I. Introduction
The legal basis of an EU legal act is a constitutional question in EU law in the strict sense.Footnote 1 It determines the competence exercised, the procedure followed, the voting rule in the Council, the role of the Parliament, and the standard of review against which the Court of Justice of the European Union (CJEU)Footnote 2 tests the resulting EU legal act.Footnote 3 It is the point at which the EU’s political ambitions meet the limits of conferred competence set out in EU primary law,Footnote 4 and the requirement that an EU legal act be founded on a specific provision of EU primary law, expressly stated, is among the oldest and most settled propositions of EU constitutional law. The choice of legal basis of EU legal acts rests on objective factors amenable to judicial review, principally the aim and content of the measure, and, where two purposes are present, on the identification of a predominant one. Where the procedures attached to two potential EU legal bases are incompatible, recourse to a dual legal basis is excluded. These issues related to the legal basis of EU legal acts, settled since Commission v Council (Generalised Tariff Preferences),Footnote 5 Commission v Council (Titanium Dioxide),Footnote 6 and Germany v Parliament and Council (Tobacco Advertising I),Footnote 7 structure every EU legal act that is adopted within the EU legal order.
This article advances a legal basis critique of the corpus of EU legal acts adopted in the past decade. Its argument is that the choice of legal basis has, in the past several years, become a central site at which the constitutional consequences of the EU’s response to its altered external environment are being worked out, and that the legal bases on which these EU legal acts rest are increasingly chosen for measures whose impetus is external to the concerns that previously animated those provisions of EU primary law. Shifts in the global environment, whether geoeconomic and geopolitical rivalry, sustained conventional war on the European continent, instability in the EU’s wider neighbourhood, recurrent supply-chain disruptions,Footnote 8 the transformation of the multilateral trading system, or the return of industrial policy in major global economies, have generated a stream of responses translated, at scale, into internal EU legal form,Footnote 9 through newly adopted EU legal acts. The EU is, in other words, trying to bring about strategic autonomy,Footnote 10 through its own legal order. What distinguishes the present moment from the past is that the permeability between the external issues and the internal issues is now self-conscious, written into the preambles of the new EU legal acts, and that the legal bases on which these EU legal acts rest have been chosen in ways that stretch, and in some instances openly contest, the limits of what those legal bases permit. The stretching is relevant, as a matter of EU constitutional law, regardless of whether any particular EU legal act is ultimately annulledFootnote 11 or invalidated.Footnote 12
The argument developed in this article is threefold. First, three horizontal provisions of EU primary law, namely, Article 114 of the Treaty on the Functioning of the European Union (TFEU) on internal market harmonisation, Article 207 TFEU on the EU’s Common Commercial Policy (CCP), and Article 122 TFEU on emergency or crisis measures, carry an overwhelming share of the load of new EU legal acts associated with the EU’s geoeconomic and geopolitical turn, and each is now employed for measures whose impetus departs from the concerns that previously animated its use. Second, that pattern has institutional consequences, not least for the Parliament, which is formally excluded from co-decision on EU legal acts adopted under Article 122 TFEU.Footnote 13 Third, the resulting body of EU legal acts will generate a slow-moving but consequential stream of legal basis litigation, the outcomes of which will structure the next phase of the development of EU constitutional law.
The article is structured as follows. Section II sets out the legal basis doctrine itself, from Generalised Tariff Preferences, through Titanium Dioxide, to Tobacco Advertising I and its progeny, in order to fix the benchmark against which the subsequent analysis is conducted. Section III examines Article 114 TFEU on internal market harmonisation and the extent to which the Foreign Subsidies Regulation, the Critical Raw Materials Act (CRMA), and the Internal Market Emergency and Resilience Act (IMERA) push the harmonisation legal basis beyond the limits laid down in Tobacco Advertising I. Section IV turns to Article 207 TFEU on the CCP and the suite of defensive trade instruments, among them the Foreign Direct Investment (FDI) Screening Regulation, the Anti-Coercion Instrument (ACI), the International Procurement Instrument (IPI), and the amended Enforcement Regulation, read through the prism of Opinion 2/15 (EU–Singapore Free Trade Agreement), the Court of Justice’s delineation of competence in the post-Lisbon EU legal order. Section V takes up analysis of Article 122 TFEU, which has migrated from a residual emergency clause into an increasingly principal channel for crisis-related, though not always genuinely crisis-driven, measures, and interrogates the institutional implications of that journey. Sections VI and VII change the angle of observation. Where Sections III–V are organised by explicit legal basis, Sections VI and VII are organised by policy field, in the two domains in which the choice of legal basis has carried the greatest financial and institutional consequence in the period under examination. Section VI traces a single activity, EU borrowing, spread across a patchwork of legal bases, in Articles 122, 175, 212, 311, and 322 TFEU, with different procedural consequences attaching to each. Section VII traces a substantive policy objective, namely, defence and industrial policy, seen in the form of defence-industrial support, pursued through the European Defence Industry Reinforcement through Common Procurement Act (EDIRPA), the Act in Support of Ammunition Production (ASAP), the European Defence Industry Programme (EDIP) Regulation, and the Security Action for Europe (SAFE) Regulation on four distinct legal bases. The two sections thereby demonstrate the same pattern from the opposite direction, in that where Sections III–V show individual legal bases carrying measures of varied policy content, Sections VI and VII show individual policy objectives distributed across multiple legal bases. Section VIII draws together the horizontal patterns that emerge from the preceding analysis, on the operation of the centre of gravity test and the accumulation of legal bases, on the institutional interaction that the choice of legal basis now mediates, on the scope and intensity of judicial review, and on the implications for the principle of conferral. Section IX concludes by reflecting on why the choice of legal basis of EU legal acts now requires more sustained constitutional attention than it has received in the past.
II. Legal basis
The legal basis of EU legal acts is of constitutional significance.Footnote 14 The EU is founded on the principle of conferral, and thus it acts only within the limits of the competences conferred upon it by the EU Member States. Each EU legal act must therefore find its basis in a specific provision of the EU Treaties, and that legal basis must be stated in each EU legal act itself, as part of the duty to give reasons.Footnote 15 These two requirements, conferral and the duty to state reasons, combine to make the choice of legal basis a constitutional question of EU law. They were given operational form by the Court of Justice in Commission v Council (Generalised Tariff Preferences),Footnote 16 in which the Court of Justice annulled two Council regulations adopted on the basis of a generic ‘having regard to the Treaty’ formulation,Footnote 17 holding that an explicit reference to the operative provision of EU primary law is indispensable where, in its absence, the parties concerned, and the Court of Justice, are left uncertain as to the precise legal basis.Footnote 18
This Generalised Tariff Preferences judgment established that the choice between two possible legal bases for an EU legal act is not legislative discretion. The Court of Justice held that ‘in the context of the organi[s]ation of the powers of the [EU,] the choice of the legal basis for a measure may not depend simply on an institution’s conviction as to the objective pursued but must be based on objective factors which are amenable to judicial review’.Footnote 19 Those factors include, in particular, the aim and the content of the measure. This became a central proposition of the Court of Justice’s case-law on legal basis. The Court of Justice eventually stated in Opinion 2/00 (Cartagena Protocol) that ‘[t]he choice of the appropriate legal basis has constitutional significance’.Footnote 20
The doctrine on the legal basis of EU legal acts acquired its current shape in Commission v Council (Titanium Dioxide),Footnote 21 decided shortly after the entry into force of the Single European Act. The case concerned a directive on procedures for harmonising the programmes for the reduction and elimination of pollution caused by waste from the titanium dioxide industry, adopted unanimously by the Council on an environmental legal basis, although the Commission considered that the correct legal basis was what was then Article 100a EEC on internal market harmonisation, which, at the time, required a procedure involving the Parliament. In Titanium Dioxide, the Court of Justice reiterated that the choice of legal basis of EU legal acts must rest on objective factors amenable to judicial review, including the aim and content of the measure.Footnote 22 It then held that the directive pursued, indissociably, both an environmental aim and an internal market aim, and that, in such a situation, a single legal basis could not be relied on for that EU legal act, to the exclusion of the other. However, the Court of Justice added a qualification, in that where two potential legal bases prescribe procedures that are incompatible with one another, recourse to a dual legal basis is excluded.Footnote 23 A choice of a single legal basis must therefore be made, and that choice of legal basis must reflect the predominant purpose of the measure.Footnote 24 In other words, for the Court of Justice in Titanium Dioxide, combining legal bases with different procedures would disturb the institutional balance in the EU.
