Democracy at the Borders: Analyzing the Role of Democratic Principles in EU External Relations Law

Table of contents: 1. Introduction – 2. Democracy in the Constitutional Case Law of the CJEU – 2.1 Protecting the Prerogatives of the European Parliament – 2.2 The Choice of Legal Basis: Titanium Dioxide – 2.3 Limits on the Ability to Delegate Powers to the Executive – 2.4 Democracy versus Institutional Balance – 3. Democracy in the EU’s Foreign Affairs Constitution – 3.1 Mixed Agreements – 3.2 Article 218(9) TFEU Council Decisions – 3.3 Non-binding Instruments – 4. Is Democracy in Need of a Constitutional Upgrade? – 5. Conclusion

Abstract: This paper critically examines the constitutional status and practical role of democracy within European Union (EU) law, with an emphasis on external relations. Although democracy is enshrined as a fundamental value in Article 2 of the Treaty on European Union and further operationalized in other Treaty provisions, its constitutional significance remains limited compared to principles such as conferral or institutional balance. The analysis reveals that democracy functions primarily as a weak background principle, informing but not directly determining the judicial review of secondary EU law measures. This status reflects the historical absence of a robust democratic framework in the EU’s institutional design and the comparatively limited powers of the European Parliament, the Union’s main democratic body. The paper traces the evolution of democratic principles within EU law, highlighting the gradual inclusion of democracy as a membership criterion and the ongoing scholarly debates over the EU’s democratic deficit. Through a descriptive review of key CJEU case law (including Roquette Frères, Titanium Dioxide, and Chernobyl), the paper shows that democracy is seldom invoked as a decisive constitutional principle. In external relations, democratic concerns have featured implicitly rather than explicitly in legal controversies. The paper concludes that meaningful democratization of EU external relations will require not only strategic litigation based on specific Treaty provisions but also substantive Treaty reform to strengthen the constitutional standing of democracy within the EU legal order.

Keywords: democracy – European Union – constitutional principles – external relations – CJEU – Treaty reform.

1.   Introduction

The legal status of democracy in the European Union’s legal order is paradoxical. On the one hand, democracy is presented in the Treaties as a fundamental principle. Democracy is listed in Article 2 of the Treaty on European Union (TEU) among the EU’s core values. Articles 9 to 12 TEU further operationalize this commitment to democracy. Article 10(1) TEU stipulates that the functioning of the EU shall be ‘founded on representative democracy’, and subsequent articles emphasize the importance of political parties, transparency, civil society and national parliaments to the EU’s democratic life. These references to democracy testify to the principle’s importance in the EU’s legal order. On the other hand, the principle plays a small role in the case law of the Court of Justice of the EU (CJEU or the Court). This may come as a surprise, in particular given how little the Treaty text tells us about what ‘representative democracy’ means and what it legally requires.

The meaning of democracy and how to achieve it remains a debated topic among political theorists. Defining democracy within the EU is particularly complex, given democracy’s historic ties to popular sovereignty and statehood.[1] This paper does not contribute to this debate. Instead, it explores the status of democracy in EU law, with a particular focus on the EU’s external relations. The paper examines what role, if any, the concept of democracy has played in the constitutional case law of the Court of Justice, and it addresses the question of whether democracy should be understood as a self-standing constitutional principle, similar to other principles that are explicitly mentioned in the Treaties, including those of conferral and institutional balance. The analysis concentrates on the requirements that democracy imposes on EU institutions, rather than on Member States; its focus is thus,horizontal, not vertical.[2]

The paper argues that, as EU law currently stands, and despite its inclusion in the list of fundamental values in Article 2 TEU, democracy cannot be understood as a self-standing constitutional principle akin to the principles of conferral or institutional balance. Rather, it is best understood as a weak background principle, informing the meaning of other principles – in particular that of institutional balance – without however, operating as a (potential) yardstick for the judicial review of measures of secondary EU law, as is the case for the principles of conferral and institutional balance.

This weak status of democracy in EU law should not come as a surprise, considering that for a long stretch of time the Treaty Framers expressed little interest in organizing the EU’s own institutional framework along democratic lines similar to those encountered at the national level. As is well documented, democracy did not play a role of significance in EU constitutional discourse in the first three decades after the Treaty of Rome.[3] To the contrary, the EU can be understood as an effort to curtail (national) democratic control over policy-making to enable Member States to credibly commit to long-term policy decisions.[4]Democracy nonetheless emerged as an issue of concern in the runup to the (then) EEC’s enlargement to countries with a recent history of authoritarianism such as Greece or Spain, leading to the inclusion of democracy as a prerequisite for EU membership in the 1993 Copenhagen Criteria.[5] Shortly thereafter, the democratic nature of EU decision-making itself came under closer scrutiny, leading to discussions in scholarship on the EU’s alleged democratic deficit.[6] Concerns over the EU’s lack of democratic legitimacy in turn fed into discussions that led to the draft European Constitution, which contained a chapter on the EU’s democratic life. This chapter found its way into the Treaty of Lisbon in its provisions on ‘democratic principles’.[7]

The inclusion of provisions on democratic principles in the Treaty on European Union is a reflection of the fact that, over time, individual citizens represented in the European Parliament have contributed to EU decision-making with an additional source of democratic legitimacy.[8] This is in addition to the indirect democratic legitimacy provided by the Council representing the Member States. Yet this direct democratic legitimacy remains weak, as is reflected also in the institutional position of the European Parliament as a representative body that lacks many of the powers enjoyed by its Member State counterparts, including a right of initiative.[9]

Based on a descriptive analysis of the current constitutional status of democracy in EU law, this paper argues that little is to be gained at present from invoking a self-standing ‘principle of democracy’ before the Court of Justice in an effort to challenge the validity of EU measures. Rather, a more rewarding strategy would be to unlock the Lisbon Treaty’s untapped democratic potential by bringing cases before the Court on the basis of more specific Treaty provisions, thereby granting the Court opportunities to interpret those provisions in a democracy-enhancing way, as it has done on several occasions in the past.[10] A more far-reaching democratization of the European Union will, however, have to pass through Treaty reform, in accordance with the procedures laid down to this end in the Treaties.

