BoAs apply the Aquind standard, the question also remains whether they are all in a position to ensure compliance with this standard. In this respect, the abovementioned distinction between BoAs composed of full-time members and those that operate on an ad hoc basis seems to re-emerge.

For the former BoAs, Aquind does not seem to pose too many challenges. For the latter BoAs, the ruling highlights two structural limits to its full reception, linked to the resources and guarantees available to exercise their full review (Navin-Jones, pp. 198, 201-7; see also Sánchez Rydelski, p. 218-9, Buchet, p. 236, Prek, pp. 242-244, and Verslype, p. 280).

As to the resources, non-full-time members are paid on a piecework basis, often for a predetermined number of hours per case, which is generally insufficient to ensure the level of instruction of the appeal required by Aquind. For some BoAs, this adds to an increase in the number and/or difficulty of cases, in the face of rather stringent procedural deadlines. Although the ECJ made it clear in Aquind (paras. 70-71) that whether BoAs’ members are employed full-time or not is irrelevant in determining the intensity of their review, the type of working relationship between the agency and the members of its BoA seems to remain a decisive factor in reaching that level of intensity.

This factor is closely linked to the second structural limit highlighted by the special issue. While enjoying significant guarantees of autonomy and impartiality under EU agencies’ establishing regulations, BoAs’ members who are not employed by the agency do not benefit from the functional immunities granted to EU civil servants by Protocol No. 7 and the Staff regulations. According to concerned BoAs members, this could affect their ability to conduct a full review, as it would expose them to damages and disciplinary actions in case of failure to comply with this standard. One may also wonder whether this is compatible with the requirement that the BoAs are “independent”, enshrined in Article 58a of the CJEU Statute.

All in all, while scholars tend to agree that the Aquind standard of review is intended to apply, or should be to applied, to all BoAs (Biondi, p. 63, Stefan, Greco, p. 254 ff., and Tovo, p. 57), the special issue has presented a more varied reality.

4. Quo vadis “full review”? A question of method, money… and scope

The Aquind judgment clarified, once and for all, that, as “administrative revision bodies”, BoAs must provide a “quick, accessible, specialised and inexpensive” “full review” of EU agencies’ decisions. Also considering the findings of the special issue in comment, however, the future (and success) of this full review, in my view, depends on three issues.

The first is a question of method: for the full review to be applied consistently and uniformly by the BoAs, it must be clear what it consists of. While the adversarial nature of the procedure clearly distinguishes the full review from a de novo assessment (i.e. a re-examination of the contested decision), the mere gravity of the errors does not suffice to clearly distinguish it from the limited review performed by the EU judges. To this end, the EU courts should clarify that full review implies that BoAs can (shall?) substitute their own assessment of the technical facts for that of the agency, to the extent that this assessment is contested before them.

The second issue is money. To make sure that the full review is also an effective review, it is necessary to put the BoAs in a position to ensure an impartial, fair, and reasonably short review. Since BoAs are administrative review bodies, the effectiveness of such review should in fact be measured by compliance with these principles, set out in article 41 of the Charter of Fundamental Rights of the EU (CFR) (Siri, p. 257, Lamandini-Ramos Muñoz, pp. 263-273, but also Bolzonello, p. 574, Ritleng, pp. 312-5, and Biondi, p. 65), among which the duty to give reasons stands out, as it is the precondition for verifying compliance will all the others (see also TenneT TSO, paras. 185-6). The more thorough the review, the more resources and support the BoAs must have to carry it out effectively. It is therefore also up to the EU agencies and the EU institutions to ensure that the full review does not remain on paper.

Finally, the third issue concerns the completeness of the full review, the real ‘elephant in the room’ of the BoAs system. Is a review full if it is not complete? Yet, many decisions taken by EU agencies currently escape the jurisdiction of their BoAs and have to be directly challenged before the GC (see also AG’s opinion in ABLV, para. 73). Of course, extending the jurisdiction of the BoAs means limiting that of the ECJ in appeal, under article 58a of the CJEU Statute. However, effective judicial protection under article 47 CFR does not require a right of access to a second level of jurisdiction (Scuola europea di Varese, para. 93). Aquind seems instead to indicate that article 47 CFR requires all EU agencies decisions to be subject to full review before the BoAs. If the ECJ has somehow spoken, it remains to be seen whether the GC will decide to have its say in the pending case RWE Supply & Trading.

—

Posted by Carlo Tovo, Researcher in EU law—University of Bologna