Two propositions of general importance followed from Titanium Dioxide, and they have been at the heart of the case-law of the Court of Justice ever since. First, where an EU legal act pursues a twofold purpose or has a twofold component, and one of those purposes or components is identifiable as the main or predominant one, the measure must be founded on the legal basis required by that main or predominant purpose. Second, dual (or multiple) legal bases are permissible only exceptionally, and only where the procedures prescribed by the two legal bases are not procedurally incompatible.
The corollary follows directly therefrom. If the predominant purpose of an EU legal act can be identified, dual legal bases are not merely unnecessary but, as a matter of EU constitutional law, in fact prohibited. An EU legal act built upon two legal bases, when one would do, is exposed to annulment for substantive infringement of EU law.Footnote 25 Where, however, two purposes are inseparably linked, with neither secondary nor incidental to the other, an EU legal act must be founded on the various corresponding legal bases. The Court of Justice has confirmed this on numerous occasions since.Footnote 26
The same set of judgments on the legal basis of EU legal acts have de facto consolidated the question of choice of legal basis of EU legal acts into a settled three-step test. First, to identify the aim and content of an EU legal act. Second, to determine whether one purpose or component is the main or predominant one of the EU legal act, in which case the measure rests on the corresponding legal basis alone. Third, only where two purposes of an EU legal act are inseparably linked, and only where the procedures are compatible, may a dual legal basis be employed.Footnote 27
That settledness, however, describes the predominant-purpose limb more accurately than the incompatibility limb, which the Court of Justice’s own practice has applied with markedly less consistency. In Commission v Parliament and Council (Rotterdam Convention II), the Court of Justice held that the contested regulation required the joint basis of then Articles 133 EC (now Article 207 TFEU) and 175 EC (now Article 192 TFEU), and dissolved the apparent procedural conflict by retaining qualified majority voting in the Council, while extending co-decision to the CCP competence, for which that procedure was nowhere provided.Footnote 28 It pressed the point further in Parliament v Council (International Fund for Ireland), combining the unanimity demanded by then Article 308 EC (now Article 352 TFEU) with the co-decision procedure of then Article 159 EC (now Article 175 TFEU), and the levelling up the participation of both the Council and the Parliament, without once referring to Titanium Dioxide.Footnote 29
Taken together, the Rotterdam Convention II and International Fund for Ireland judgments gave the appearance of having abandoned the incompatibility rule of Titanium Dioxide, or at least of confining it to its own facts.Footnote 30 The Court of Justice’s later reaffirmation of that rule, however, in Parliament v Council (Smart Sanctions) Footnote 31 did not dispel the tension so much as sharpen it, in that the two lines of authority could be reconciled, if at all, only by treating levelling-up solution as an artefact peculiar to combinations involving the flexibility clause of Article 352 TFEU, a distinction for which the case-law of the Court of Justice supplies no explicit support. As observed, once it is accepted that the pairing of any two legal bases is itself a signal that neither suffices alone, there is no reason why procedural incompatibility should evaporate, the moment the flexibility clause enters the legal basis, yet continue to bind every other combination.Footnote 32
That said, if Titanium Dioxide was the foundational ruling on the procedural dimension of legal basis of EU legal acts, then Tobacco Advertising I is the foundational ruling on its substantive dimension, at least so far as the central internal market harmonisation provision, what is now Article 114 TFEU, is concerned. In this latter seminal judgment, the Court of Justice annulled, in its entirety, a directive prohibiting the advertising and sponsorship of tobacco products.Footnote 33 The reasoning is well-known, but bears partial restatement because it has come back to the centre of debate in the new generation of EU legal acts considered later in this article. In Tobacco Advertising I, the Court of Justice held that Article 100a EEC (now Article 114 TFEU) ‘does not vest in the [EU] legislature a general power to regulate the internal market’.Footnote 34 For a measure to be lawfully adopted on that legal basis, it must genuinely have as its object the improvement of the conditions for the establishment and functioning of the internal market, meaning, in practice, that it must contribute to eliminating obstacles to the free movement of goods or services, or to eliminating appreciable distortions of competition.Footnote 35
The directive at issue in Tobacco Advertising I failed both tests, in that, according to the Court of Justice, it applied not only to advertising in tradeable media, but to forms of promotion with no genuine cross-border dimension, such as ashtrays, parasols, and posters within hotels and restaurants. As to the distortions of competition it sought to eliminate, both in the advertising and in the tobacco product sectors, the Court of Justice stated that they were not appreciable.Footnote 36 Thus, EU legal acts adopted on Article 114 TFEU must, the Court of Justice emphasised, ‘genuinely’ pursue internal market objectives, as mere abstract risks to those objectives, or mere disparities between national laws, were insufficient.
The Court of Justice has, however, been at pains in subsequent case-law to make clear that this limit is not, in itself, a prohibition, and its record of enforcing the limit has been permissive. In Germany v Parliament and Council (Tobacco Advertising II), a redrafted version of the same directive was upheld, with the Court of Justice accepting that a sufficiently robust evidentiary record on cross-border effects and on the existence, or likelihood, of appreciable distortions of competition can sustain recourse to Article 114 TFEU as a legal basis for an EU legal act.Footnote 37 Subsequent judgments, on roaming charges in Vodafone,Footnote 38 and on seal products in Inuit,Footnote 39 among others, confirmed that the provision is capable of supporting wide-ranging EU legal acts on a relatively light evidentiary showing. In Vodafone, the future disparities in national laws said to justify the measure were improbable, but the Court of Justice upheld the legal basis anyway. Tobacco Advertising I remains the near-solitary instance in which the Court of Justice has annulled an EU legal act for want of a genuine internal market object. Article 114 TFEU has, in short, always been an elastic legal basis, and the Court of Justice has policed its conditions episodically at most.
The elasticity is most visible in the interaction between Article 114 TFEU and provisions of the EU Treaties that exclude harmonisation. Article 168(5) TFEU excludes harmonisation under the EU’s public health competence, yet the Court of Justice has repeatedly accepted Article 114 TFEU as the legal basis for measures whose principal regulatory concern is the protection of public health, as in Tobacco Advertising II itself, Swedish Match,Footnote 40 and Philip Morris.Footnote 41 That acceptance sits uneasily with the predominant purpose logic of Titanium Dioxide, for if the predominant purpose of a measure is public health, the provision corresponding to that purpose excludes harmonisation, and yet Article 114 TFEU carries the measure regardless.Footnote 42 The point is relevant, because the willingness of the Court of Justice to sustain internal market harmonisation, whose principal concern lies in a field where the EU lacks harmonisation competence, is long-standing, and it is against that permissive record, not against an imagined history of rigour, that the present generation of EU legal acts falls to be assessed.
It is against this background that the central argument of this article should be read. Article 114 TFEU, and the other legal bases considered in the following sections, have long been used for regulatory purposes beyond a narrow market-making function. The new generation of EU legal acts departs from that record in the impetus of the measures concerned. These legal bases now increasingly carry measures whose primary impetus is externally generated, in economic security, strategic autonomy, and the management of dependency on third states, where their prior use responded to internally arising concerns such as public health or consumer protection. The legal character of the choice of legal basis test has remained steady. Yet the impetus of the measures adopted upon the legal bases it governs demonstrably has, and it is that shift which places the case-law of the Court of Justice under fresh and considerable strain.
