To What Extent Can the CJEU Contribute to Increasing the EU Legislative Process' Transparency?

Alongside other actors such as the European Ombudsman, the Court of Justice of the European Union (CJEU) plays what looks like, at first sight, a key role in improving the transparency of EU legislative procedures. To take two relatively recent examples, theDe Capitani v. European Parliament (2018) judgmentwas perceived as a victory by those in favor of increased transparency of EU legislative procedures at the stage of trilogues, as was the ClientEarth v. European Commission (2018) judgment regarding the pre-initiative stage. Both rulings emphasize the need for “allowing citizens to scrutinize all the information which has formed the basis of a legislative act...[as] a precondition for the effective exercise of their democratic rights” (ClientEarth v. European Commission, 2018, §84; De Capitani v. European Parliament, 2018, §80). Nevertheless, while the CJEU’s case law may indeed contribute to improving the legislative process’ transparency, its impact on the latter is inherently limited and even bears the potential of having a perverse effect. This article sheds light on the limits of the CJEU’s capacity to act in this field and the potential effects of its case law on the EU institutions’ attitudes or internal organization.


Introduction
Many individuals or NGOs active in the field of transparency are confronted with confirmatory decisions of EU institutions rejecting their requests for access to documents. These individuals and NGOs then place a lot of hope in actions for annulment of those decisions brought before the Court of Justice of the European Union (hereafter CJEU), be it in first instance the General Court or in appeal the Court of Justice (Driessen, 2012, pp. 254-260). Alongside other actors such as the European Ombudsman, the CJEU serves what looks like, at first sight, a key role in improving the EU legislative procedures' transparency.
Under EU law, transparency relates to the broader concept of openness. Whereas the Treaties do not provide us with a definition of what the principle of openness entails, the academic literature offers a wide range of diverse understandings, often using, similarly to the CJEU itself, the concepts of openness and transparency interchangeably (Curtin & Mendes, 2011, p. 103). In our opinion, openness of decision-making should be understood as entailing two aspects, namely transparencydefined restrictively as the possibility for any individual to access information (de Fine Licht & Naurin, 2016, p. 217;Wyatt, 2018)-and participation-defined as the actual possibility to participate in the decision-making process (Alemanno, 2014). These two elements are interlinked in the sense that meaningful participation necessitates fully-fledged transparency. In other words, the former is dependent on the latter. This is particularly salient in the context of the EU legislative process, in which the purpose of transparency of ongoing procedures is about public scrutiny-or accountability to the public-but also allowing, in a timely manner, the participation of any interested citizen while a legislative act is in the making. It is this understanding of openness that the CJEU gives when stating that "allowing citizens to scrutinize all the information which has formed the basis of a legislative act…is a precondition for the effective exercise of their democratic rights" (e.g., ClientEarth v. European Commission, 2018, §84;De Capitani v. European Parliament, 2018, §80). Even though it offers us such a bold statement, the CJEU's ability to satisfy the expectations of pro-transparency activists is limited for various reasons. Moreover, while some rulings appear at first glance to foster the transparency of the EU legislative process, they can paradoxically lead to the contrary.
This article aims to shed light on these limitations and pitfalls by focusing on the limits of the impact of the judiciary on the legislative process' transparency. First, it gives a short overview of the key provisions of EU law relevant for the debate on the transparency of the legislative process. Second, it highlights three inherent limits-the limits of interpretation, the principles of institutional balance and institutional autonomy, and timeof the CJEU's action on the matter. It does so by drawing on some lessons from the case law, taking the judgments De Capitani v. European Parliament (2018) of the General Court and ClientEarth v. European Commission (2018) of the Court of Justice as cases of reference for this discussion, given they both draw on previous case law on the transparency of the legislative process (such as the cases Council v. Access Info Europe, 2013;Herbert Smith Freehills v. Council, 2016;Sison v. Council, 2007;Sweden and Others v. API and Commission, 2010;and Sweden and Turco v. Council, 2008), but also build on the latter on crucial stages of the legislative procedures. Third, this article touches upon the issue of the risk of perverse effect of the CJEU's case law in this field. It concludes by saying that if the CJEU indeed has the capacity to improve the openness of the legislative process, this capacity is more limited than we might think and bears a potential perverse effect. On a final note, it raises some suggestions on how to rebalance the dynamics in place between the judiciary and the legislative branch to increase the capacity of the CJEU to act in this field.

