Introduction

Workload and length of proceedings have always been the main driver of each reform of the EU Judiciary. The growing competences of the former European Communities and the further enlargement in the early 2000s prompted a debate on the structural changes which the European judicial architecture ought to undergo in order not to succumb to its own success (A. Dashwood, A. Johnston, p. 56; B. Nascimbene, p. 107). Such a priority led to the adoption in a brief timespan of several reforms of the Protocol No 3 on the Statute of the Court of Justice of the European Union (Regulation (EU, Euratom) 2015/2422 of the European Parliament and of the Council of the 16 December 2015; Regulation (EU, Euratom) 2016/1192 of the European Parliament and of the Council of 6 July 2016; Regulation (EU, Euratom) 2019/629 of the European Parliament and of the Council of 17 April 2019; Decision of the Council 2022/0906(COD) of 6 March 2024 adopting the amendments to Protocol No 3 on the Statute of the Court of Justice), whose rationale and solutions have sparked off scepticism, if not open criticism (on the doubling of the General Court’s members, F. Dehousse, p. 55; A. Alemanno, L. Pech, p. 157; M. Derlén, L. Lindholm, p. 12; on the transferring to the General Court of the competence to render preliminary ruling on certain subjects, C. Amalfitano, p. 39).

Against this backdrop, the role of EU Agencies’ Boards of Appeal (hereinafter, BoAs) is often overlooked. However, the conclusion of the experience of specialised Tribunals established pursuant to Article 257 TFEU and a renewed characterization of the General Court (hereinafter, “GC”) as a generalist instance caused the request for specialised review to be channelled through the activity of hybrid bodies, such as the Boards of Appeal, formally placed outside the EU judicial architecture, yet projected into a more judicial dimension by recent reforms of the CJEU Statute (J. Alberti, 2023, p. 72).

While data on the CJEU’s activities are publicly available since many years on, far less information are provided on those of EU Agencies’ BoAs. However, over the last years these bodies have been performing an increasing role in the EU system of judicial protection. According to Aquind (General Court, 18^th of November 2020, T-735/18, Aquind v. ACER, ECLI:EU:T:2020:542), they have to perform a full review over the Agencies’ decisions, so as to balance the limited nature of the review conducted by the EU courts over complex technical, scientific and economic assessments and, thus, to enhance the overall protection offered to individuals. Moreover, according to Article 58a of CJEU Statute (entered into force in 2019 for the BoAs of EUIPO, CPVO, EASA and ECHA and soon to be extended also to those of ACER, ESAs, SRB and ERA, thanks to the further reform of CJEU Statute which is set to enter into force on September 1^st, 2024) they are also entitled to reduce the workload of the CJEU, delivering decisions that could be appealed before the GC but not before the Court of Justice (hereinafter, ‘CJ’) unless they raise ‘an issue that is significant with respect to the unity, consistency or development of Union law’.

Against this framework, this contribution follows up a previous similar work (J. Alberti, 2019, p. 16) and aims at assessing the magnitude of litigation before BoAs by delivering data on their activity. In particular, the number of yearly adjudicated cases by the BoAs will be assessed, noting how many procedures were concluded with the repeal of contested decisions or the dismissal of the appeal. Information on proceedings’ duration will be presented as well, taking into account time limits, whenever set out by establishing regulations. This will allow to compare BoAs’ performance with statistics on proceedings before the CJEU. Indeed, swiftness has been one of the main factors fostering the institution and consolidation of BoAs, hence, it amounts to a parameter of legitimacy of the devolution process they are being part of. This contribution therefore conducts an empirical analysis on BoAs’ responsiveness, which serves as one of the indicators of the quality of their output. For this latter purpose, BoAs’ decisions and CJEU’s rulings are organised according to the date of the final decision.

Furthermore, the research also aims at investigating the impact of BoAs’ activity over the CJEU’s overall workload, assessing the number of appeals brought before the GC and the CJ, as well as the acceptance by the EU Judges, monitoring whether and to what extent BoA’s decisions have been confirmed or not by the CJEU. The issue of the “untapped potential” is also addressed, investigating the volume and nature of litigation falling out of BoAs’ competence and arriving directly before the CJEU. Since the timeframe of the present research stretches from January 1st 2018 to December 31st 2023, the impact on this latter issue given by the entry into force of Article 58a of CJEU Statute will also be taken into account (for a theorical perspective on the filter, see in this Review, R. Torresan, forthcoming). For a detailed account of the application of the filter, the research was conducted on the Curia database.