III. Internal market harmonisation
Article 114 TFEU empowers the Parliament and the Council, acting in accordance with the ordinary legislative procedure, on foot of a proposal of the Commission, to adopt measures for the approximation of the laws of the Member States which have as their object the establishment and functioning of the internal market.Footnote 43 For many decades now, this internal market harmonisation provision has operated as the principal workhorse of EU secondary law.Footnote 44 It has also been the provision most consistently tested against conferral-based limits, most famously in Tobacco Advertising I,Footnote 45 where the Court of Justice annulled a directive on the ground that the provision did not vest the EU with a general power to regulate the internal market. An EU legal act adopted on that legal basis, the Court of Justice held, must genuinely have as its object the improvement of the conditions for the establishment and functioning of the internal market. Mere disparities between national laws, or an abstract risk that such disparities might later obstruct the internal market, do not suffice.Footnote 46
The conditions laid down in Tobacco Advertising I were applied permissively in the subsequent case-law. In Tobacco Advertising II,Footnote 47 Swedish Match,Footnote 48 Vodafone,Footnote 49 Poland v Parliament,Footnote 50 and Philip Morris,Footnote 51 among others, the Court of Justice demonstrated a willingness to accept a relatively light evidentiary showing of obstacles. The assessment of that permissive line has been mixed. The case-law in this regard has been characterised as a ‘drafting guide’ for the EU legislature, rather than a meaningful constraint.Footnote 52 Its implications for democratic legitimacy have been examined,Footnote 53 and questions have been raised as to the manner of EU law-making across a range of policy areas.Footnote 54 Against this background, the claim advanced in this section is not that the permissive line has been abandoned, or that Article 114 TFEU is newly elastic. The claim is that the new wave of EU legal acts adopted under Article 114 TFEU tests the conditions of Tobacco Advertising I against an altogether different type of measure, namely, EU legal acts whose centre of gravity lies in the management of external dependencies, rather than in the removal of intra-EU obstacles, or in the elimination of appreciable distortions of intra-EU competition.
For example, Article 114 TFEU simultaneously carries the digital acquis, including the Digital Markets Act (DMA), the Digital Services Act (DSA), the Markets in Crypto-Assets Regulation (MiCAR), and the Artificial Intelligence Act (AI Act),Footnote 55 as well as the Commission’s proposal for a so-called 28th regime for companies operating across the internal market.Footnote 56 The impetus for those EU legal acts is internally arising, in the contestability of digital markets, the accountability of platforms, the integrity of financial markets, and the protection of consumers and fundamental rights. Any external effect they produce, through what has been termed the ‘Brussels effect’,Footnote 57 is an effect of the legislation, rather than the explicit impetus for it.
The Foreign Subsidies Regulation (the FSR Regulation) is the most analytically interesting instrument of the current era,Footnote 58 and a natural place to begin. The FSR is based on dual legal bases of Articles 114 and 207 TFEU, combining internal market harmonisation and the CCP to address state aid by third states that distort competition inside the EU. The logic is straightforward in substantive terms, in that EU state aid control, under Articles 107–108 TFEU, disciplines state aid granted by EU Member States but has never reached state aid granted to undertakings by third states.Footnote 59 A state-owned enterprise from a third state could therefore acquire an undertaking in the EU, or win a bid under EU public procurement law using capital, in the form of state aid, that would have been prohibited as EU state aid had it originated in an EU Member State. The FSR closes that gap by conferring on the Commission an ex officio power of review, a mandatory notification regime for concentrations above specified thresholds, and a further notification regime for public procurement procedures above certain monetary amounts.Footnote 60 Redressive measures include potential divestment, the unwinding of concentrations, and the prohibition of contract award, backed by fines of up to 10% of aggregate worldwide turnover.Footnote 61
The interest, for present purposes, lies in the legal basis of the FSR Regulation, however. The FSR Regulation reframes an external phenomenon, state aid granted by a third state, as a distortion of the EU internal market, and, on that footing, legislates in part under Article 114 TFEU, and in part under Article 207 TFEU. The dual legal basis of the EU legal act has not, to date, been challenged before the Court of Justice, which might suggest that the approach is constitutionally orthodox.Footnote 62 Nevertheless, an observation is in order in this regard. The Court of Justice insisted in Tobacco Advertising I that Article 114 TFEU requires that an EU legal act based upon it genuinely harmonises national laws of EU Member States. The FSR Regulation does not, in any obvious sense, harmonise divergent national laws of EU Member States on foreign subsidies, because EU Member States generally had no such laws, given that state aid law was regulated at EU level.
Next, the CRMA,Footnote 63 a regulation adopted in 2024, illustrates the same tension more acutely. Its legal basis is Article 114 TFEU alone. The CRMA sets benchmarks for the EU’s annual consumption of strategic raw materials. That is to say, at least 10% must come through extraction in the EU, at least 40% through processing in the EU, at least 25% through recycling, and no more than 65% coming from any single third state.Footnote 64 The CRMA confers on the Commission the power to identify ‘Strategic Projects’, including projects located in third states, which benefit from accelerated permitting and from a framework that compresses the timelines for extraction and processing permits. It additionally provides for a series of ‘Strategic Partnerships’ with third states, which function, in substance, as instruments of the EU’s external relations.Footnote 65
In light of the substance of the CRMA, the question is whether Article 114 TFEU, as a legal basis, can provide for this. The conditions of Tobacco Advertising I require that an EU legal act respond to genuine obstacles arising from divergent national law, or to appreciable distortions of competition among EU Member States. It is difficult to identify divergent national laws of EU Member States on the extraction, processing, or recycling of critical raw materials that were distorting trade in the internal market prior to the adoption of the CRMA. What EU Member States shared, rather, was a common external vulnerability through their collective dependency on suppliers for rare earths, lithium, cobalt, and similar materials in third states. The centre of gravity of the CRMA, reading its recitals and its operative provisions together, lies in securing supply from third states. On that reading, Article 194 TFEU concerning energy, or Article 207 TFEU concerning the CCP, to the extent that the Strategic Partnerships are effectively commercial arrangements, would have provided more exact legal bases.
Whether the Court of Justice will revisit the line of authority originating in Tobacco Advertising I in relation to the CRMA is speculative, as no action for annulment was brought, and to date no validity challenge has arisen. The question is nevertheless a live one because the CRMA is not an isolated case. The Net-Zero Industry Act, another regulation, adopted alongside it, employs Article 114 TFEU in similar ways.Footnote 66 The broader point is that Article 114 TFEU was (and is) being asked to perform a function, namely, to serve as the legal basis for a managed response to common external vulnerabilities of EU Member States that does not map cleanly onto the rationale of Article 114 TFEU, which is to remove obstacles to intra-EU trade or correct distortions of intra-EU competition. Whether the Court of Justice will police the provision as it did in Tobacco Advertising I, and as it signalled it might in Opinion 2/94 (EU Accession to the ECHR I) that preceded it,Footnote 67 is an open question.
Also in this regard is the IMERA,Footnote 68 a regulation applicable since 2026 and establishing a standing emergency framework for the internal market.Footnote 69 Its legal bases are Article 21 TFEU on Union citizenship, Article 46 TFEU on the free movement of workers, and Article 114 TFEU. On the emergency side, the IMERA furnishes the Commission with what is, in substance, a power of civilian industrial mobilisation in the form of priority-rated orders, derogations from public procurement law, and a framework for coordinated production ramp-ups in designated crisis sectors, and related matters. It is a ‘standing’ instrument in the sense that it is adopted in non-crisis conditions, and held in reserve for activation when the criteria provided for therein are met.