The remainder of the paper is structured as follows. A first section explores the role of democracy in the Court’s constitutional case law on the democratic credentials of the EU, highlighting seminal cases including Roquette Frères, Titanium Dioxide and Chernobyl. A second section focuses on the role of democracy in EU external relations, and how constitutional controversies in this area have had important democratic aspects, even though they have not been discussed in those terms. A third and final section tackles the question of whether democracy is in need of a constitutional upgrade in the EU. It develops the point that the weak status of democracy in the EU’s unwritten constitution is in keeping with the comparatively weak powers of the European Parliament as the main institutional manifestation of democracy at EU level. It also develops the point that litigation before the CJEU can be a useful means to incrementally democratize EU decision-making in external relations, but that a more far-reaching democratization requires Treaty change.

2.   Democracy in the Constitutional Case Law of the CJEU

As far as the case law of the Court of Justice goes, in numerical terms, democracy does not play a significant role. A search on the CJEU website for the terms ‘principle of democracy’ yields 81 results.[11] This stands in contrast to the ‘principle of conferral’ (200 results), the ‘principle of subsidiarity’ (728 results) and certainly the ‘principle of proportionality’ (8923 results). ‘Institutional balance’ is mentioned 426 times, and the ‘principle of institutional balance’ 125 times. The ‘principle of sincere cooperation’ is mentioned 800 times. In quantitative terms, then, democracy is not often discussed in Luxembourg.

Numbers of course only tell us that much. If we look at the Court’s constitutional case law, we see that the Court has never formally recognized ‘democracy’ as a self-standing general principle of EU law. General principles, Takis Tridimas explains, are ‘fundamental unwritten principles of law which underlie the [EU] law edifice’.[12] Within EU law’s hierarchy of norms, they rank alongside other rules and principles of primary EU law, such as Treaty provisions, and thus above norms of secondary EU law such as directives or regulations. They have three functions: (i) they allow the CJEU to fill gaps left in the written primary EU law, (ii) they steer the interpretation of norms of primary and secondary EU law, and (iii) they operate as yardsticks for the judicial review of norms of secondary EU law.[13] The CJEU discovers general principles as it fulfils its task of ensuring that in the interpretation of the Treaties ‘the law’ is observed.[14] There is no specific list of criteria a principle must meet to quality as a general principle. The CJEU’s approach is flexible, drawing on different sources – including the common constitutional traditions of the Member States – in an effort to discover general principles that best ‘fit’ the EU legal order. 

Under this conception, has the CJEU recognized a ‘principle of democracy’ in EU law? How does the Court itself understand the constitutional status of democracy under the EU Treaties? To begin to answer these questions, this paper looks at three bodies of case law in which ‘democracy’ has played a role, either directly or indirectly: First, case law that directly concerns the powers of the European Parliament. Second, case law about the choice of legal basis of EU measures. And third, case law about the delegation of powers by the legislature to the executive, be it the Commission or decentralized agencies. As will become clear, while a principle of democracy has occasionally influenced how the Court has interpreted (other) rules or principles of primary EU law, the Court has never annulled a measure of EU law violating a principle of democracy. To be sure, this should not be understood to imply that democracy lacks constitutional significance. Several judgments can be read as efforts by the CJEU to strengthen the democratic credentials of EU decision-making, thereby recognizing a modest but real constitutional role for a principle of democracy as a background principle of EU law – a principle that may not operate as a self-standing yardstick for judicial review, but which does influence and steer the interpretation of other norms of primary EU law.

2.1.  Protecting the Prerogatives of the European Parliament

If we look at the case law of the Court in which ‘democracy’ has played a role, we see that on several instances the Court has interpreted the Treaties in such a way as to strengthen the position of the European Parliament, thereby positioning itself as a promotor of EU democracy.[15]

Roquette Frères is a well-known example.[16] In this case, decided shortly after the first direct elections for the European Parliament in 1979, the Court held that the obligation on the Council to consult with the European Parliament in the process of adopting a regulation constituted an essential procedural requirement. Failure to comply with the obligation thus makes the ensuing regulation susceptible to annulment by the Court. The Court invoked the abovementioned ‘fundamental democratic principle’ in Roquette Frères, but it did so in conjunction with the notion of an ‘institutional balance’, thereby raising the question whether ‘democracy’ should be considered a self-standing constitutional principle, or whether it is subsumed by an overarching principle of institutional balance. The Court:

‘The consultation provided for in the third subparagraph of Article 43 (2), as in other similar provisions of the Treaty, is the means which allows the Parliament to play an actual part in the legislative process of the Community. Such power represents an essential factor in the institutional balance intended by the Treaty. Although limited, it reflects at Community level the fundamental democratic principle that the peoples should take part in the exercise of power through the intermediary of a representative assembly. Due consultation of the Parliament in the cases provided for by the Treaty therefore constitutes an essential formality disregard of which means that the measure concerned is void’.[17]