As to the methodology implemented, the research relies on the Common Database of EU Agencies’ Boards of Appeal for BoAs’ decisions and subsequent judicial developments. The official database of the CJEU (Curia) has also been consulted in order keep track of pending proceedings and recently rendered judgments and orders which might be of interest. Finally, official reports published by Agencies have been used to gather information on their activity.

A quantitative assessment of BoAs’ activity

The reader should be advised that, given the main purpose of the research – and whenever it is not stated otherwise (such as for EUIPO) – sections concerning judgments and orders by the CG and the CJ do not report the total amount of decisions related to BoAs’ activity on that specific year, rather how many cases decided by the BoA on that year have subsequently been under scrutiny by the GC and by the CJ and the following outcome, notwithstanding the date of the final judgement. Decisions stemming from relaunched procedures following an annulment by the CJEU are deemed as the result of autonomous procedure.

 

a. EUIPO’s Boards of Appeal

Data on BoAs’ activity is taken from Official Statistics published by EUIPO. Data on CJEU’s rulings are taken from EUIPO Database eSearch Case Law. As the information provided here is simply a re-elaborated version of official statistics, decisions are organised accordingly.

Year	BoAs decisions			GC decisions		Filter Art. 58a Statute		CJ decisions
	Upheld	Dismissed		Upheld	Dismissed	Allowed	Inadmissible	Upheld	Dismissed
2018	2 600			342		—	—	63
	634		1 855	53	215			5	58
2019	2 514			303		0	28	61
	557		1 596	47	215			8	53
2020	2 558			248		0	34	29
	620		1 559	47	157			11	18
2021	2 687			320		1	46	2
	690		1 644	59	213			0	2
2022	2 490			300		2	39	8
	548		1 619	26	195			0	8
2023	2 621			273		4	45	4
	653		1 806	25	204			1	3

According to data provided by the Agency, most of decisions issued concerned inter partes proceedings under the EUTM Regulation (2, 515 in 2023, 2, 328 in 2022, 2, 598 in 2021, 2, 415 in 2020, 2, 402 in 2019, 2, 497 in 2018).

The CJ has made extensive use of the filter under Article 58a of the Statute since its introduction in 2019. Such a strict implementation not only did prevent further pressure on the Court, but also caused a plunge in the number of cases under scrutiny. It is premature to conduct an assessment on the systemic relevance of admitted cases and on their outcome. In fact, the first decision following a leave for an appeal was rendered in February 2024, annulling the previous General Court’s judgment (Court of Justice, 27^th of February 2024, C-382/21 P, EUIPO c. The KaiKai Company Jaeger Wichmann Gbr, ECLI:EU:C:2024:172); in this Review, M. Coli, p. 2).

 

b. CPVO’s BoA

Year	BoA decisions		Appealed before GC		Filter Art. 58a Statute		Appealed before CJ
	Upheld	Dismissed	Upheld	Dismissed	Allowed	Inadmissible	Upheld	Dismissed
2018	2		1		0	0	0
	0	2	0	1			0	0
2019	2		1		0	1	0
	1	1	0	1			0	0
2020	2		2		0	1	0
	0	2	0	2			0	0
2021	3		0		0	0	0
	2	1	0	0			0	0
2022	3		0		0	0	0
	0	3	0	0			0	0
2023	2		1 pending		0	0	0
	0	2	—	—			0	0

Caseload before the CPVO’s BoA is chronically scarce, except for delimited surges. For example, back in 2016 the Board issued 21 decisions (J. Alberti, 2019, op. cit., p. 18).

So far, the Court of Justice has adopted two orders of inadmissibility pursuant to Article 58a Statute, all related to actions brought against BoA’s decisions prior to the timeframe object of the present study (Court of Justice, order of 16^th of September 2019, C-444/19 P, Kiku v. CPVO, ECLI:EU:C:2019:746; order of March 3^rd 2020, C-886/19 P, Pink Lady America v. CPVO, ECLI:EU:C:2020:146). Therefore, contrary to the methodology implemented in the rest of the research, here such decisions are labelled on the basis of the date of their issuing.