Read again against Tobacco Advertising I, the IMERA tests the outer limits of Article 114 TFEU, even if adopted on the other legal bases in addition. It is not, in the Court of Justice sense, a measure that harmonises divergent national laws of EU Member States. It is rather a framework for EU-level emergency management of the internal market, triggered by events (eg, pandemics, conflicts, supply-chain failures) that arise largely from outside the EU. To the extent that the IMERA codifies responses to external disruption, it performs a function that Article 122 TFEU was historically thought to occupy. The choice of Article 114 TFEU as a legal basis, rather than Article 122 TFEU, preserves the Parliament’s role as co-legislator, but does so at the cost of further stretching what Article 114 TFEU can be used for.Footnote 70
IV. Common Commercial Policy
Article 207 TFEU, as interpreted in the case-law of the Court of Justice, confers on the EU exclusive competence over the CCP, exercised under the ordinary legislative procedure for the setting thereof and under procedures laid down in Article 207(3) TFEU for the conclusion of international agreements.Footnote 71 The extent of the scope of that exclusivity post-Lisbon was set out in Opinion 2/15 (EU–Singapore FTA), where the Court of Justice held, inter alia, that the CCP covers direct investment but not non-direct investment (ie, portfolio investment), and that certain aspects of the then proposed international agreement with a third state fell within shared competence of the EU, rather than exclusive competence of the EU.Footnote 72
EU legal acts adopted on Article 207 TFEU in recent years have shared a common substantive feature, in that they operationalise the EU’s capacity to respond to the conduct of specific third states. This is a shift away from the posture that characterised the CCP for much of its history, in which the EU’s competence over external trade was exercised primarily through WTO-consistent, non-discriminatory EU legal acts. Newer EU legal acts, among them the FDI Screening Regulation (and its proposed recast), the Anti Coercion Instrument (ACI), the International Procurement Instrument (IPI), and the amended Enforcement Regulation, are defensive by nature and by design. Trade policy in EU law has become, more openly, geoeconomic policy.Footnote 73 Each of these EU legal acts raises distinct legal basis questions.
First, the FDI Screening Regulation rests on Article 207(2) TFEU as its sole legal basis.Footnote 74 That sole legal basis reflects the Court of Justice’s confirmation, in Opinion 2/15 (EU–Singapore FTA), of the EU’s exclusive competence over FDI as part of the CCP. Whilst the regulation de facto demanded investment screening by EU Member States,Footnote 75 the regulation did not formally establish a centralised EU-level screening authority. For screening at EU Member State-level, coordination must be done vertically (with the Commission) and horizontally (among EU Member States). Upon notification, other EU Member States may comment, and the Commission may issue opinions, with a sharper role where the transaction affects a project or programme of EU interest.Footnote 76
The scope of the FDI Screening Regulation was partially clarified in Xella Magyarország, where the Court of Justice held that the FDI Screening Regulation applies only to investors constituted or organised under the law of a third state.Footnote 77 Here, authorities in an EU Member State had vetoed an intra-EU acquisition of an undertaking in that EU Member State by an undertaking in a third state, but which in turn was ultimately controlled by an undertaking in another EU Member State. The Court of Justice thus concluded that the target undertaking, being incorporated in an EU Member State, was entitled to rely on the freedom of establishment in Articles 49 and 54 TFEU, and that the veto by the EU Member State was an unjustified restriction on that freedom. In other words, the FDI Screening Regulation was not, in fact, relevant for such a situation.
The judgment of the Court of Justice in Xella Magyarország furnished the background to the Commission’s proposed recast of the FDI Screening Regulation, tabled thereafter in 2024,Footnote 78 which proposed to add Article 114 TFEU as a second legal basis to a recast regulation. The recast proposal responded to the gap exposed in Xella Magyarország by extending EU-level screening to intra-EU transactions by subsidiary undertakings that may be ultimately controlled in third states, and by harmonising the minimum sectoral scope of national screening mechanisms, and their procedural features. The proposed recast also imposes a two-phase procedure with a 15-month retroactive call-in power. The direction of travel, whatever the final shape of the text agreed to become the recast regulation, is clear, in that it will result in more intrusive screenings, more harmonised procedures, and more centralisation at EU level.
The proposed dual legal basis for the recast FDI Screening Regulation, Articles 207 and 114 TFEU, raises a similar concern. Here, Article 114 TFEU as legal basis would be deployed to reach inside the Xella Magyarország line of argument, to capture intra-EU transactions by EU-incorporated undertakings controlled from outside, on the footing that the harmonisation of national FDI screening laws improves the functioning of the internal market. That is a defensible argument in line with Tobacco Advertising I, because national FDI screening laws do genuinely differ, and those differences affect the establishment decisions of undertakings controlled from third states. Whether the Article 114 TFEU aspect of the proposed dual legal basis is doing independent work, or is merely instrumental to the Article 207 TFEU legal basis, whose objective is the control of outside investment, will be a question for the Court of Justice should the eventual recast FDI Screening Regulation be challenged, if adopted as it is envisaged in the Commission’s proposal. The broader point, for now, is that the FDI Screening Regulation, initially with a modest application of Article 207(2) TFEU as its legal basis, is in the process of being transformed into a hybrid EU legal act focusing on both the internal market and the CCP, through the addition of an Article 114 TFEU legal basis.Footnote 79
An adjacent question concerns outbound investment screening, which the Commission is considering under a separate initiative.Footnote 80 Outbound investment screening reaches undertakings established in the EU, investing in third states, typically in sensitive technologies.Footnote 81 Whether a competence for the EU is grounded in Article 207 TFEU on the CCP, in Articles 63–66 TFEU on the free movement of capital (which includes a potential third state dimension) as legal bases,Footnote 82 or in both, turns on the character of the investments reached. The Court of Justice in Opinion 2/15 (EU–Singapore FTA) drew the line at ‘direct investment’.Footnote 83 If the outbound screening EU legal acts were to reach portfolio investment, Article 207 TFEU alone would not suffice, and Articles 63–66 TFEU would have to be added, or substituted.Footnote 84
Second, the ACIFootnote 85 was adopted in 2023 in response to two specific episodes, namely, China’s economic coercion of one particular EU Member State in 2021 following that Member State’s decision to permit the opening of a Taiwanese Representative Office on its territory, and the customs duties imposed by the United States on imports of steel and aluminium from the EU. The legal basis of the ACI is Article 207 TFEU. In substance, the ACI creates a structured procedure at EU level for identifying and responding to economic coercion by third states of the EU and its Member States. For its invocation, the Council may decide, by qualified majority, on a Commission proposal, whether economic coercion exists within the meaning of the regulation. Once coercion is established, the Commission may select responses from a menu of options that includes increases on customs duties, other restrictions on trade in goods, restrictions on trade in services, exclusions from public procurement, restrictions on FDI, and access to capital markets in the EU, inter alia.Footnote 86
The breadth of the potential responses of the EU within the ACI is its most constitutionally significant feature. Read conservatively, the ACI is a mere framework for the exercise of existing powers of the EU under the CCP. On this reading, the legal basis of the ACI as Article 207 TFEU is unobjectionable. Read more liberally, however, as to the capacity of the EU to deploy a suite of economic-statecraft responses, coordinated under a single determination of ‘coercion’, this means that the ACI can function as a quasi-foreign policy EU legal act.Footnote 87 The latter reading sits uneasily with the division between CCP, which is implemented by the Council and the Commission under qualified majority, and the EU’s Common Foreign and Security Policy (CFSP), which is decided upon by the Council alone, acting near-always in unanimity under Articles 23–46 TEU. The Court of Justice has been careful to police the boundary between the CCP and the CFSP, given the different forms of institutional decision-making that apply, as seen in Parliament v Council (Mauritius),Footnote 88 Parliament v Council (Tanzania),Footnote 89 and Commission v Council (Kazakhstan),Footnote 90 all of which form part of the choice of legal basis case-law. The ACI has not been tested in this respect, and its operation to date has been limited to investigation, rather than the imposition of response measures.