In a similar spirit, in the 1990 judgment in Chernobyl the Court held that the European Parliament has standing to bring an action for annulment to challenge the validity of EU legislative acts, despite this possibility not being explicitly mentioned in the then EEC Treaty.[18] The Court held that denying the Parliament this right ran counter the institutional balance whereby the Treaties ‘set up a system for distributing powers among the different [EU] institutions, assigning to each institution its own role in the institutional structure of the [EU] and the accomplishment of the tasks entrusted to the [EU]’.[19] The Court neither referred to Roquette Frères, nor did it mention the ‘fundamental democratic principle’ introduced in that case. However, by referring to the institutional balance as a ‘system’, and on that basis recognizing a right to the European Parliament that was not explicitly mentioned in the Treaty, the Court suggested the institutional balance was more than the legal translation of the intention of the Treaty framers. The ‘fundamental democratic principle’ introduced in Roquette Frère arguably provides the missing link, enabling the Court to fill the gap it had identified.

More recently, in the case of Mauritius the Court used the language introduced in Roquette Frères to give a broad meaning to the requirement in Article 218(10) TFEU that the European Parliament be kept fully and immediately informed of negotiations to conclude international agreements.[20] Somewhat formalistically, the Council had argued that it had complied with the requirement by publishing the decision to conclude the agreement in the Official Journal of the European Union. The Court rejected this narrow interpretation of the Treaty provision and held, as was the case for the obligation to consult the European Parliament at issue in Roquette Frères, that the obligation to inform the European Parliament laid down in Article 218(10) TFEU constituted an essential procedural requirement. It did so in by now familiar language:

‘That rule is an expression of the democratic principles on which the European Union is founded. In particular, the Court has already stated that the Parliament’s involvement in the decision-making process is the reflection, at EU level, of the fundamental democratic principle that the people should participate in the exercise of power through the intermediary of a representative assembly…’[21]

In these cases, we see the Court effectively strengthening the Parliament’s powers, and thus the democratic legitimacy of EU law-making. Yet it justifies these interventions primarily not by reference to a self-standing principle of democracy, but by reference to the intention of the Treaty framers who had set up an institutional balance of which the European Parliament is an integral part. At the same time, however, the Court also seems to use democracy as an (additional) interpretative tool to interpret existing Treaty provisions, as was the case in Mauritius. Faced with different interpretative options – one advocated by the Council, the other by the Parliament – the Court opted for the latter over the former because it interpreted the Treaty text in light of the ‘fundamental democratic principle’ that the people should participate in the exercise of power through the intermediary of a representative assembly. While in Mauritius the Court also saw itself as enforcing the institutional balance laid down by the Treaty framers, a principle of democracy seems to lead the Court to interpret that balance in a democracy- (read: EP-) enhancing way.

2.2.  The Choice of Legal Basis: Titanium Dioxide

Considerations related to the democratic nature of EU decision-making also played a role in the Court’s early case law on the choice of legal basis. In Titanium Dioxide, the Court was asked to review the choice of legal basis of a 1989 directive on ‘procedures for harmonizing the programmes for the reduction and eventual elimination of pollution caused by waste from the titanium dioxide industry’.[22] Titanium dioxide is a white pigment used in various products like paints and food. In 2022, it was banned in the EU as a food additive due to concerns over its potential carcinogenic effects when inhaled in nanoparticle form.[23] In the late 1980s, however, the EU was in the business of regulating the waste generated by the products containing titanium dioxide.

The controversy in Titanium Dioxide revolved around the question of whether the directive ought to have been adopted on the basis of the EU’s internal market competence or its competence in the field of environmental policy. The Court determined that both legal bases were necessary, but it added that the decision-making procedures prescribed by the two legal bases were incompatible. On the one hand, the internal market legal basis provided for decision-making in the Council by qualified majority vote (QMV) along with a co-decision power for the European Parliament. On the other hand, the environmental policy legal basis called for unanimity in the Council and merely required the Council to consult the European Parliament. Combining both decision-making procedures would ‘undermine’ the ‘essential element of the cooperation procedure’, the Court held. This would, in turn, ‘jeopardize … the very purpose of the cooperation procedure, which is to increase the involvement of the European Parliament in the legislative process of the Community…’ Here, the Court once more cited Roquette Frères, stating that ‘that participation reflects a fundamental democratic principle that the peoples should take part in the exercise of power through the intermediary of a representative assembly’.[24] For this reason, the EU legislature could not have opted for a dual legal basis, and instead had to adopt the directive on a single legal basis.

In contrast to the cases discussed in the previous section, in Titanium Dioxide we do not find an explicit coupling of democracy with institutional balance. Yet implicitly, the connection is there. The Court sees itself in Titanium Dioxide as protecting the institutional balance envisaged by the Treaty framers when they introduced the cooperation procedure, a procedure involving the European Parliament as co-legislature on an equal footing with the Council. By combining legal bases, that equal footing would be undermined, as the Parliament would come to face a unified Council acting by unanimity rather than QMV. Here, too, we see the Court first and foremost enforcing what it understands as being the institutional balance as provided for by the framers. In addition, though, ‘democracy’ can be understood as providing the Court with additional support for a reading of the institutional balance that protects the prerogatives of the European Parliament.

It is important to appreciate not only the importance, but also the limits of the role of a ‘democracy principle’ in choice of legal basis controversies. In Titanium Dioxide, the Court made clear that, in the presence of two components of equal weight, a combination of decision-making rules may be impermissible if that combination would undermine a legislative procedure that aims to endow the ensuing legislative act with democratic legitimacy. By contrast, where there are no two components of equal weight to begin with, democratic considerations do not factor into the choice of legal basis analysis. The exercise is (merely) one of enforcing the intention of the EU legislature by looking at the aim and content of the measure.