 

c. EASA’s BoA

Year	BoA decisions		Appealed before GC		Filter Art. 58a Statute		Appealed before CJ
	Upheld	Dismissed	Upheld	Dismissed	Allowed	Inadmissible	Upheld	Dismissed
2018	1		1		0	0	0
	0	1	0	1			0	0
2019	0		0		0	0	0
	0	0	0	0			0	0
2020	0		0		0	0	0
	0	0	0	0			0	0
2021	0		0		0	0	0
	0	0	0	0			0	0
2022	1		0		0	0	0
	0	1	0	0			0	0
2023	1		0		0	0	0
	0	1	0	0			0	0

The only action brought before the General Court did not lead to any decision on the merit, as the appeal was discontinued (order of the President of the General Court of 11 September 2018, case T-371/18, Reiner Stemme Utility Air Systems GmbH v. EASA).

 

d. ERA’s BoA

Year	BoA decisions		Appealed before GC		Appealed before CJ
	Upheld	Dismissed	Upheld	Dismissed	Upheld	Dismissed
2018	0		0		0
	0	0	0	0	0	0
2019	0		0		0
	0	0	0	0	0	0
2020	0		0		0
	0	0	0	0	0	0
2021	2		0		0
	2	0	0	0	0	0
2022	0		0		0
	0	0	0	0	0	0
2023	0		0		0
	0	0	0	0	0	0

 

e. ECHA’s BoA

Year	BoA decisions		Appealed before GC		Filter Art. 58a St.		Appealed before CJ
	Upheld	Dismissed	Upheld	Dismissed	Allowed	Inadmissibile	Upheld	Dismissed
2018	9		0		0	0	0
	6	3	0	0			0	0
2019	11		1		0	0	0
	6	5	0	1			0	0
2020	16		2		0	1	0
	7	9	0	2			0	0
2021	12		2 (1 pending)		0	0	0
	4	8	0	1			0	0
2022	9		0		0	0	0
	2	7	0	0			0	0
2023	11		1 pending		0	0	0
	5	6	—	—			0	0

The only CJ’s order stating on the admissibility under Article 58a Statute relates to a BoA’s decision first issued back in 2020. However, the order refusing the appeal to proceed was rendered just recently (see Court of Justice, order of 28th May 2024, C-79/24 P, Cruelty Free Europe v. ECHA, ECLI:EU:C:2024:430).

ECHA Agency has made use of the possibility to rectify or withdraw in its entirety the initial decision, as provided for by Article 93 (1) of the REACH Regulation. In some cases the appeals were dismissed due to supervened amendment to the original decision by the Agency by integration of a allegedly lacking motivation (See cases A-005-2022, decision of 22 august 2022; case A-007-2022, decision of 6 october 2022). However, this does not entail a new approach by the Agency, as other examples can be found in previous case law (for further references, see M. Chamon – A. Volpato – M. Eliantonio, p. 90). Even more interesting, in one case the dismissal was justified upon the fact that the contested decision had been revoked by the Agency itself due to “deficencies in the Agency’s assessment of the relevant available information” (see case A-002-2023, decisions of 25^th of April 2023).

 

f. SRB’s Appeal Panel (AP) 

Year	BoA decisions		Appealed before GC		Appealed before CJ
	Upheld	Dismissed	Upheld	Dismissed	Upheld	Dismissed
2018	17		1		0
	12	5	0	1	0	0
2019	15		0		0
	4	11	0	0	0	0
2020	8		0		0
	2	6	0	0	0	0
2021	1		0		0
	0	1	0	0	0	0
2022	3		0		0
	2	1	0	0	0	0
2023	9		0		0
	5	4	0	0	0	0

 The Table depicts a not homogenous trend, marked by a quite intense activity from 2017 (when 54 decisions were adopted, see J. Alberti, 2019, op. cit.) to 2019, followed by an overall lightening in the following years.

The occasional surges might be correlated with higher workload that hit the Appeal Panel after important resolution decisions adopted by the SRB. In fact, while it is true that the adoption of resolution schemes is not per se an act subject to challenge before the AP, nonetheless it prompts the filing of requests of access to documents by concerned investors. This has repercussions on the duration of proceedings as well (see infra).

In spite of a marginal gain in 2023 (9 cases concluded), one might infer that the workload before the AP has been decreasing over the last years. This is further confirmed by the latest activity of the AP: in 2024 four decisions have been issued so far. This might suggest that the latest resolution decision will not pave the way to another surge in the AP’s activity.