Legal Background
According to the Treaties, the functioning of the EU relies on two complementary models of democracy (Curtin & Leino-Sandberg, 2016, p. 4). First, it is founded on the model of representative democracy (Consolidated version of the Treaty on European Union [TEU], 2016, Article 10(1-2)), according to which citizens elect representatives who in turn should be held accountable to the citizens for the decisions they take. Second, it is equally founded on the model of participatory democracy as the Treaties expressly foresee that "[e]very citizen shall have the right to participate in the democratic life of the Union." Thus, "[i]n order to promote good governance and ensure the participation of civil society," EU institutions shall conduct their work and take their decisions "as openly as possible to the citizen" (Consolidated version of the Treaty on the Functioning of the European Union [TFEU], 2016, Article 15(1); TEU, 2016, Article 10(3)). Concretely, EU "institutions shall, by appropriate means, give citizens and representative associations the opportunity to make known and publicly exchange their views in all areas of Union action" (Marxsen, 2015;Mendes, 2011;TEU, 2016, Article 11(1)). In addition, the Treaties add that each institution "shall ensure that its proceedings are transparent" (TFEU, 2016, Article 15(3)), and both the Treaties and the Charter of Fundamental Rights of the EU recognize the right of any EU citizen, and any natural or legal person residing or having its registered office in a Member State, to access EU institutions' documents, subject to limitations on grounds of public or private interest as fixed under EU secondary law (Charter of Fundamental Rights of the European Union, 2016, Articles 42 and 52; TFEU, 2016, Article 15 (3)). It is through these various provisions that the principle of openness, as defined in the introduction, arises.
The central piece of secondary legislation on transparency is Regulation (EC) European Parliament, Council and Commission Documents (2001;hereafter Regulation No 1049/2001. This Regulation fixes the conditions governing the right of access to documents and its limits. It poses that the "widest possible access to documents" should be the norm while the denial of access, on the grounds foreseen under Article 4, be the exception (Article 1; Recitals 4 and 11). It also emphasizes that "[o]penness enables citizens to participate more closely in the decision-making process and guarantees that the administration enjoys greater legitimacy and is more effective and more accountable to the citizen in a democratic system" and "contributes to strengthening the principles of democracy" (Recital 2). Despite some attempts to revamp the already 20 year old Regulation, interinstitutional negotiations have not led to any compromise as "Member States are divided in the Council into those thinking reform ought to mean going forward towards increased openness, and those wishing to turn the clock back" (Leino, 2011(Leino, , p. 1216. Legislative procedures (TFEU, 2016, Articles 289 and 294) benefit from increased transparency requirements. In particular, the Treaties emphasize that "[t]he European Parliament shall meet in public," as shall the Council when considering, deliberating and voting on a draft legislative act (TFEU, 2016, Article 15(2); TEU, 2016, Article 16(8)). Hence the Treaties seem to establish a close relationship between legislative procedures and transparency (Council v. Access Info Europe, Opinion of Advocate General Cruz Villalón, 2013, § §39-40). However, the Treaties underline that access to documents relating to legislative procedures also suffer from limitations fixed in the above-mentioned Regulation (TFEU, 2016, Article 15(3)). As emphasized in the Regulation, "[w]ider access should be granted to documents in cases where the institutions are acting in their legislative capacity" and "[s]uch documents should be made directly accessible to the greatest possible extent," yet "the effectiveness of the institutions' decision-making process" should be preserved (Articles 2(4) and 12(2), Recital 6). In any case, according to a well-known legal principle, exceptions to the right of access must be interpreted and applied strictly, even more so, as induced from the considerations above and a well-established case law, in a legislative context, in which "the discretion left to the institutions not to disclose documents that are part of the normal legislative process is extremely limited or non-existent" (see e.g., Council v. Access Info Europe, 2013, §63; De Capitani v. European Parliament, 2018, §40).