Third, the defensive reorientation of Article 207 TFEU extends beyond these two EU legal acts. The IPI, a regulation adopted in 2022 on Article 207 TFEU, empowers the Commission to investigate restrictions on EU economic operators in the public procurement markets of third states, and, where consultations fail, to impose score adjustments on, or the exclusion of, tenderers from the third state concerned in procurement procedures within the EU.Footnote 91 The amended Enforcement Regulation, revised in 2021 in direct response to the paralysis of the Appellate Body of the World Trade Organization (WTO), authorises the EU to adopt countermeasures where dispute settlement is blocked, and extended the reach of those countermeasures to services and certain intellectual property rights.Footnote 92 The recast Dual-Use Regulation of 2021, also based on Article 207 TFEU, brought autonomous, human rights-driven controls on cyber-surveillance items within the EU’s export control regime.Footnote 93 Each of these EU legal acts conditions access to the EU internal market, or restricts departures from it, by reference to the conduct of specific third states. The reorientation of the CCP accordingly rests on a legislative programme, of which the FDI Screening Regulation and the ACI are the most constitutionally salient parts.
Taken together, these EU legal acts based on Article 207 TFEU establish the defensive aspects of the CCP, reconfigured for a world in which trade is increasingly a vector of strategic competition among global powers. That the CCP reaches matters adjacent to foreign policy is, in itself, unremarkable. EU law has long operated through functional connections, rather than neatly demarcated policy fields. Overlaps between policy areas are inevitable, and measures of the CCP will regularly carry foreign policy effects. The constitutional question is one of degree, namely, at what point the functional use of Article 207 TFEU as a legal basis becomes overstretched. The density and scope of the present suite of EU legal acts, and the extent to which their operation depends on politically charged determinations, that a given measure constitutes ‘economic coercion’, or that a given investment engages the ‘security or public order’ of the EU or its Member States, place that question of degree at the centre of the CCP for the first time.
V. Emergency clause
Article 122 TFEU is a short, two-paragraph provision, an emergency clause, largely removed from the traditional pattern by which the EU adopts secondary law. First, it empowers the Council, on a proposal from the Commission, to decide ‘in a spirit of solidarity between Member States, upon the measures appropriate to the economic situation, in particular if severe difficulties arise in the supply of certain products, notably in the area of energy’.Footnote 94 Second, it empowers the Council to grant EU financial assistance to an EU Member State in difficulties or seriously threatened with severe difficulties caused by natural disasters or exceptional occurrences beyond its control.Footnote 95 Both paragraphs operate by qualified majority voting within the Council. In particular, as regards the second aspect, it provides that ‘the President of the Council shall inform the European Parliament of the decision taken’. In other words, as a legal basis, it contains no parliamentary involvement whatsoever.
EU primary law contains no general definition of an emergency. The concept is distributed across a series of provisions, each construed on its own terms. Article 122 TFEU itself speaks of ‘severe difficulties’ in supply and of an EU Member State ‘seriously threatened with severe difficulties caused by natural disasters or exceptional occurrences beyond its control’, language that the Court of Justice, in Pringle, read as confining Article 122(2) TFEU to financial assistance granted by the EU itself.Footnote 96 Beyond Article 122 TFEU, the EU Treaties contain a family of emergency and safeguard legal bases such as Article 66 TFEU on safeguard measures concerning the movement of capital to or from third states, Article 78(3) TFEU on provisional measures in the event of a sudden inflow of nationals of third states, Articles 143 and 144 TFEU on balance of payments difficulties of EU Member States outside the euro area, Article 213 TFEU on urgent financial assistance to third states, Article 222 TFEU on the solidarity clause, and Article 347 TFEU on wholly exceptional situations, alongside Article 42(7) TEU on mutual assistance in the event of armed aggression.Footnote 97 What these provisions share, structurally, is a Council-centred procedure with a reduced role, or no role at all, for the Parliament, which is the institutional price that EU primary law attaches to exigency. In Slovakia and Hungary v Council (Relocation), the Court of Justice upheld provisional relocation measures of considerable scope under Article 78(3) TFEU, accepting that emergency legal bases can sustain measures that derogate from existing EU legal acts, provided the measures are temporary and respond to the exigency identified.Footnote 98 The defining features of an emergency legal basis are, accordingly, exceptionality, temporariness, and responsiveness to a specific exigency, and the question that the recent use of Article 122 TFEU raises is whether those features continue to characterise the measures adopted upon it.
Before 2020, Article 122 TFEU had been used as a legal basis for EU legal acts sparingly, most notably with the adoption of the European Financial Stabilisation Mechanism (EFSM) during the sovereign debt crisis.Footnote 99 As a potential legal basis, it was treated in institutional practice as a residual or last-resort legal basis, a provision reserved for genuine exigency, available where the ordinary procedures could not reach.Footnote 100 Two elements contributed to that reading, namely, the explicit emergency language of both paragraphs of Article 122 TFEU, and the minimal role assigned to the Parliament, which was difficult to reconcile with the general orientation of the EU Treaties towards democratic co-decision under the ordinary legislative procedure.
Since 2020, Article 122 TFEU has been deployed with far greater frequency, as seen in the EU’s response to Covid-19 through the European instrument for temporary support to mitigate unemployment risks in an emergency (the SURE Regulation),Footnote 101 and the European Union Recovery Instrument (EURI), for energy-price stabilisation, including gas demand reduction and emergency energy-price intervention, for support to Ukraine, and most recently for joint defence financing through the SAFE Regulation.
A comparison will first be drawn with the Gas Storage Regulation, adopted around the same time on an Article 194 TFEU legal basis, which illustrates that the ordinary legislative procedure can, in fact, deliver on EU objectives at comparable speed, before two of these uses are analysed. The EU’s response to the 2022 energy crisis, linked to the Russian invasion of Ukraine,Footnote 102 produced a package of seven EU legal acts adopted in rapid succession. These EU legal acts can be divided reasonably cleanly into two groups. One group of EU legal acts was adopted on legal bases that ensured the ordinary legislative procedure, with the Parliament as co-legislator. The other was adopted under Article 122 TFEU, with the Council acting on a Commission proposal, and the Parliament merely informed.
The Gas Storage Regulation 2022 belongs to the first group.Footnote 103 It was proposed by the Commission in March 2022,Footnote 104 and adopted by the Parliament and Council in June 2022 on the basis of Article 194 TFEU, a turnaround of approximately three months from proposal to adoption. The significance of that three-month turnaround bears directly on the Commission’s and Council’s argument, advanced in support of subsequent uses of the emergency clause in Article 122 TFEU as a legal basis for EU legal acts, that only it can deliver the necessary speed for the adoption of EU legal acts. The Gas Storage Regulation, adopted under the ordinary legislative procedure, moved from proposal to adoption in a time frame commensurate with the gravity of the energy crisis at the time. The subsequent resort to Article 122 TFEU for reducing the demand for gas, and intervention measures on energy prices, cannot, on the basis of that comparator, be justified purely on grounds of speed.Footnote 105
As for the second group, EU legal acts from 2022 based on Article 122 TFEU introduced gas-demand reduction targets and two revenue-side measures, a revenue cap on inframarginal electricity producers, and a temporary solidarity contribution on the fossil fuel sector.Footnote 106 The latter attracts particular interest on the basis that the solidarity contribution was, in substance, a tax, and that an EU legal act of that kind required the use of Article 113 TFEU as a legal basis, which concerns harmonisation of indirect taxes, or alternatively, Article 115 TFEU as a legal basis on the approximation of laws affecting the functioning of the internal market, either of which would have required unanimity in the Council, rather than qualified majority.Footnote 107 Such use of Article 122 TFEU as a legal basis for these two EU legal acts by the Commission and the Council raises genuine questions as to what amounts to a coordinated fiscal measure, imposed at the EU level by an emergency clause, and without parliamentary co-decision, not to mention the lack of need for unanimity within the Council. This use of Article 122 TFEU amounts to a substantive expansion of the effective reach of an ‘emergency’ clause.