 

g. ESAs’ Joint Board of Appeal (JBoA)

Year	BoA decisions		Appealed before GC		Appealed before CJ
	Upheld	Dismissed	Upheld	Dismissed	Upheld	Dismissed
2018	0		0		0
	0	0	0	0	0	0
2019	2		0		0
	1	1	0	0	0	0
2020	3		0		0
	0	3	0	0	0	0
2021	3		0		0
	0	3	0	0	0	0
2022	1		0		0
	0	1	0	0	0	0
2023	1		1 pending		0
	0	1	—	—	0	0

 

h. ACER’s BoA

Year	BoA decisions		Appealed before GC		Appealed before CJ
	Upheld	Dismissed	Upheld	Dismissed	Upheld	Dismissed
2018	1		1		1
	0	1	1	0	0	1
2019	4		1		0
	1	3	1	0	0	0
2020	7		3		2 pending
	2	5	1	2	—	—
2021	7		8 (7 pending)		0
	1	6	0	1	0	0
2022	4		2 pending		0
	1	3	—	—	0	0
2023	6		5 pending		0
	0	6	—	—	0	0

All seven pending appeals introduced in 2021 contested the same decisions (A-001-2021). From a substantive point of view, it would not appear that the Aquind judgement has led to an increased rate of annulments, but rather to a stricter duty to state motivation and, consequently, a deeper scrutiny from the Board.

 

Review of CJEU rulings on Agencies’ decisions falling outside BoAs’ competence

The “untapped potential” of Agencies’ BoAs has been already widely discussed by scholars and it refers to the possibility to extend the BoAs’ powers to all the acts adopted by the Agency to which it belongs (and not only a selection thereof), to actions others than that of annulment, to engage into a more innovative approach on the capability of technical acts to produce legal effects, so as to expand the reviewability of EU Agencies’ soft law (C. Tovo, p. 353 ; J. Alberti, 2018, p. 213; on the ESAs’ JBoA and the SRB’s AP, M. Chamon, A. Volpato, M. Eliantonio, p. 30; on the reviewability of ESMA’s soft law under Article 263 TFEU by BoAs and the CJEU, M. van Rijsbergen, p. 256).

Indeed, by broadening the BoAs’ mandate not only the CJEU would be further discharged of its workload, but a thorough review of more Agencies’ acts would be ensured as well, especially whenever their adoption entails complex technical or scientific assessments. In such domains, the CJEU has always refrained from plunging into a comprehensive review, limiting itself to the ascertainment of a “manifest error” or a “misuse” of the discretion conferred to the body. Conversely, technical expertise provided by Boards may constitute a remedy to such blind spots, improving Agencies’ accountability (on the financial regulatory framework, J. Pollak, P. Slominski, p. 140; on the gaps within ECHA’s BoA competence, A. Bartosch, p. 441).

As the theorical aspects have already been discussed, this research intends to delve into the quantitative aspect, focusing in particular on how many acts adopted by EU Agencies are challenged directly before EU Courts, bypassing BoAs. As Boards have always been designed to deal with litigation triggered by non-privileged applicants, actions before the CJEU brought by Member States are not taken into account, as they do not amount to an example of “untapped potential”. Still, the numbers are negligible as currently two cases are pending against ACER, both brought by Germany (T-283/19, Germany c. ACER, appeal of May 2^nd 2019; T-612/23, Germany c. ACER, appeal of September 28^th 2023) and only two rulings involving ECHA have been delivered so far (General Court, 20^th of September 2019, T-755/17, Germany c. ECHA, ECLI:EU:T:2019:647; General Court, 15^th of September 2021, T-127/20, France c. ECHA, ECLI:EU:T:2021:572). Case T-540/20, France c. SRB, was recently closed on July 20^th 2024, hence, being outside the scope of research.

It bears noting that some domains have been excluded from the scope of research. First of all, as this contribution intends to focus on direct actions, preliminary rulings questioning Agencies’ acts validity have not been included. While we acknowledge the existence of important precedents, such as the FBF case, for the sake of homogeneity – and given a more complex traceability of preliminary proceedings – this contribution only addresses contentious proceedings. The litigation on civil service matters is purposely excluded from the scope of research as well, since this domain is typically dealt with by the CJEU as an autonomous policy field. Judgments are classified according to date of their issuing.