Three Limits to the CJEU's Contribution
The Treaties give one single yet important mission to the CJEU, namely to "ensure that in the interpretation and application of the Treaties the law is observed" (TEU, 2016, Article 19). This role as central judicial actor of the EU legal order comprises, as in any democratic system, the duty to provide checks and balances against abuse of other branches of power, which constitute an important "judicial influence in the political process" (Alter, 1998, p. 124). Like any other actor though, "courts have political limits," or what Alter (1998, p. 138) describes as "some area of 'acceptable latitude,' beyond which they cannot stray." This 'acceptable latitude' of action in the field of transparency appears limited by three factors: the limits of interpretation, the principles of institutional balance and of institutional autonomy, and the importance of time in this area.

The Limits of Interpretation
The framing of the above-mentioned role of the CJEU means that its margin of maneuver is inevitably limited by the provisions of EU law at stake and, consequently, the limits to their interpretation. No lawyer would find anything surprising here. Even if, through interpretation, the boundaries of the understanding of provisions may be dynamic, evolve over time, sometimes dramatically, they remain boundaries. It must be borne in mind that the CJEU, built on the model of the French Conseil d'État, follows a civil law tradition despite featuring some minor common law procedural characteristics. In interpreting EU law, no matter which method of interpretation it uses (literal/grammatical, historical, systematical, purposive or 'effet utile'; Grimmel, 2012, p. 532), the Court must strike a delicate balance between "[making] sense of the political compromises embodied in the relevant legislation" and paying what Arnull (2009Arnull ( , p. 1238 calls "[j]udicial respect" for the "essential elements of those compromises." It is crucial for the CJEU to keep, at least, "the appearance of judicial neutrality, which is the basis for parties accepting the legitimacy of [its] decisions" (Alter, 1998, p. 135). It involves paying due respect to the nature and contexts of the case at stake. Any misstep always bears the risk of giving more ground to authors describing the CJEU as "an uncontrolled authority generating law" (Alter, 1998, p. 129).
Beyond the inherent limits of interpretation, the CJEU's involvement in the transparency debate is also limited in the sense that it can only interpret what it is given a chance to interpret. To give one example, so as not to appear as ruling ultra petita, the General Court explicitly underlined in the De Capitani case that its ruling did not in any way conclude that "direct access to ongoing trilogue work within the meaning of Article 12 of Regulation No 1049/2001" should be ensured, as the case concerned solely the access to the fourth column of documents "on specific request," lodged pursuant to the same Regulation (see below-De Capitani v. European Parliament, 2018, §86). The General Court was not offered the possibility to rule beyond this specific context. The mere fact that the General Court underlines what it does not say shows its acknowledgment of its limits, and that it feels the necessity to emphasize it. In other words, the CJEU needs to receive cases to judge, and more specifically here, cases that mobilize provisions whose interpretation could constitute an opportunity for improving the transparency of the legislative process. Yet the possibilities for cases to reach the CJEU are very limited due to procedural constraints. Cases on transparency brought by individuals or NGOs are inevitably not numerous as any individual willing to launch an action for annulment of a decision refusing the access to documents must demonstrate an interest in bringing proceedings, comply with the other restrictive conditions set in Article 263 TFEU (2016), act within a limited period and have the legal and financial resources to do so.
As a result, if some sort of judicial activism can take place in this field, its ambition is inherently limited.
When the CJEU states that "allowing citizens to scrutinize all the information which has formed the basis of a legislative act…is a precondition for the effective exercise of their democratic rights" (e.g., ClientEarth v. European Commission, 2018, §84;De Capitani v. European Parliament, 2018, §80), it can comfortably rely on the primary law and secondary law mentioned above. Yet if such a general consideration by the CJEU is welcome, it does not per se lead to any subjective right. The CJEU gets more adventurous when, in the ClientEarth case, it asserts that "allowing divergences between various points of view to be openly debated…also contributes to increasing…citizens' confidence in those institutions" (ClientEarth v. European Commission, 2018, §75) and that: [B]y increasing the legitimacy of the Commission's decision-making process, transparency ensures the credibility of that institution's action in the minds of citizens and concerned organizations and thus specifically contributes to ensuring that that institution acts in a fully independent manner and exclusively in the general interest. (ClientEarth v. European Commission, 2018, §104) The ClientEarth judgment was rendered on appeal of a judgment of the General Court. This ruling concerned the stage preceding the submission by the Commission of draft legislative proposal to the co-legislators. The NGO ClientEarth had sent the Commission two requests for access to specific documents, namely one for the draft impact assessment report regarding access to justice in environmental matters and one for the impact assessment report regarding inspections and surveillance in environmental matters, together with the respective opinions of the Impact Assessment Board. Both requests had been rejected by the Commission. In first instance, the General Court dismissed the actions introduced by ClientEarth for annulment of the decisions of the Commission. ClientEarth, unsurprisingly supported in this endeavor by Finland and Sweden, appealed the judgment of the General Court before the Court of Justice.