The Court made the above point clearly in the Restrictive Measures case, decided in 2012.[25] At issue was the choice of legal basis of a set of restrictive measures taken against alleged terrorists. The Council had adopted the measures on the basis of Article 29 TEU and Article 215 TFEU. The European Parliament brought a case before the Court, arguing that Article 75 TFEU on sanctions in the framework of the Area of Freedom, Security and Justice was the correct legal basis. Part of its argumentation touched on the democratic legitimacy of the decision-making procedure to adopt CFSP sanctions. In particular, the Parliament noted that, internally, measures that touch on fundamental rights can only be adopted under the ordinary legislative procedure or with the consent of the Parliament (in itself a factually inaccurate premise), and that, by analogy, external measures that touch on fundamental rights therefore cannot be adopted without involvement of the European Parliament, either. For this reason, Article 75 TFEU was the appropriate legal basis, as this legal basis did grant the Parliament co-decision rights.

The Court dismissed the Parliament’s argument and concluded that the measures were lawfully adopted on the basis of Article 215 TFEU due to, inter alia, the measures’ close connection to United Nations Security Council resolutions imposing sanctions. It did not matter, the Court held, that the Article 215 TFEU procedure did not involve the Parliament, because ‘it is not procedures that define the legal basis of a measure but the legal basis of a measure that determines the procedures to be followed in adopting that measure’.[26] In the two paragraphs that followed, the Court made clear that democracy, as realized through the involvement of the European Parliament in EU decision-making, matters, but that the intention of the Treaty framers matters even more:

‘Admittedly, participation by the Parliament in the legislative process is the reflection, at Union level, of the fundamental democratic principle that the people should participate in the exercise of power through the intermediary of a representative assembly …

Nevertheless, the difference between Article 75 TFEU and Article 215 TFEU, so far as the Parliament’s involvement is concerned, is the result of the choice made by the framers of the Treaty of Lisbon conferring a more limited role on the Parliament with regard to the Union’s action under the CFSP’.[27]

While the choice of legal basis is thus first and foremost the EU legislature’s choice to make, a ‘principle of democracy’ does enter into play where a combination of legal bases would undermine the position of the European Parliament. As discussed, in Titanium Dioxide, as was the case for the cases concerning the European Parliament’s institutional powers discussed in the previous section, democracy comes into play primarily indirectly, as part of the institutional balance. Yet arguably a principle of democracy also operates directly, in conjunction with the institutional balance principle, as an interpretative aid, steering the Court toward interpretations of Treaty provisions that enhance democracy.

2.3.  Limits on the Ability to Delegate Powers to the Executive

A third and final area where democratic considerations have played a role in the constitutional case law of the Court of Justice concerns the delegation of powers by the legislature to the executive – typically the European Commission, but increasingly also decentralized agencies. 

In Meroni, the leading case on the topic of delegating powers to decentralized agencies, the Court put limits on the ability of the legislature to delegate powers to external bodies.[28] In particular, the Court limited the ability of the legislature to delegate to ‘clearly defined executive powers’.[29] By contrast, ‘discretionary powers’ could not be delegated to a decentralized agency. This was the case, as delegating a discretionary power would boil down to a ‘transfer of responsibility’ from the delegator (the legislature) to the delegate (the agency). An agency endowed with discretionary powers could in effect ‘replace the choices of the delegator by the choices of the delegate’.

When Meroni was decided in 1957, democracy was not high on the agenda. The Treaty of Rome had not yet been adopted; there was no directly elected European Parliament; and the institutional balance within the then European Coal and Steel Community was a balance between a technocratic High Authority and an intergovernmental Council. That said, as the European Parliament gained in prominence, the Meroni doctrine – and the institutional balance principle on which it relies – gained a new democratic meaning. Today, in most instances the European Parliament is the delegator along with the Council. Discretionary powers delegated to an agency are thus powers transferred from the Parliament and the Council to the agency. By putting limits on the ability of the legislature to do so, the Meroni principle protects the institutional prerogatives of both the Parliament and the Council. There is no explicit reliance on ‘democracy’ at play here; democracy is subsumed, as it is in other cases involving the institutional prerogatives of the Parliament, by the principle of institutional balance. 

A related issue is that of delegating powers to the European Commission. The Lisbon Treaty introduced a distinction between delegated and implementing acts.[30] The former are non-legislative acts of general application to supplement or amend certain non-essential elements of a legislative act.[31] The latter are non-legislative acts whereby the Commission – and, in exceptional circumstances, the Council – are called upon, as the Court put it in the Biocidal Products case to ‘provide further detail in relation to the content of a legislative act, in order to ensure that it is implemented under uniform conditions in all Member States’.[32] The distinction between both types of acts is notoriously difficult to draw, and the Court has itself conceded that, while the distinction is an objective matter, the EU legislature nonetheless has a degree of discretion in choosing between both types of instruments as it delegates powers to the executive, leading the Court to only marginally review the legislature’s choice.[33]

The choice between delegated and implementing acts has an important democratic aspect to it, as the European Parliament is involved in the process of adopting a delegated act, but not in that of adopting an implementing act. This renders the delegated act a more attractive option from the vantage point of the Roquette Frères principle that the people should participate in the exercise of power through the intermediary of a representative assembly. An argument can be constructed that the legislature should exercise the abovementioned discretion in a democracy-enhancing way, prioritizing delegated over implementing acts.[34] Yet thus far there is no trace in the case law of such a suggestion. Instead, the Court relies on an, arguably artificial, distinction between implementing legislation on the one hand, and supplementing and amending it on the other.