Rulings issued by the GC (and the CJ) on Agencies’ decisions falling outside BoAs’ competence

Year	EUIPO	CPVO	EASA	ERA	ECHA	SRB	ESAS	ACER
2018	0	0	0	0	5	2	0	0
2019	0	0	0	0	5 (1)	3	0	0
2020	0	0	0	0	3	3 (1)	0	0
2021	0	0	0	0	3 (1)	1 (6)	1	0
2022	0	0	0	0	2	7 (4)	0	0
2023	0	0	0	0	1 (3)	19 (1)	0	0
Pending cases	0	0	0	0	4 (1)	250	0	0

The following Table offers a more detailed overview, pointing out the outcome of the rulings rendered by the GC (and the CJ). 

Year	ECHA		SRB			ESAs
	Upheld	Dismissed	Upheld	Dismissed		Upheld	Dismissed
2018	5		2			0
	1	4	0	2		0
2019	5 (1)		3			0
	0	5 (1)	3	0		0
2020	3		3 (1)			0
	0	3	3		0 (1)	0
2021	3 (1)		1 (6)			1
	0	3 (1)	(2)	1 (4)		0	1
2022	2		7 (4)			0
	0	2	(2)	7 (2)		0
2023	1 (3)		19 (1)			0
	0	1 (3)	7	12 (1)

The Tables above show that a significant part of litigation arising from ECHA’s and SRB’s decisions eludes the remittal of the respective BoAs.

As to rulings issued in respect of ECHA, they mostly pertain to decisions adopted pursuant to Articles 57 and 59 and including certain substances in Annex XIV of Regulation (EC) No 1907/2006 (13 out of 24). The second most relevant topic is the accordance of tariffs reduction for SME (5 out of total).

According to the Curia database, 4 cases are currently pending before the CJEU, 3 concerning substances evaluation and 1 on access to documents.

The Tables also deliver empirical validation of doubts raised in respect of the narrow jurisdiction of the SRB Appeal Panel (D. Ramos-Muñoz, M. Lamandini, p. 119), providing a quantitative perspective on the topic. In fact, cases mostly concern decisions on banks resolutions and ex ante contributions. In its latest Official Report (SRB 2023 Annual Report, p. 51) the Agency mentions that out of 250 pending cases before the CJEU on the 31^st December 2023 almost a half (121) concerned SRB decisions taken in the context of the exercise of the resolution powers in relation to Banco Popular Espanol, ABLV, PNB Banka and Sberbank. 126 actions concerned decisions in the matter of ex ante contributions. The huge disproportion between pending affaires and cases closed per year points out the bottleneck caused to the CJEU by the activity of the SRB. This has caused actions first introduced in 2017 to be adjudicated after more than 5 years.

The only decision pertaining to the ESAs is General Court, order of October 10^th 2021, in case T-760/20, Stasys Jakeliunas v. ESMA, adjudicating on the Agency’s refusal to start an inquiry on potential violation of Union law.

 

Average duration of proceedings before BoAs

As to the length of the proceedings before BoAs, it bears noting that several Agencies’ establishing Regulations set forth a specific time frame within which the BoA has to take the final decision. As further detailed below, however, this term is often related to the sheer deliberative part of the proceedings, i.e. to the time in which the BoA has to draft the final decision after the conclusion of the written and oral part.

Therefore, the question on the overall length of the proceedings before BoAs remains open and it has been here addressed through a case by case analysis that calculates the overall duration from the notice of appeal to the date of final decision.  Relaunched procedures following an annulment by the CJEU are deemed as autonomous proceedings.  Moreover, whenever the relevant regulation or established practice sets a dies a quo other than the lodging of the appeal, this is also taken into account and presented in a different Table. As for EUIPO, given the lack of official statistics and the huge amount of yearly workload managed by the latter, following the methodology used by the Common Database on EU Agencies’ Boards of Appeal this research has covered only the decisions taken by the Grand Board (hereinafter, “GB”).