To motivate its refusal, the Commission relied on the first subparagraph of Article 4(3) of Regulation No 1049/2001. It argued that "impact assessments were intended to help it in preparing its legislative proposals and that the content of those assessments were used to support the policy choices made in such proposals." Therefore, the disclosure, at this "very early and delicate stage," "would seriously undermine its ongoing decisionmaking processes" in restricting "its room for maneuver, [reducing] its ability to reach a compromise" (ClientEarth v. European Commission, § §15, 17).
To the greatest joy of ClientEarth, the Court of Justice annulled the judgment of the General Court. The Court acknowledged that as "the impact assessment procedure takes place upstream of the legislative procedure sensu stricto," the Commission "does [indeed] not itself act in a legislative capacity" at that stage. However, "policy choices made [by the Commission] in its legislative proposals [are] supported by the content of those assessments." The latter contain "information constituting important elements of the EU legislative process, forming part of the basis for the legislative action." As a result: [T]he disclosure of those documents is likely to increase the transparency and openness of the legislative process as a whole…and, thus, to enhance the democratic nature of the EU by enabling its citizens to scrutinize that information and to attempt to influence that process. (ClientEarth v. European Commission, 2018, §92) These considerations led the Court to qualify those documents as legislative and, further in the judgment, to reject-contrary to the General Court-the general presumption of confidentiality of those documents (ClientEarth v. European Commission, 2018, § §68-70, 85-86, 89-93, 102 et seq.).
If the ClientEarth case was crucial in determining by which set of transparency standards the pre-initiative stage of the legislative process is governed, it joins what constitutes the vast majority of the CJEU's case law in the field of transparency: the case law defining the limits of Article 4 ("Exceptions") of Regulation No 1049No /2001No (2001. As exceptions, the CJEU must proceed to a restrictive interpretation. In this endeavor, the CJEU must decide which institutions' arguments for confidentiality are acceptable and which are not. The CJEU sometimes goes beyond this binary exercise by giving, through an obiter dictum, an indication on how the institutions' arguments should be reframed in the future to be acceptable de lege lata. This case law has progressively framed the spectrum of what should or should not be transparent, despite presenting sometimes some obscurities and contradictions (Adamsky, 2009;Maiani, Villeneuve, & Pasquier, 2010, p. 16).
For instance, on what documents can be deemed 'sensitive,' the General Court underlined in the De Capitani case that the mere fact that the documents at stake relate to a sensitive field of EU law "cannot per se suffice in demonstrating the special sensitivity of the documents" (De Capitani v. European Parliament, 2018, §89). To hold otherwise would exempt this whole field of EU law from the transparency requirements. The sensitivity point is even less relevant considering that the documents in that case concerned "a draft regulation, of general scope, binding in all of its elements and directly applicable in all the Member States, which naturally concerns citizens" and affect their rights (De Capitani v. European Parliament, 2018, §90). The General Court specified that the reason of sensitivity could only be successful if the information contained in a document is "particularly sensitive to the point of jeopardizing a fundamental interest of the EU or of the Member States if disclosed" (De Capitani v. European Parliament, 2018, §97). On the 'risk of external pressure' that would result from making documents publicly available, the General Court insisted in the same case that "co-legislators must be held accountable for their actions to the public" (De Capitani v. European Parliament, 2018, §98 ). Yet the General Court acknowledged that the risk of external pressure could constitute a legitimate ground for restricting access to documents if its reality is "established with certainty, and evidence must be adduced to show that there is a reasonably foreseeable risk that the decision to be taken would be substantially affected" (De Capitani v. European Parliament, 2018, §99). We see here a concrete example of that delicate balance we referred to above: The General Court keeps the door open to pay judicial respect to the text, but the door is so narrowly ajar that it appears difficult to go through.