Similarly, there has been the SAFE Regulation, adopted by the Council in 2025, the most constitutionally consequential of the EU legal acts to date based on Article 122 TFEU.Footnote 108 The SAFE Regulation provides for up to €150 billion in EU-issued loans to EU Member States for defence procurement. The choice of Article 122 TFEU, rather than a legal basis within the CSDP in Articles 42–46 TEU, or the legal basis related to industrial policy in Article 173(3) TFEU, has a number of consequences. The centre of gravity of the SAFE Regulation could rather be in defence policy, in industrial policy, or in support for defence procurement, and it might even relate to numerous areas, even if a dual or multiple legal basis here would be unworkable, given the different decision-making procedures of the potential alternative legal bases other than Article 122 TFEU.
Concerns have rightly been raised about the SAFE Regulation,Footnote 109 with the broader concern that the emergency clause of Article 122 TFEU risks gradually developing into a new super-legal basis, to be used without effective democratic scrutiny. The democratic dimension deserves to be stated plainly. The Parliament is the only directly elected institution of the EU, and the choice of Article 122 TFEU as the legal basis removed it from the adoption of a €150 billion defence financing commitment, among the largest fiscal undertakings in the history of the EU. That exclusion operates at precisely the point where democratic scrutiny is most in need, in the commitment of public resources on a scale that will bind the EU and its Member States for decades, and in a policy field, defence, in which national parliaments have traditionally held the purse strings.
The Parliament has contested the legal basis of the SAFE Regulation, and seeks its annulment on the ground that the Council misused Article 122 TFEU as a legal basis to bypass the Parliament’s institutional position.Footnote 110 The fundamental question that the SAFE Regulation poses, for the Parliament, is whether Article 122 TFEU has become a free-standing emergency clause to be deployed whenever the Commission and the Council have identified a sufficiently exceptional ‘economic situation’ or ‘exceptional occurrences’, or whether the provision retains a structural limit that confines it to measures whose centre of gravity is responsive to a genuine emergency. Parliament v Council (SAFE) will be a monumental judgment in this regard as to the contours of this emergency clause. Article 122 TFEU was drafted as a deliberately narrow emergency clause,Footnote 111 and it may not necessarily be utilised for means that recent practice has placed upon it without a corresponding loss of its emergency character.
The institutional implications of the Article 122 TFEU pattern are significant. The Parliament is largely silenced on what are, by almost any measure, among the most consequential EU legal acts of the period, as the SAFE Regulation is. Taking this regulation and other similar regulations together, these various EU legal acts commit the EU to expenditure in the hundreds of billions of euros. In each case of the use of the emergency clause, the Parliament was informed, rather than consulted or involved in co-decision. EU primary law can tolerate this under the emergency clause that Article 122 TFEU was designed to embody if the EU legal acts are genuine emergencies, and narrow in scope, but a series of invocations of Article 122 TFEU as an exceptional legal basis may mean that it no longer occupies the category of ‘emergency’, and has become something closer to a standing alternative to the ordinary legislative procedure.
There are structural limitations that could apply to the use of Article 122 TFEU as a legal basis for EU legal acts. First, the centre of gravity test could be applied with rigour, such that EU legal acts whose primary substantive matter is in reality defence, taxation, or industrial policy fall outside its scope. Second, the principle of institutional balance, anchored in Article 13(2) of the Treaty on European Union (TEU), requires that major commitments of the EU involve the Parliament unless EU primary law plainly excludes that involvement, such as within the CFSP.Footnote 112 Third, the emergency character of Article 122 TFEU requires a genuinely exceptional, unforeseeable, and transient set of circumstances, criteria that were met by Covid-19 initially, but are not met by a general and ongoing defence financing instrument of the type represented by the SAFE Regulation.Footnote 113 When an emergency clause becomes a routine channel for EU legal acts, it ceases to operate as an emergency clause. The frequency of invocation is itself an argument against continued characterisation of each new invocation as exceptional.
VI. EU borrowing
Joint borrowing by the EU is, in quantitative terms, a substantial innovation of the past decade as regards the EU’s finances, and in EU constitutional terms, it is arguably the most consequential.Footnote 114 Before 2020, the EU’s cumulative borrowing capacity was modest. The literature was sparser still.Footnote 115 Since 2020, however, a succession of EU legal acts, the SURE Regulation, the European Union Recovery Instrument (EURI) funded through the enlarged own-resources ceiling, the Recovery and Resilience Facility (RRF), the Macro-Financial Assistance Plus Regulation, the Ukraine Facility, the Ukraine Loan Cooperation Mechanism (ULCM), and the SAFE Regulation, inter alia, has brought cumulative authorised borrowing of the EU to a figure comfortably approaching one trillion euros in issued and yet-to-be issued debt.Footnote 116 The legal bases on which the underpinning EU legal acts rest are assorted, comprising, as will be demonstrated, Articles 122; 175, third paragraph; 212; 311, third paragraph; and 322(1) TFEU. The array of legal bases is significant because it reveals a pattern of legal basis-matching in which the choice of provision is made instrument-by-instrument, with the consequent institutional consequences.
The SURE Regulation, in May 2020, was based on Article 122 TFEU and provided up to €100 billion in EU loans to EU Member States to finance short-time work schemes and analogous measures in response to Covid-19.Footnote 117 The preamble invoked both Article 122(1) TFEU, for the guarantee scheme of voluntary EU Member State contributions, and Article 122(2) TFEU for the loan facility as such.Footnote 118 Procedurally, the SURE Regulation was adopted by the Council, on foot of a Commission proposal. As the SURE Regulation’s purpose was explicitly framed as financial assistance to EU Member States facing an exceptional occurrence in the form of a pandemic, the use of Article 122 TFEU as the legal basis attracted limited controversy in itself at the time of adoption.Footnote 119
The larger and more contested instrument was the pairing, later in 2020, of the Own Resources Decision (ORD) with the EURI. A decision of the Council amended the own-resources regime under Article 311 TFEU, third paragraph, to authorise the Commission to borrow significant funding on capital markets, solely and exclusively to address the consequences of the Covid-19 pandemic.Footnote 120 The ORD was adopted under a special legislative procedure in the form of Council unanimity, Parliament consultation, and subsequent national ratification. The EURI of the same date, based on Article 122 TFEU, channelled the borrowed funds as ‘external assigned revenue’ to several spending programmes, chief among them the RRF.Footnote 121 The RRF itself was based on Article 175 TFEU, third paragraph, the legal basis for EU cohesion policy that preserved co-decision, and anchored the reform and investment conditionality architecture.Footnote 122 Cumulatively, the ORD–EURI–RRF construction represents one of the more intricate legal basis puzzles of the time period. The ORD carried the borrowing authorisation and the expanded own resources ceiling that served as the guarantee for the debt. The EURI translated the borrowed funds into revenue available for EU spending. And the RRF was the principal programme that disbursed those funds, subject to reform and investment conditions.