 

a. Overall length of proceedings before BoAs (from the lodgement of the appeal to the final decision)

Year   Time limits	EUIPO (Gran Board)     none	CPVO    3 months Art. 52 (1) Reg. 874/2009	EASA        none	ERA   3 months Art. 62 (1) Reg. 2016/796	ECHA       none	AP    1 month Art. 85 (4) Reg. 806/2014	JBoA    2 months Art. 60 (2) ESAs Reg.	ACER   4 months Art. 28 (2) Reg. 2019/942
2018	22,6 months	12,3 months	9,7 months	—	13,4 months	6,3 months	—	2 months
2019	26,1	12,6	—	—	11,1	5	3,8	2,3
2020	37,1	30,7	—	—	21,8	2,5	2,3	4,2
2021	40,1	11,8	—	3,4	11,8	5,2	2,6	3,9
2022	42,1	11,9	9	—	12,4	7,8	3,2	6,9
2023	45,1	30,8	7	—	12,5	3,6	2,1	6,8

No relevant distinction could be observed between proceedings concluded with the dismissal or the upholding of the appeal.

Particularly interesting are the relevant differences among BoAs. Duration ranges from a mere couple of months (see, for instance JBoA and ERA) up to almost 4 years (see latest EUIPO GB trends). This might be the result of both technical complexity of cases under scrutiny and workload which the Board is charged with. Indeed, a more far reaching competence of the Board might be correlated with longer proceedings (see EUIPO, CPVO, ECHA, as opposed to JBoA and AP).

Another aspect worth discussing is a seemingly inconsistent trend over the years. In fact, some columns present either a progressive surge or consecutive rises and dips. EUIPO’s trend could be easily explained by the structural high demand of registrations under the EUTMR (as well as, of course, the specificity of the chamber under scrutiny in this research, namely the Grand Board). In other cases, such as ECHA’s, this is due to both organisational and exogenous factors. Indeed, the 2023 Annual Report from the Chairman of the Board of Appeal provides clarifications on the 2020 “anomaly”: complexity of the cases, in itinere changes to the composition of the Board and restrictions resulting from Covid caused delays.

On the other hand, the progressive rise pertaining to procedures before ACER’s BoA might hint at a correlation with the Aquind judgement (General Court, 18^th of November 2020, case T-735/18, Aquind v. ACER, cit., confirmed by the Court of Justice, 9^th of March 2023, C-46/21 P, ECLI:EU:C:2023:182). Such ruling – and the duty of full review which is prescribed therein – might have led to a more cautious approach by the BoA, resulting in a stretched duration. As a matter of fact, proceedings introduced in 2021 and 2022 had a duration which is in stark contrast with that of the previous years, in spite of a comparable workload (see above, at para 1). This trend might suggest that the Board aligned with the GC’s indications on the scope of review well before the final word on the Aquind case given by the CJ on the appeal case decided in 2023. Such an interpretation could explain the climb from 2, 3 months in 2019 to 6, 8 months in 2023. This correlation may be better visualized arranging gathered data according to the date of the filling of the appeal, as presented in the table below.

Year	Avg. duration before ACER BoA (months)
2018	2
2019	3, 6
2020	3, 82
2021	6, 8
2022	7, 7
2023	4, 2

As to 2023, the proceedings (A-004-2019_R, A-003-2019_R) were in the wake of previous BoA’s decisions and CJEU rulings. The fact that the investigation had already been carried out and that the GC provided indications on point of law might account for shorter time required to issue the final decision (3 months the former, 5 months and 13 days the latter). This makes the data for 2023 not as much reliable and representative.

Conversely, the statistics related to the ECHA BoA – which also has to apply high standards of review thanks to the BASF judgement (General Court, 20th of September 2019, T-125/17, BASF c. Germany, ECLI:EU:T:2019:638), which has been the blueprint for Aquind – do not show similar extensions in the duration of proceedings. Even though the reasons may well be founded simply in the peculiarities of the cases and in the differences among the two policy fields, it seems relevant to point out that ECHA’s BoA has a permanent nature (i.e., the majority of members is employed by the Agency on a full-time basis). This certainly helps this body in better handling its workload.

As regards the SRB, the procedures before its Appeal Panel show the same inconsistencies as to duration. The trend is marked by occasional increases, such as in 2018 (6, 3 months), 2021 (5, 2 months) and 2022 (7, 8 months). The causes may be further assessed by organizing cases on the basis of the date of their introduction and carrying out a comparison with the adoption of the most relevant resolution decisions by the Single Resolution Board.