A Delicate Position in View of the Principles of Institutional Balance and of Institutional Autonomy
The limits to the right of access to documents expressly recognized by EU primary and secondary law prevent the CJEU from consecrating an absolute right of access (ClientEarth v. European Commission, 2018, §77;De Capitani v. European Parliament, 2018, §112), even for legislative documents. In the De Capitani case, the General Court acknowledged, as suggested by the three institutions, that the "widest possible access" as provided for in-inter alia-Article 1 of Regulation No 1049/2001, cannot be regarded as equivalent to "absolute access" (De Capitani v. European Parliament, 2018, §112). The same consideration by the Court of Justice in the ClientEarth case indicated to the European Commission that it can reject a request at the condition to duly justify its decision by motivating why the disclosure would seriously undermine its ongoing decisionmaking processes, quod non in that case (ClientEarth v. European Commission, 2018, § §123-124). This is partly explained by the limits of interpretation we elaborated on above, but also by a more fundamental consideration. As we can see in the case law on Regulation No 1049/2001, the CJEU's margin is thin: Its ruling options are binary in essence, namely to consider that the institution should have given access or not, despite the debate on openness being more complex. It is not only about giving access or not, but rather when and how; in other words, about allowing or not the scrutiny and participation of any interested citizen while a legislative act is in the making. The CJEU recognizes this by highlighting transparency as a precondition for the exercise by EU citizens of their democratic rights. However, the CJEU fails to give full meaning to that exercise and to the substance of those rights, although the elliptical provisions that are Articles 10 and 11 TEU contain the ingredients for a more ambitious agenda. Nevertheless, it cannot be blamed for this failure, as its mandate and institutional position prevent it taking further steps in this field.
The very limit that prevents the CJEU from doing so can surely be found in the principle of separation of pow-ers, or in its sort of substitute in the EU context, which is the principle of institutional balance, and in the principle of institutional autonomy. The judiciary should refrain from ultra vires marked interventions into the working of the sacrosanct legislative branch. Looking at the whole debate of determining what is the principle and what should be the exceptions through these lenses gives the debate a particularly sensitive taste, especially when considering the stances taken by the institutions brought before the CJEU by individuals or NGOs.
The De Capitani case was particularly telling in this aspect. Mr Emilio De Capitani, a former civil servant of the European Parliament, asked the latter to be granted access to the so-called four-column documents of all ongoing trilogue negotiations. The term trilogue refers to informal tripartite meetings that gather representatives of the Commission, the European Parliament, and the Council, with the aim of finding compromises on-in the present context-legislative files (Giersdorf, 2019). This is where, behind closed doors, the political agreement on a legislative file is sealed. Hence trilogues constitute the "decisive phases of the legislative process" (European Parliament, 2016, § §22, 26) or, in other words, "a substantial phase of the legislative procedure, and not a separate 'space to think"' (European Parliament, 2011, §29).
The four-column document is the central piece of the negotiations. In this document, the first column contains the proposal of the European Commission, the second the position of the European Parliament on the latter and its suggestions for amendments (if any), the third the position of the Council, and the last a tentative compromise or the preliminary positions of the Presidency of the Council in relation to the amendments proposed by the European Parliament. Often, the final text as adopted is a copy-paste of the final version of the fourth column. While this document does not report on all exchanges happening during the negotiations, it gives a clear insight into what position each institution involved defends behind closed doors, and how this position evolves during the negotiations. As Mr De Capitani's request targeted all ongoing procedures, the European Parliament rejected it as processing it would create an excessive administrative burden. Mr De Capitani therefore introduced a confirmatory application limiting the scope of the request to ongoing procedures related to specific areas. As a result, the European Parliament gave full access to five of the seven four-column documents it identified, but limited the access to the last two, refusing to disclose the fourth column of those two documents. Mr De Capitani challenged this refusal before the General Court.