Three points merit attention. First, the ORD’s legal basis of Article 311 TFEU, third paragraph, was the only element that required unanimity and national ratification because of the magnitude and the constitutional character of the own resources expansion. Second, Article 122 TFEU was used as the legal basis for the EURI, despite the fact that the EURI was not itself, in any conventional sense, an emergency act. Rather, it operated downstream of the ORD. Third, the RRF was placed on Article 175 TFEU, third paragraph, and hence subject to the ordinary legislative procedure.Footnote 123
The architecture constructed for Ukraine has been more varied still.Footnote 124 The Macro-Financial Assistance Plus Regulation of 2022, providing concessional loans, rested on Article 212 TFEU, which authorises economic, financial, and technical cooperation measures with third states.Footnote 125 The Ukraine Facility of 2024, with a specified of certain financial backing for the 2024–27 period, combined Article 212 TFEU with Article 322(1) TFEU as legal bases to cover both the external cooperation dimension and the horizontal financial rules.Footnote 126 The Ukraine Loan Cooperation Mechanism, also of 2024, implemented the EU’s share of the G7 ‘Extraordinary Revenue Acceleration’ (ERA) loans package, whose distinctive feature is that it is serviced and repaid from extraordinary revenues on immobilised Russian sovereign assets held in the EU, predominantly at Euroclear.Footnote 127
The ERA loan illustrated the interplay of EU legal acts within the CFSP and other EU legal acts that were not related to the CFSP. The immobilisation of financial assets of the Central Bank of Russia was achieved in two steps. First, a decision of the Council under Article 29 TEU,Footnote 128 and second, a regulation of the Council, under Article 215 TFEU,Footnote 129 which cumulatively required central securities depositories to pay 99% of the net profits generated on the immobilised financial assets to the EU on a biannual basis, with 90% channelled to the European Peace Facility (EPF) and 10% to the Ukraine Facility. The Ukraine Loan Cooperation Mechanism (ULCM) Regulation then created the borrowing vehicle for the EU’s contribution to the G7 package, servicing the resulting EU loans from those extraordinary revenues. The legal basis of the ULCM Regulation was Article 212 TFEU, but its operation depends, materially, on the immobilised financial assets providing the revenue stream acted upon under the CFSP. In December 2025, the Council adopted a regulation,Footnote 130 based on Article 122 TFEU, which prohibited transfers of assets of the Central Bank of Russia until decided otherwise.
The permanence effected by this December 2025 regulation is noteworthy for two reasons. First, it accomplished, through an EU legal act based on Article 122 TFEU and adopted by qualified majority on a Commission proposal, an outcome that under the CFSP would have required unanimity in the Council, and without the Commission’s involvement. The choice of Article 122 TFEU as the legal basis can be defended on the ground that the revenue stream coming from the immobilised financial assets was thereafter to be part of the EU’s fiscal architecture, servicing the EU’s borrowing under the ULCM Regulation, and therefore falls within the remit of Article 122 TFEU, given the emergency nature. That said, whether the Court of Justice would accept that framing is an open question. Second, and more generally, the permanence of the 2025 regulation in this regard shows how an EU legal act initially characterised as part of the CFSP can, over a relatively short period, be re-based into an EU legal act based on a legal basis outside of the CFSP, with different procedural features.
The SAFE Regulation, examined in Section V, principally as an EU legal act based on Article 122 TFEU, also relates to EU borrowing. The €150 billion amount it authorised was to be raised by the Commission on capital markets and lent onwards to EU Member States for defence procurement. The servicing architecture differs from that of the borrowing under NextGenerationEU (NGEU). Whereas the EURI provided that borrowing is guaranteed by the expanded own resources ceiling under the ORD, loans under the SAFE Regulation are serviced by participating EU Member States under contractual arrangements. In other words, the SAFE Regulation does not formally require a new ORD because the loans are to be repaid by the participating EU Member States, rather than by the EU budget.Footnote 131
Three observations emerge from the pattern of EU borrowing in these recent times. First, the procedural rule selected in each example has had significant institutional consequences for the role of the Parliament, which varies from full co-decision under Article 175 TFEU, third paragraph, to mere information under Article 122 TFEU. Second, the reliance on Article 122 TFEU in successive EU legal acts on borrowing, the SURE Regulation, the EURI, and the SAFE Regulation, inter alia, demonstrates a pattern whose cumulative implication is that the emergency clause is becoming a routine channel for the adoption of major EU legal acts.Footnote 132 Third, the judicial review of these arrangements has been limited, though challenges may yet come.Footnote 133
VII. Defence and industrial policy
EU legal acts adopted since 2022 related to defence and industrial policy form a distinct category within the present analysis because EU primary law contains specific provisions on defence, and the questions of legal basis of EU legal acts in this regard arise because of the substantive content of these new EU legal acts. The CFSP provides that, ordinarily, decisions are taken by unanimity within the Council.Footnote 134 The CSDP forms an ‘integral part’ of the CFSP.Footnote 135 Article 346 TFEU recognises the essential security interests of EU Member States, and permits derogation from the law of the internal market for trade in arms, munitions, and war material.Footnote 136 Article 42(2) TEU envisages the progressive framing of a common EU defence policy, which ‘will lead to a common defence’ when the European Council, acting unanimously, so decides. Taken together, these provisions related to CFSP matters and non-CFSP matters make the area of defence an untraditional policy domain for the EU. Yet, recently, there have been various EU legal acts making use of Article 173(3) TFEU on industrial policy, Article 114 TFEU on internal market harmonisation, Article 212 TFEU on assistance to third states, and Article 122 TFEU on crisis measures, as legal bases to pursue objectives that are, in substance, defence oriented.
The EDIRPA was adopted in 2023 on the basis of Article 173(3) TFEU on industrial policy, which empowers the EU to take ‘specific measures in support of action taken in the Member States’.Footnote 137 The EDIRPA furnishes financial incentives for EU Member States undertaking joint defence procurement, with the stated aim of addressing the most urgent gaps in defence capabilities in the EU arising from the war in Ukraine. The legal basis is unproblematic on its face because Article 173(3) TFEU explicitly authorises supportive measures for industrial policy, and the defence-industrial dimension is encompassed. What is noteworthy, however, is the substantive pivot of Article 173(3) TFEU, historically deployed for general industrial policy, and now utilised in the service of an adjacent objective, defence-industrial policy.Footnote 138
The ASAP,Footnote 139 a regulation adopted in 2023, illustrates the legal construction more sharply. The ASAP rests on a dual legal basis of Articles 114 and 173(3) TFEU. The use of Article 114 TFEU on internal market harmonisation has been seen as necessary because it introduces, by effect if not by the explicit terms of the regulation, a priority-rated orders mechanism to require defence manufacturers to prioritise orders.Footnote 140 A power of that kind cannot sit on Article 173(3) TFEU on industrial policy alone, because Article 173(3) TFEU excludes, in its second sentence, ‘any harmonisation of the laws and regulations of the Member States’. The combination of legal bases produces, in substance, an industrial-mobilisation power for defence applications framed as internal market harmonisation.
There is a question whether the ASAP construction fully satisfies the conditions of Tobacco Advertising I. Priority-rated orders do not, in any obvious sense, harmonise divergent national laws of EU Member States on defence-industrial production. They rather impose a new EU-level prioritisation regime that differs from what any EU Member State had in place. The ASAP expired on 30 June 2025, succeeded in function by the more comprehensive EDIP Regulation.Footnote 141 The Court of Justice is unlikely ever to address the legal basis question in relation to the ASAP, given that it has lapsed, but the same reasoning is carried forward into the EDIP Regulation. Adopted in 2025, the EDIP Regulation combines four legal bases. Article 114 TFEU supports the internal market harmonisation elements, including the priority-rated orders architecture, de facto inherited, though more advanced in textual form, from the ASAP, and Article 173(3) TFEU supports the general measures in support of industrial policy. Article 212 TFEU supports the Ukraine Facility, which channels resources under the EDIP Regulation to Ukrainian defence production, and Article 322(1) TFEU provides for horizontal financial measures. The combination of legal bases is notable for its breadth, but all four operate under the ordinary legislative procedure, so no limitation on grounds of procedural incompatibility arises in formal terms.Footnote 142
The question around the centre of gravity of the EDIP Regulation, however, is more pressing. The aim and the content of the EDIP Regulation, as the recitals and operative provisions make plain, are to reinforce the EU’s defence industrial and technological base, and to sustain Ukrainian defence production. Framed in those terms, the centre of gravity could plausibly be located in the CSDP, squarely within the CFSP. Article 41(2) TEU, in particular, provides that ‘expenditure arising from operations having military or defence implications’ shall not be charged to the EU budget. The EDIP Regulation works around this by using other legal bases. That non-CFSP legal bases carry measures with defence implications is, in itself, within the ordinary functional operation of EU law, in which policy fields overlap and legal bases reach matters adjacent to their core. The constitutional question, here as with the CCP, is one of degree, and the EDIP Regulation approaches the threshold at which functional use becomes repurposing, for its defence orientation is the substance of the measure, and Article 41(2) TEU states a limit in terms that the construction is designed to avoid. A further issue arises from the composition of the benefits conferred by the EDIP Regulation. It includes provisions restricting eligibility, in certain respects, to products originating from undertakings established in EU Member States or in associated third states as a ‘European preference’ feature.Footnote 143
The SAFE Regulation, examined earlier, completes the defence picture. Its legal basis, Article 122 TFEU, sits alongside the EDIRPA (Art 173(3) TFEU), the ASAP (Arts 114 and 173(3) TFEU), and the EDIP Regulation (a combination of four legal bases). The result is that a substantive policy area, defence support, is, across four principal EU legal acts, legislated under four distinct legal bases. The Parliament is a co-legislator under the EDIRPA, the ASAP, and the EDIP Regulation, but not under the SAFE Regulation. Each of the four can be defended on its own terms as a choice of legal basis compliant with Tobacco Advertising I or with the centre of gravity test. Taken together, however, they show that the same policy objective is being pursued through EU legal acts that, constitutionally, are being treated as four different kinds of EU legal act.