Year	Duration from the lodging (months)
2018	6, 3
2019	2, 5
2020	2, 64
2021	5
2022	7, 8
2023	3, 33

The Table enables the reader to locate more accurately the origin of backlogs. In fact, the most time expensive cases were introduced back in 2018 and 2022. Resolution decisions of Banco Popular and Sberbank date back respectively to 2017 and 2022. This, together with the fact that in 2017, 2018, 2019 and 2023 the SRB has adopted more decisions compared to other years shows that sudden workload increase indirectly linked to resolution decisions negatively impacted on swiftness granted by the AP.

Finally, the limited workload of the EASA and ERA’s BoAs makes it difficult to describe a trend.

 

b. Do BoAs respect the maximum length of the proceedings set forth in their establishing Regulations (if any)?

As anticipated above, some regulations provide that the proceedings before BoAs shall not exceed specific time-limits and, to this end, establish a dies a quo other than the lodging of the appeal. In the case of EASA, despite lacking a clear legal basis, by established practice the appeal is deemed to have been lodged after the conclusion of an interlocutory phase in occasion of which the Agency may revoke, modify or hold its decision.

The following Table calculates the length of the proceedings starting from that other circumstance, assessing whether BoAs respect the time limits set forth in their establishing Regulation.

Year   Dies a quo	EUIPO   Lodging of the appeal	CPVO   Closure of oral proceeding (art. 52 (1))	EASA   End of interlocutory review	ERA   Lodging of the appeal	ECHA   Lodging of the appeal	AP   Completion of evidence (art. 20 RoP)	JBoA   Lodging of the appeal	ACER   Lodging of the appeal
2018	As above	0,5 months	6,7 months	As above	As above	20 days	As above	As above
2019		2,5	—			15 days
2020		2,1	—			10, 2   days
2021		0,7	—			14, 3 days
2022		2	7			20 days
2023		2	6,3			2, 5 days

The imposition of a dies a quo other than the lodging of the appeal – whether be it by provision (CPVO and SRB) or by established practice (EASA) – significantly curbs down the duration. This is especially true for CPVO’s BoA and the SRB AP, whose time limit refers to the sheer deliberative part. More in detail, before the AP the deliberative part only accounts for a marginal fraction of the total duration of the proceeding. This may suggest that the investigation phase is the most time expensive.

 

Duration of proceedings before the CJEU

Another aspect worth assessing is the impact in terms of swiftness of a prior review carried by a BoA on the following appeals before the CJEU. The Tables below show the differences in length between appeals contesting a decision issued by a Board of Appeal and other actions brought directly before the EU Judiciary (last column on the right). It bears noting that data on direct actions before the CJEU are extracted from the 2022 and 2023 Annual Reports drafted by the Court of Justice.

 

a. Length of appeals before the GC against BoAs’ decisions (in months)

Year	EUIPO GB	CPVO	EASA	ERA	ECHA	SRB	ESAs	ACER	Direct actions before GC
2018	16,6	—	1,7	No further litigation	—	See below	No further litigation	—	20
2019	—	—	—		31,1			27,7	16,9
2020	—	14,6	—		21,1			20,9	15,4
2021	—	15,6	—		—			—	17,3
2022	—	—	—		—			34,9	16,2
2023	10,1	—	—		32,1			28,4	18,2

As to SRB’s AP, only a single order from the CJEU could be reviewed. Case T-514/18, Antonio della Valle Ruiz et al. v. SRB, was first introduced by appeal in August 2018. The action was ultimately discontinued by the claimants, leading to removal from the register by order of January 18th 2024. No decision on substantive matters could be tracked on the Common Database.

 

b. Length of further actions before the CJ against a GC act delivered on the basis of a BoA decision (in months)

Year	EUIPO GB	CPVO	EASA	ERA	ECHA	SRB	ESAs	ACER	Appeals before the CJ
2018	—	No further litigation	1,6	No further litigation	No further litigation	No further litigation	No further litigation	—	13,4
2019	18,5		—					—	11,1
2020	—		—					—	13,8
2021	—		—					—	15,1
2022	—		—					—	11,9
2023	—		—					24,5	13,9

From these data, Boards of Appeal stand out for delivering swift decisions. In spite of a general trend towards the extension of proceedings’ duration, BoAs’ responsiveness is still far greater than that of EU Courts. Nonetheless, whenever a decision issued by a BoA is further challenged before the GC, the time frame significantly stretches, as far as 34,9 months. Moreover, this trend seems to concern mainly those Agencies (and respective BoAs) involved in highly technical scientific and economic assessments, namely ECHA and ACER.