The European Parliament invoked the first subparagraph of Article 4(3) of Regulation No 1049No /2001No (2001, arguing that the requested disclosure would: [A]ctually, specifically and seriously undermine the decision-making process of the institution as well as the inter-institutional decision-making process in the context of the ongoing legislative procedure and [that] no overriding public interest which outweighs the public interest in the effectiveness of the legislative procedure had been identified in the present case. (De Capitani v. European Parliament, 2018, §6) It maintained "that the principle of transparency and the higher requirements of democracy do not and cannot constitute in themselves an overriding public interest" (De Capitani v. European Parliament, 2018, §8). In support of its position, and in addition to the sensitivity and risk of external pressure points mentioned earlier, the European Parliament put forward two other arguments. First, the disclosure: [W]ould make the Presidency of the Council [warier] of sharing information and cooperating with the Parliament negotiating team and, in particular, the rapporteur; moreover, the Parliament negotiating team would be forced, on account of the increased pressure from national authorities and interest groups, to make premature strategic choices…which would 'complicate dramatically the finding of an agreement on a common position'. (De Capitani v. European Parliament, 2018, §7) Second, since the principle according to which 'nothing is agreed until everything is agreed' is "very important for the proper functioning of the legislative procedure," the disclosure "before the end of the negotiations of one element, even if it is itself not sensitive, may have negative consequences on all other parts of a dossier" and "disclosure of positions that have not yet become final risks giving an inaccurate idea of what the positions of the institutions actually are" and therefore "significantly compromise the credibility of the legislative process and of the co-legislators themselves." The European Parliament concluded that "access to the whole of the fourth column should be refused until the text agreed has been approved by the co-legislators." By making these points, the European Parliament calls in short for a time-limited confidentiality of the fourth column, "for a very brief period of time" (De Capitani v. European Parliament, 2018, § §7, 43-45, 47, 50). As the European Parliament had given access to five out of the seven identified fourth columns, the European Parliament appeared to show a quite positive attitude towards the publication of the four-column document in principle. Hence the European Parliament insisted, perhaps strategically, on the exceptional nature of its refusal.
The arguments put forward by the European Parliament demonstrated how sensitive the case was and how deep in the intricacies of the legislative procedure, "the closed technocratic machinery of the institutions" (Lea & Cardwell, 2015, p. 79), the General Court was invited to intervene. Despite its delicate position for the reasons exposed above, the General Court under-stood that allowing such arguments to be successful would inevitably open Pandora's Box. Indeed, those arguments did affect the very essence of the legislative process. The four-column documents "form part of the legislative process" (De Capitani v. European Parliament, 2018, § §38, 75, 78, 80, 98). As recalled above, it follows that they should, in principle, be made public, as: [I]t is precisely openness in the legislative process that contributes to conferring greater legitimacy on the institutions in the eyes of EU citizens and increasing their confidence in them by allowing divergences between various points of view to be openly debated. (De Capitani v. European Parliament, 2018, §78) A time-limited non-disclosure of the fourth column as requested by the European Parliament would in essence prevent citizens from exercising their rights at a very crucial point in time. As the General Court importantly notes, transparency requirements cannot be undermined by objectives of protecting the effectiveness and integrity of the legislative process (De Capitani v. European Parliament, 2018, § §81, 83). The efficiency of the process is therefore not a successful argument to refuse access to documents.
Yet the judgment of the General Court gives a middleground solution in the sense that one could argue that debates should be livestreamed to give full publicity to the exchanges. Indeed, the General Court fell short of saying that trilogue meetings should take place in public. On the contrary, it accepts with deference the necessity to keep the "possibility of a free [in the sense of confidential] exchange of views" (ClientEarth v. European Commission, 2018, §106) between the co-legislators, although the Treaties explicitly foresee the publicity of activities of the co-legislators when considering legislative files. Therefore, the General Court preserved a certain margin of maneuver for institutions to reorganize their relations in the framework of legislative procedures. Again, we see here the expression of that delicate balance that the CJEU must strike. It can explain why the European Parliament decided not to appeal the judgment of the General Court, not to risk obtaining a more unfavorable position from the Court of Justice. Such a balanced position of the CJEU is also exemplified by the position of the Court of Justice when, in its ClientEarth ruling, it gives a moderate interpretation of Article 11(2) TEU (2016) by saying that "that provision in no way means that the Commission is required to respond, on the merits and in each individual case, to the remarks it may have received following disclosure of a document" (ClientEarth v. European Commission, 2018, §106).