2.4.  Democracy versus institutional balance

As these cases demonstrate, there are hints in the case law of the Court of a ‘democracy principle’. Yet, this principle operates within the limits of the powers attributed both to the EU as a whole and to the individual institutions by the Treaty framers. The Treaty framers allocated powers to the EU. For each power – each legal basis – they settled on a specific institutional balance, which translates into varying degrees and forms of European Parliament involvement in the legislative process. The constitutional requirement that the EU operate democratically, laid down in Article 10(1) TEU mentioned earlier, operates rather as a background principle, informing the interpretation of the Treaty text (as in Mauritius) and the Treaty structure (as in Chernobyl), but arguably not as a self-standing constitutional principle that parties can invoke before the Court to challenge the validity of secondary EU law, akin to the role played by the democracy principle in the constitutional law of some Member States.[35]

The preceding analysis makes clear that the EU Treaties as they are currently conceived are best understood as a constitutional framework in which federal values outweigh democratic values, even if both sets of values play a role in EU law.[36] ‘Federal’ values are values that are connected to the protection of the autonomy of the Member States and that aim to protect the interests of those Member States within the EU’s institutional architecture. ‘Democratic’ values are values that are connected to the protection of EU citizens and that aim to protect the interests of those citizens within the EU’s institutional architecture. Where it comes to the institutional balance, the former outweigh the latter in the sense that the institutional balance is best understood as the reflection of the intentions of the framers, rather than the expression at EU level of a trias politica as understood in the liberal democratic philosophical tradition with its own, independent normative value. 

Jean-Paul Jacqué made the point that the institutional balance is nothing more than the allocation of powers by the Treaty Framers to the EU institutions quite clearly:

‘The principle of institutional balance does not imply that the authors of the treaties set up a balanced distribution of the powers, whereby the weight of each institution is the same as that of the others. It refers simply to the fact that the Community institutional structure is based on the division of powers between the various institutions established by the treaties’.[37]

Democracy is increasingly part of the balance intended by the Treaty Framers, in that, exceptions notwithstanding, the European Parliament is granted a role in the EU legislative process. That role should however be interpreted narrowly, in keeping with the framers’ intentions, rather than broadly, to realize any overarching notion of popular sovereignty, or of a Montesquieu-inspired separation of government functions on which the constitutions of many Member States are based. This is not to say that ideal theories have no role to play in the development of the EU constitution. However, as will be discussed further below, such theories ought to inform how the Treaties are amended by the Treaty framers, not how they are interpreted by the courts.[38]

3.   Democracy in the EU’s foreign affairs constitution

In the EU’s foreign affairs constitution, the democratic principle seems to play a minor role at best.[39] To be clear, Article 21(1) TEU does require that ‘the Union’s action on the international scene shall be guided by the principles which have inspired its own creation, development and enlargement, and which it seeks to advance in the wider world’. Democracy is the first principle listed in this provision, which seems to testify to its importance.[40] Likewise, the abovementioned Article 10(1) TEU on the grounding of EU decision-making in representative democracy applies as much to external action as it does to internal decision-making. 

Regardless, in the CJEU’s foundational external relations case law the democracy principle is not discussed much. Several of the abovementioned cases – in particular Restrictive Measures and Mauritius – did concern the EU’s external relations. Yet in those cases that have really defined the shape and content of the EU’s foreign affairs constitution – we can think of ERTA, Opinion 1/75, PFOS – democracy did not play a role of significance. This is interesting, because several of the questions raised in those cases could be reconceptualized as democracy issues. Three such issues are worth mentioning: mixed agreements, Article 218(9) TFEU Council decisions, and non-binding instruments.[41] They each raise democracy questions, yet in none of the cases before the Court were these questions touched upon.

3.1.  Mixed Agreements

Take the Court’s case law on mixed agreements. Mixity raises deep questions of democratic legitimacy. By having Member States and the EU conclude international agreements jointly, national parliaments are empowered, but the European Parliament is weakened. Some argue that ‘mixity’ makes EU treaty-making more democratic through the involvement of national parliaments.[42]Others argue the opposite and consider that by involving the national parliaments EU decision-making becomes less democratic as the possibility for a pan-European democratic debate to take place is stifled.[43] To make this latter point, it is useful to draw a parallel to the abovementioned Titanium Dioxide case, where the Court concluded that the combination of QMV and European Parliament consent on the one hand, and unanimity and European Parliament consultation on the other, are incompatible.

The process of concluding and ratifying a mixed agreement is in at least one respect similar to the Titanium Dioxide scenario. In a (vertical) mixity scenario, as in a (horizontal) Titanium Dioxide scenario, we typically have a combination of QMV and European Parliament consent on the one hand, and unanimity in the Council on the other. The Council's decisions are made by unanimity, even when the Treaties require a Qualified Majority Vote (QMV), because mixed agreements need to be ratified by all Member States. This ratification requirement casts a shadow over the decision-making process in the Council: it de facto turns that process into a unanimity-based process.[44] Hence, here too we have a combination of QMV and unanimity in the Council. This combination has repercussions for the institutional position of the European Parliament, which is now faced with a unified Council, making it harder for the European Parliament to bargain with the Council. This in turn weakens the Parliament’s bargaining power with the Council.