On the one hand, this is not surprising, given the fact that BoAs have been established exactly for reaching these results (sometimes setting forth specific time-limits, as discussed above) and given their very limited workload. On the other hand, the increased resources brought by the doubling of the General Court as well as the fact that both EU Courts can rely upon the preliminary assessment of facts and findings made by the BoAs raise some doubts on the former’s efficiency, at least in the highly technical policy fields in which BoAs operate.

Moreover, the added value given by these latter bodies is demonstrated also by the further increase in the average length of the proceedings pending before EU Courts in absence of a prior review by a BoA (which happens in the cases, discussed above, where the power to review the legitimacy of an act of those agencies has not been given to BoAs, falling directly and solely within the jurisdiction of the EU Judiciary).

 

c. Length of proceedings before the GC (and the CJ) in absence of prior review by a BoA (in months)

Year	EUIPO	CPVO	EASA	ERA	ECHA	SRB	ESAs	ACER
2018	No litigation	No litigation	No litigation	No litigation	19,9	21,2	—	No litigation
2019					21 (18,7)	41,3	—
2020					15,5	39,2 (13,4)	—
2021					24,4 (25,1)	25,3 (17,7)	9, 8
2022					19,9	53,1 (17)	—
2023					13,2 (23,5)	36,1 (13)	—

As already discussed above, the most relevant – and time expensive – litigation before the CJEU stems from SRB’s decisions, which for the most part elude the AP’s jurisdiction. The results of the research show that serial litigation prompted by banks resolution may take many years to be addressed by the GC in the first place. Moreover, as the majority of pending cases still concern the Banco Popular resolution, it is likely that future statistics might reach even higher values. While the current absence of comparable case law before the AP prevents any attempt to forecast how much time would take the appellate body to adjudicate on matters of banks resolutions, one can not neglect that the AP still enjoys greater technical specialization as well as human resources capable of tackling such task. Moreover, contrary to the constant flow of appeals brought before the CJEU, the AP’s activity is densely concentrated in defined timeframes, allowing for the adoption of the necessary organisational measures. In any case, as entrusting the AP with a far reaching competence would require substantial amendments to the founding Regulation, the body’s structure could be re-devised accordingly, for example requiring a full-time commitment from its members.  These considerations may persuade that endowing the AP with the power to review all SRB’s decisions would indeed positively affect responsiveness.

The second most relevant source of contentieux are decisions adopted by ECHA. Actions brought against the introduction of a substance in the List under Annex XIV (Article 57 and 59 of the REACH Regulation) led to the longest proceedings, lasting up to almost 3 years.

 

Conclusive remarks 

The quantitative analysis of BoAs workload in the years 2018-2023 has confirmed the great diversity which characterizes the functional and operational aspect of such bodies. Workload is massively concentrated in field of intellectual property, whereas in other sectors, such as aviation and railway, litigation is almost non-existent. In addition to that, the analysis of the cases that are brought directly before EU Courts, circumventing the review by BoAs, has demonstrated that these latter bodies do have a certain untapped potential, particularly in the fields of chemicals, banking resolution and finance. Expanding the acts reviewable by BoAs in these policy fields would contribute in easing the pressure on the CJEU, while enhancing the overall review offered to individuals, ensuring an extra level of specialised scrutiny.

These remarks further shore up the objections which have been raised in regard to the extension of the filter under Article 58a Statute of the Court of Justice to ACER, SRB, the ESAs and ERA. If BoAs’ powers and competences are not expanded, such reform seems to be set to produce marginal results, at least from a purely quantitative perspective.

Overall, BoAs have proven to be a valuable resource when it comes to address highly technical assessments and regulatory measures adopted by Agencies. This is demonstrated by the low rate of appeals and subsequent dismissals by the GC.

Research on proceedings’ length shows that most of the BoAs abide by the time limits set forth by their establishing Regulations, outperforming the General Court. This holds true even in the light of the discussed impact of Aquind. This suggests that BoAs can amount to a fast side-track to traditional means of judicial protection. However, length considerably stretches whenever BoAs’ decisions are further challenged before the GC. Nonetheless, as it was already noted above, most of the litigation is closed before the Boards without the need for the CJEU to state on the matter, thus offering to individuals a swift and valuable response to their needs.