Time Is of the Essence
Another limitation to the CJEU's substantive contribution to improving transparency of the legislative process has to do with a simple yet consequential concept: time.
As both the General Court in the De Capitani case and the Court of Justice in the ClientEarth case ruled, the appropriate exercise by EU citizens of their democratic right to participate in the legislative process requires that they gain access to the information in a timely manner, at a critical stage of the procedure, namely while the debate is still ongoing, and a decision has not yet been taken.
Since the key political debates surrounding legislative proposals are taking place in trilogue meetings, this is therefore essential to give access in extenso et omni tempore at least to the only written source giving a dynamic account of the discussions. The access to this information is required by, on one hand, the model of representative democracy, in which citizens should be able to hold their elected representatives accountable for the positions they take and, on the other hand, to allow the same citizens, in one way or another, to take part directly in an ongoing procedure (De Capitani v. European Parliament, 2018, § §36, 41). However, in the position it defended in the De Capitani case, the European Parliament called, in short, for a time-limited confidentiality of the fourth column, "for a very brief period of time" (De Capitani v. European Parliament, 2018, § §7, 43-45, 47, 50). But, as already underlined, a time-limited non-disclosure of the fourth column would in essence prevent citizens from exercising their rights at the very crucial point in time. On the provisional nature of the information contained in those documents, the General Court insisted on the fact that the public "is perfectly capable of understanding that the author of a proposal is likely to amend its content subsequently" (De Capitani v. European Parliament, 2018, §102). In the ClientEarth case, the Court of Justice emphasized that: [T]he possibility for citizens to scrutinize and be made aware of all the information forming the basis for EU legislative action…presupposes not only that those citizens have access to the information at issue…but also that they may have access to that information in good time, at a point that enables them effectively to make their views known regarding those choices [before any decision is taken]. (ClientEarth v. European Commission, 2018, §84, read in conjunction with § §46-47; emphasis by the author) To this end, the Court considered-again contrary to the General Court-that: Not only acts adopted by the EU legislature, but also, more generally, documents drawn up or received in the course of procedures for the adoption of acts which are legally binding in or for the Member States, fall to be described as 'legislative documents' and, consequently, subject to Articles 4 and 9 of that Regulation, must be made directly accessible. (ClientEarth v. European Commission, 2018, §85, read in conjunction with § §68-70) Nevertheless, the length of the procedures before the CJEU renders the latter unable to satisfy the expectations of individuals appearing before it in that participatory aspect. Judges and legislators have "fundamentally different time horizons" (Alter, 1998, p. 122). In both the De Capitani and the ClientEarth cases, all documentssome in their final instead of intermediary version-at issue had been made available to the public, and decisions taken in the respective (pre)legislative procedures, by the time the courts gave their judgments. In addition, on a procedural note, the necessary interest in bringing proceedings that must be demonstrated by the individual is still subject to debate in the present field, especially in the case of an appeal (see, for instance, the recent judgment in the case Päivi Leino-Sandberg v. European Parliament [2021]).
In other words, today's claimants are fighting for tomorrow transparency's activists. This is only true if, and only if, judgments are followed by effective changes of the transparency policies of the institutions concerned. On this note, obtaining a judgment of the CJEU still poses the question of its enforcement; should a judgment of any of the two courts of the CJEU not be respected by an institution, the only remedy available to an individual would be an action for damages under Article 340 TFEU (2016).