In Titanium Dioxide, such a combination of voting rules was considered incompatible. A similar question was at stake in Restrictive Measures, but the Court avoided the issue by concluding that the contested measure required only a CFSP legal basis. The issue did re-emerge in the Court’s case law on so-called ‘hybrid decisions’. In US Air Transport Agreement, decided in 2015, the Commission had challenged the Council’s practice of merging Council decisions and intergovernmental decisions in the process of concluding a mixed agreement.[45] This practice represented a formalization of the informal dynamics described in the previous paragraph: in a mixed scenario, the Council and the Member States proceed by unanimity. The Court annulled the hybrid decision as it violated the Treaty rules on Council decision-making: the Treaty required that the Council decide by QMV; hence a Council decision adopted by unanimity violated the Treaty.

The US Air Transport Agreement case was litigated as a clash between the EU legal order's autonomy, including the decision-making rules established in the Treaties, and the principle of sincere cooperation. This principle necessitates that Member States and EU institutions collaborate effectively to ensure the practical implementation of ‘mixity’. This duty of loyalty first emerged in the Court’s 1990s case law on mixity. In Opinion 1/94 on the WTO agreements, for example, the Court reminded the parties that 

‘where it is apparent that the subject-matter of an agreement or convention falls in part within the competence of the Community and in part within that of the Member States, it is essential to ensure close cooperation between the Member States and the Community institutions, both in the process of negotiation and conclusion and in the fulfilment of the commitments entered into. That obligation to cooperate flows from the requirement of unity in the international representation of the Community’.[46]

Because of the pleas advanced by the Commission, the case revolved mainly around federal problems, i.e. issues that had to do with the relationship between the EU and the Member States, rather than democratic problems that touched on the institutional position of the European Parliament. To be clear, the Parliament did intervene to support the Commission’s plea that the Council had infringed Article 13(2) TEU, laying down the principle of institutional balance, but the Parliament did not go as far as to argue that, by adopting hybrid decisions, the Council was undermining the Parliament’s institutional position and thereby acting undemocratically. This should not surprise in the particular context of the case: the contested measure was a decision to sign an international agreement, and signing a treaty does not require European Parliament consent.[47] Yet the constitutional problem at issue in US Air Transport Agreement was of course broader: it concerned not only decisions to sign agreements, but also decisions to conclude them. Council decisions to conclude international agreements do require European Parliament consent in most instances.

In Opinion 1/19, similar questions arose as in US Air Transport Agreement.[48] Opinion 1/19 was an opinion procedure, allowing the Court to take a broader look. The European Parliament had brought the case, suggesting it was aware of the democratic issues at stake. The Parliament asked questions about the possibility and necessity of splitting Council decisions to conclude the proposed international agreement (i.e. the Istanbul Convention) into separate decisions, one to be adopted by QMV and the other by unanimity. The Parliament also asked whether it was permissible for the Member States to act by ‘common accord’ in the Council, meaning that they would only adopt decisions by unanimity, even where the Treaty provided that it could act by QMV.

Against the backdrop of the Court’s Titanium Dioxide case law, the democratic implications were clear: by dividing a Council decision that, had it been a separate decision, would have required QMV, into separate decisions, one of which required unanimity, the Council aimed to ensure that the entire agreement needed unanimous Member State support. This is reminiscent of the combination at issue in Titanium Dioxide, where the combination of legal bases ‘undermined’ the (then equivalent of) the ordinary legislative procedure. The same held true for the practice of proceeding only by common accord: it hollowed out the QMV voting rule in the Council. This, too, had repercussions for the institutional position of the European Parliament. Both practices led the European Parliament to face a unified Council rather than one that could move forward by QMV.

Given these democratic stakes, and despite the fact that the procedure was initiated by the European Parliament, it is noteworthy that a democracy principle was not discussed in Opinion 1/19. Democracy was not mentioned a single time in the 52-page opinion. In its treatment of the question about the practice of deciding by ‘common accord’, the Court reaffirmed US Air Transport Agreement but refused to extend it to informal practices that had the same effect as formal hybrid decisions, such as the practice of only adopting decisions by common accord. Instead, the Court held that Member States in the Council are not obliged to proceed by common accord, but EU law does not preclude them from doing so. The Court did point to the Council’s rules of procedure, which allow a regular majority of Member States to force a vote, requiring all Member States to take position.[49] By doing so, the Court again aimed to protect the autonomy of EU law, including EU voting rules. At the same time, it rejected arguments based on the principle of loyalty that proceeding by common accord is legally required to protect the unity in the international representation of the EU.

Democracy played no visible role in these controversies, even though these federal issues related to the autonomy of EU law and the obligation that the EU and the Member States work together in a spirit of sincere cooperation have democratic implications – in particular for the European Parliament, for whom close EU-Member State collaboration boils down to a higher Member State weight in the EU treaty-making process, and thus less influence for the European Parliament.

3.2.  Article 218(9) TFEU Council Decisions

Article 218(9) TFEU points to another aspect of external relations that raises democratic questions, but which has not yet been debated in those terms. This provision allows the Council to adopt decisions ‘establishing the positions to be adopted on the Union's behalf in a body set up by an agreement, when that body is called upon to adopt acts having legal effects, with the exception of acts supplementing or amending the institutional framework of the agreement’. This procedure simplifies the ordinary treaty-making procedure in one important respect: the European Parliament is not involved. Excluding the European Parliament makes for a more practical procedure from the vantage points of Commission and Council, yet it also makes for a procedure with a lower level democratic legitimacy.

An issue that has not yet been litigated is whether Article 218(9) TFEU puts any limits on the use of this procedure. Such limits may be needed.[50] International agreements often set up bodies. These bodies sometimes have the authority to adopt decisions binding on the parties to the initial agreement. Moreover, as the Court made clear in the case of OIV, decided in 2014, the Article 218(9) TFEU procedure may be used even where the international body concerned does not hold binding decision-making powers, provided pre-existing EU legislation declares the decisions of the body to be binding as a matter of EU law.[51]

The number of international bodies for which the Council can set EU positions through the Article 218(9) TFEU procedure is thus potentially significant. The EU’s trade agreements with South Korea and Canada offer examples of agreements that delegate meaningful powers to international bodies. CETA, for example, empowers an international body to conclude agreements on the mutual recognition of professional qualifications.[52] CETA committees may also adopt decisions that are legally binding on the parties (without however going as far as to have direct effect within the EU legal order.[53]) Without wanting to exaggerate the significance of such agreements, they are nonetheless agreements with binding force that are adopted without any European Parliament involvement. This, in turn, raises questions about the democratic legitimacy of the procedure.

A parallel can be drawn to the internal context. Internally, as mentioned earlier, the Meroni doctrine puts limits on the ability of the EU legislature to delegate discretionary powers to agencies, thereby ensuring that such delegations do not alter the institutional balance established by the Treaties, including the powers of the European Parliament. No Meroni doctrine has been articulated for external delegations of powers to international bodies, even though in the external sphere, too, there is an institutional balance set out in Article 218 TFEU that includes European Parliament involvement. By articulating an external Meroni doctrine, the Court of Justice could indirectly put limits on the use of Article 218(9) TFEU. Such a doctrine could be based on the principle of institutional balance, as is the case for the (internal) Meroni principle.

An alternative option would be to involve the European Parliament in the Article 218(9) TFEU decision-making process by enforcing the requirement laid down in Article 218(10) TFEU that ‘the European Parliament be kept fully and immediately informed at all stages of the procedure’; yet doing so would run counter the rationale of the provision, which was precisely to exclude the Parliament to streamline decision-making.[54]

3.3.  Non-binding Instruments

Non-binding instruments are yet another part of the EU’s external relations that raise democratic questions, which, however, have not yet been tackled in the case law. Non-binding instruments represent an important part of the EU’s external action. They extend from mundane arrangements between administrations to facilitate cooperation, to agreements with important political or economic consequences. Non-binding agreements on critical raw materials, including with countries with atrocious human rights track records such as Rwanda, spring to mind as examples of the latter.[55]

In the Swiss MoU case, decided in 2015, the Court made clear that decisions to conclude non-binding instruments that involve acts of policy-making require Council approval and can thus not be made independently by the European Commission.[56] The case was brought by the Council, challenging a Commission decision to conclude a non-binding agreement with Switzerland to set Switzerland’s contribution to the EU budget to keep access to the internal market following Croatia’s accession to the Union. 

The Commission sought and received the Council’s approval to initiate negotiations with Switzerland. However, it did not return to the Council to seek approval for the resulting agreement. The Council argued the Commission had violated Article 16 TEU, which grants the Council the power to make policy, and that it had gone beyond its powers of external representation laid down in Article 17 TEU. The Court agreed with the Council and annulled the Commission decision. Following Swiss MoU, the Council and the Commission have agreed to a modus operandi whereby the Commission shares information and requests Council approval of non-binding instruments.[57]

As is the case with Article 218(9) TFEU, recourse to non-binding instruments excludes the European Parliament from the decision-making process. This was not discussed in the Swiss MoU case, as the European Parliament was not a party to the proceedings and did not even intervene in the case. Had it intervened, it could have made the case that Articles 16 and 17 TEU, which operate as a lex generalis in cases where the lex specialis of Article 218 TFEU does not apply, should be read in harmony with Article 14 TEU, which grants the European Parliament powers of ‘political control’. Perhaps ‘political control’ ought to be understood as requiring that the Parliament holds a power of consent over non-binding agreements – at the very least those agreements that involve ‘policy-making’, and which on that basis require Council approval.[58] If this were the case, then the Treaties effectively provide for an institutional balance covering non-binding instruments that mirrors the institutional balance that applies to binding international agreements. Such an arrangement may be impractical, but this ignores the point that Council approval is already required for non-binding agreements. It is difficult to see how an agreement can be deemed sufficiently important to require Council approval, but not sufficiently important to require Parliament approval.

On issues such as this, the argument of the Court in Roquette Frères that the ‘fundamental democratic principle that the people should take part in the exercise of power through the intermediary of a representative assembly’ may offer the missing link. The Treaties are not clear on whether or not the Parliament should have a right of consent over some or all non-binding instruments. Article 14 TEU empowers the European Parliament to exercise ‘functions of political control’, but also adds that the Parliament is to exercise these functions ‘as laid down in the Treaties’. This qualifier could be understood to mean that the European Parliament’s political control powers do not extend to soft law instruments.[59] However, a background principle of democracy as discussed earlier could be understood to require one to read Article 14 TEU in a democracy-enhancing way, whereby the qualifier ‘as laid down in the Treaties’ is to be read down to protect the full effectiveness of the Parliament’s power to exercise political control. To read Article 14 TEU as being limited in scope to legally binding instruments would allow Commission and Council to circumvent Parliament’s oversight powers.