A Risk of Perverse Effect
Furthermore, the CJEU's action in this field does not only suffer from these three limitations, but also entails the risk of having a perverse effect. Indeed, the mission to interpret the transparency requirements entails a risk for transparency activists, as it is fixing the borders between what should be transparent and what should not. In the cases above-mentioned this was particularly salient when the CJEU qualified what qualified as sensitive information and what did not. Stéphanie Novak's (2014) work on the transparency of the Council highlighted such a potential perverse effect by emphasizing the wide margin of discretion that institutions enjoy in the implementation of transparency rules. Moreover, when it comes to dealing with informal mechanisms such as trilogues, pushing for more transparency inevitably pushes the informality a little further. A judgment like the De Capitani ruling can have the negative consequence of showing to the institutions concerned how they should organize elements or discussions they want to stay confidential. Despite a formal procedure being laid down in Article 294 TFEU (2016), the ordinary legislative procedure has become increasingly informal in the last two decades (Reh, Héritier, Bressanelli, & Koop, 2011). The De Capitani case epitomizes the difficulty of grasping the substance of informal exchanges for the sake of transparency. It equals the endeavor of trying to make transparent what is inherently-to some extentpassing below the radar. That the General Court says that the fact that a document has "been produced or received in a formal or informal context has no effect on the interpretation" of the rules on access to documents (De Capitani v. European Parliament, 2018, §101) has little effect on the inherent difficulties linked to that phenomenon.

Conclusions
Is the CJEU able to contribute to improving the openness of the EU legislative process? If the CJEU indeed has the capacity to do so, its capacity is limited, and its actions can lead to paradoxically reducing the transparency of the legislative process. This article aimed to shed light on the limits to the role that the CJEU can play in the transparency of the legislative process debate.
The four considerations highlighted in this contribution-the limits of interpretation, the principles of institutional balance and institutional autonomy, time, and the risk of perverse consequences that the interpretation exercise entails-impede a broader and more ambitious action of the CJEU in the field of transparency of the legislative process. These issues cannot all be addressed. The fourth issue-the risk of perverse consequences-is inherent to the interpretation exercise. However, the EU legislator could reduce the effects of the three limitations in the context of a revision of Regulation No 1049/2001. It could define a new equilibrium in the dynamics in place between the CJEU and other institutions by providing the public but also the CJEU with provisions increasing the transparency requirements of legislative documents, and delineating the scope of the legislative action, drawing from the case law. However, such a reform, awaited for more than a decade, remains politically sensitive (Curtin & Leino-Sandberg, 2016, p. 5;Driessen, 2012, pp. 269-270). Interestingly, Hillebrandt, Curtin, and Meijer (2014, p. 15) note that "progressive clarification of Regulation No 1049/2001 by the courts has rendered it more difficult for a Council majority to accept this regulation as a starting point." The EU institutional law aficionados can still nourish the secret hope that the upcoming Conference on the Future of Europe could constitute a momentum. In any case, a revision of the Treaties is not necessary, as long as the CJEU fully grasps the potential of some of its provisions such as Articles 10 and 11 TEU (2016). Nevertheless, the substantial improvement that a revision of Regulation No 1049/2001 could bring would be to insert a new type of urgent procedure allowing the CJEU to act swiftly when a request is rejected. This would lead to a new institutional balance in the matter, and as such would need to carefully avoid slowing down the legislative procedure, otherwise risking again to damage the principles of institutional balance and institutional autonomy. The obvious question is: Why would the European Parliament and the Council proceed to a revision in that direction as they would be considerably impacted in their legislative work? One could argue that the potential enormous number of requests that could ensue from such a revision, as feared by these institutions, might be overestimated by the latter. On a more positive note, the co-legislators could grasp the political interest of enjoying a greater legitimacy thanks to a stronger transparency apparatus. Yet the insertion of such an urgent procedure might involve amending the Treaties or, at least, the Statute of the CJEU.
Whatever reform could take place, transparency activists should refrain from putting all their hopes in judicial interventions, as only self-regulating exercises by the three institutions concerned seem to lead to concrete and tangible results. As the current President of the CJEU put it, writing about the principle of democracy: [I]t is by progressively narrowing the gap between our conception of an ideal form of government and the government which actually rules over us that the former becomes less utopian, as society grows more receptive to the practical reforms implied by those ideals and more of them come to be realized. Ironically, we may never close that gap…since new utopian thoughts have always been the dynamic force through which mankind has moved forward. (Lenaerts, 2013, pp. 314-315) This might hold true for transparency too.