4.   Is democracy in need of a constitutional upgrade?

The analysis undertaken in this paper reveals that democratic concerns have played a role in the constitutional development of the European Union, including in the external relations context. However, democracy has never been a central animating principle of the EU’s constitutional acquis. Federal values – see the emphasis on the intention of the Treaty framers and on the principles that limit the scope and exercise of EU competences such as conferral, subsidiarity and proportionality – have been recognized as constitutional principles, whereas democratic values operate indirectly, influencing the interpretation of Treaty provisions but not operating as grounds of review of secondary law. This is the case not only along the vertical axis (i.e. the EU-Member State relation), but also along the horizontal axis, where enforcing the institutional balance as envisaged by the Treaty framers is the priority of the Court of Justice, and it is only where that balance is unclear (as in Chernobyl), where it pulls in conflicting directions (as in Titanium Dioxide), that democratic considerations play a more overt role as a principle that influences constitutional interpretation.

Overall, and despite the innovations introduced by the Lisbon Treaty in 2009, a ‘principle of democracy’ remains a weak background principle in the EU legal order, ranked hierarchically below other principles that reflect the intention of the Treaty framers, such as conferral or institutional balance, and which operate as yardsticks for judicial review of secondary law. And yet, while the principle is thus weak, it nonetheless exists. This is visible in a case such as Mauritius, where a conception of an institutional balance is not sufficient to tip the (interpretative) scales in favour of a broad interpretation of the Parliament’s right to be kept informed.

The existence of a principle of democracy, however weak and incomplete, reflects the fact that the European Union has gradually expanded its sources of legitimacy to include not only the indirect democratic legitimacy that stems from the Member States, but also a direct democratic legitimacy that stems from the citizens of the EU directly.[60] That said, the powers of the European Parliament remain weak when compared to national parliaments, and, more broadly: the Treaty text dedicated to ‘democratic principles’ in the TEU remains scarce when compared to the TFEU’s very expansive treatment of federal values – we can think of the competence catalogue in Articles 3 to 6 TFEU, or the specific legal bases that represent the bulk of the TFEU’s provisions. In light of this scarcity of Treaty text, it should come as no surprise that this ‘principle of democracy’ in EU law is a weak principle.[61]

That democracy exists within the EU constitutional framework, but only as a weak constitutional principle, should inform the European Parliament’s litigation strategy before the Court. It serves little purpose for the Parliament to take cases to the Court of Justice, claiming that arrangement A or B, or measure X or Y does not comply with the Treaties because it lacks sufficient ‘democratic’ elements. As we have seen, democracy is not a self-standing constitutional principle in the same way as conferral or institutional balance are such principles; it does not operate as a ground of review.

More promising, however, are strategies to leverage the textual hooks that the existing Treaties already offer in an effort to further strengthen the position of the European Parliament, or to try to extend doctrines that protect the prerogatives of the European Parliament internally to the external sphere. For example, as discussed: Article 14 TEU arguably offers the European Parliament more powers than it is currently aware it has at its disposal. The Parliament could bring actions for annulment against non-binding agreements to which the European Commission is a party, or against the Council decision to approve of such agreements.[62] The Parliament could advocate for a reading of Article 14 TEU that aligns with a conception of the institutional balance in which the Parliament can fulfil its democratic role as representative of the citizens of the EU. This could mean that it must be granted a power of consent or, at the very least, be kept fully informed in the same way as it has to be kept informed of negotiations to conclude binding international agreements.

Similarly, the Parliament could bring a challenge against a Council decision adopted on the basis of Article 218(9) TFEU to invite the Court to consider introducing Meroni style limits to the ability of the EU treaty-maker to delegate powers to international bodies. Here, too, the Parliament could invoke the principle of institutional balance to make the point that a delegation of discretionary powers to international bodies would upset the institutional balance, including the prerogatives of the European Parliament. Or, finally, the European Parliament could try to extend the teachings of Mauritius, which applied to CFSP agreements, to Article 218(9) TFEU decisions by challenging such decisions on the ground that the Council did not immediately and fully inform the European Parliament during the decision-making process.

5.   Conclusion

While such litigation strategies are important, ultimately, however, it will not be through constitutional litigation that the European Union in general, or the EU’s external relations more specifically, will democratize in any fundamental way. Meaningful changes to the institutional balance must pass through Treaty change if they are themselves to be democratically legitimate.[63] Despite the introduction of provisions on the ‘democratic life’ of the EU in the Amsterdam Treaty and its successors, there arguably has not been a universally shared intention among the Member States to genuinely democratize the EU. To argue, then, that further democratization should proceed through litigation before the CJEU, itself not a democratic institution, is a strategy that faces important limitations that must be kept in mind.

As has been argued as long ago as the 1984 Spinelli report, full democratization would consist in granting the European Parliament a right of initiative akin to the ability of individual Members of Congress in the United States to propose bills – bills that can be enacted moreover against the President’s will, provided a two-thirds majority can be found[64]. It would consist, also, in abolishing the myriad of special legislative procedures that exclude the European Parliament and allow the Council to proceed by unanimity rather than QMV as we see, for example, in the context of the Common Foreign and Security Policy. In the external sphere, the Parliament could be granted a general right to grant (or withhold) its consent to international agreements the EU wishes to conclude, and the European Parliament could be granted an oversight role in the adoption of non-binding instruments, akin to the role currently played by the Council.

Each of these reforms ideally requires Treaty change to be democratically legitimate. While there is increasing recognition that Treaty reform will be required, whether to prepare for a future accession of Ukraine, to enhance the EU’s effectiveness in foreign policy, or to expand its fiscal capacity for industrial policy, such reforms will take time to implement.[65] In the interim, CJEU litigation provides opportunities for piecemeal democratization: by leveraging the textual hooks provided by the Lisbon Treaty, the European Parliament could offer the CJEU opportunities to read the Lisbon Treaty in a democracy-enhancing way. The Parliament should utilize these opportunities fully by bringing cases to the Court.