An observation of longer-term significance is that the provisions related to the CSDP, Articles 42–46 TEU, have played a markedly secondary role in the recent EU legal acts, despite being the provisions of EU primary law nominally governing the policy area concerned. Permanent Structured Cooperation (PESCO), the European Defence Agency (EDA), the European Peace Facility (EPF), and the CSDP missions remain active, but the principal financial and industrial instruments have migrated outside the CFSP. This reinforces the assessment that a ‘communitarisation’ of defence is well underway,Footnote 144 through the back door of legal bases, or, put less charitably, that non-CFSP legal bases are being used to bypass the politically costly unanimity rule that governs the CFSP in the Council. Whether the reliance on non-CFSP legal bases for defence matters is sustainable over the longer term depends in part on whether EU Member States continue to acquiesce, and in part on whether the Court of Justice will be called upon to police the boundary between CFSP matters and non-CFSP matters, a jurisdictional power specifically conferred upon it.Footnote 145
VIII. Patterns
The preceding sections have approached the choice of legal basis from two directions, first by legal basis, in Sections III–V, and then by policy field, in Sections VI and VII.Footnote 146 Read together, they disclose four horizontal patterns, which concern the operation of the centre of gravity test, the institutional interaction that the choice of legal basis now mediates, the scope and intensity of judicial review, and the condition of the principle of conferral.
The first pattern concerns the centre of gravity test itself. The doctrine set out in Section II directs the EU legislature to a single legal basis wherever a predominant purpose can be identified, and admits dual or multiple legal bases only exceptionally, where purposes are inseparably linked and the attached procedures are compatible.Footnote 147 The legislative practice examined earlier proceeds otherwise. It exhibits, on the one hand, an accumulation of legal bases, in that the FSR Regulation combines Articles 114 and 207 TFEU, the recast proposal for the FDI Screening Regulation adds Article 114 TFEU to Article 207 TFEU, the ASAP combines Articles 114 and 173(3) TFEU, and the EDIP Regulation rests on four legal bases.Footnote 148 It exhibits, on the other hand, an avoidance of legal bases, in that the SAFE Regulation rests on Article 122 TFEU alone, in preference to any alternative that would have exposed the measure to the predominant-purpose inquiry on contestable terms.Footnote 149
What unites the two techniques is that neither infringes the compatibility rule of Titanium Dioxide. The rule is, instead, engineered around. The combinations selected are combinations whose procedures happen to align, principally because some constituent legal basis carries the ordinary legislative procedure, so the procedural incompatibility that compelled a choice in Titanium Dioxide never materialises.Footnote 150 The predominance question, whether one purpose is the main one, and the others merely incidental, is answered in advance by recitals asserting that the purposes are inseparably linked, and no institution whose procedural position is preserved by the combination has any incentive to litigate the assertion. The doctrine therefore remains formally intact while its selective function, the function of forcing the institutions onto the one legal basis that corresponds to what a measure is predominantly about, is quietly suspended. In other words, the centre of gravity test is honoured in form and displaced in operation.
The second pattern concerns institutional interaction. The choice of legal basis has partially become an instrument of institutional preference. The Commission frames proposals, and the Council adopts measures, with an eye to the procedure that a given legal basis carries. Article 122 TFEU excludes the Parliament from co-decision, while the accumulated combinations noted earlier are assembled precisely so that the ordinary legislative procedure applies throughout, and no procedural objection can arise.Footnote 151 The Parliament’s attempt at a obtaining a remedy, where the choice operates to its detriment like Article 122 TFEU does, is litigation, and Parliament v Council (SAFE) is the current expression of that remedy.Footnote 152 Section VII demonstrated the clearest exemplar of the pattern. A single substantive policy objective, defence-industrial support, has been pursued through the EDIRPA on Article 173(3) TFEU, the ASAP on Articles 114 and 173(3) TFEU, the EDIP Regulation on four legal bases, and the SAFE Regulation on Article 122 TFEU, with the Parliament a co-legislator under the first three, and excluded under the fourth.Footnote 153 The same policy is thereby processed, constitutionally, as four different kinds of EU legal act, and the variable that determines the Parliament’s position is the choice of legal basis alone. There is a certain symmetry in this. The doctrine itself was forged in inter-institutional litigation, from Generalised Tariff Preferences through Chernobyl to Titanium Dioxide, in which the Court of Justice recognised the Parliament’s standing to defend its prerogatives precisely because the choice of legal basis determines them.Footnote 154 What has changed is the scale of what now turns on the choice, and the deliberateness with which the choice is made. The principle of institutional balance in Article 13(2) TEU, which requires each institution to act within the limits of the powers conferred on it, and with due regard for the powers of the others, is the constitutional standard in which this pattern falls to be assessed.Footnote 155
The third pattern concerns the scope and intensity of judicial review. The choice of legal basis is substantive in form, since it directs the Court of Justice to assess the aim and content of the measure, and procedural in consequence, since what follows from the answer is who decides, under which procedure, and by what majority in the Council. The intensity of substantive review should be appropriate to the procedural stakes of the choice under challenge. Where every candidate legal basis carries the ordinary legislative procedure, the choice among them alters little in the institutional settlement, and the permissive review that has in fact characterised the case-law on Article 114 TFEU is a tolerable equilibrium.Footnote 156
Where, by contrast, the choice of legal basis displaces the Parliament, as Article 122 TFEU does, the substantive conditions of the legal basis should be scrutinised in the mode of Tobacco Advertising I,Footnote 157 rigorously and against the text, because the substantive inquiry is at that point the only safeguard the institutional balance has. This possibility is available to the Court of Justice without any alteration of EU primary law. The standards themselves are judicial constructs, as seen by Tobacco Advertising I.Footnote 158 The scope and the object of the Court of Justice’s scrutiny can develop, with further judicial refinement of the centre of gravity and institutional balance doctrines, including resetting the intensity of review. A particular standard of review is a lesser constitutional event than treaty amendment, and enforcing a particular standard of review lies within the Court of Justice’s own hands.
The fourth pattern concerns conferral. The principle in Article 5(2) TEU confines the EU to the competences conferred upon it, and the choice of legal basis is the point at which that confinement is given effect in each individual EU legal act. The question whether the EU may act at all is, in reality, always answered affirmatively because a legal basis is always found, whether through the elasticity of Article 114 TFEU, the exclusivity and breadth of Article 207 TFEU, the exigency of Article 122 TFEU, or even through the flexibility clause of Article 352 TFEU. The operative question becomes how the EU acts, under which procedure, and with which institutions. Conferral is, on this analysis, hollowed, rather than breached as a constraint. The function of conferral, policing the boundary between the EU and its Member States, recedes, and its residual force is exercised horizontally, in the allocation of power among the institutions in the choice of legal basis of EU legal acts. The consequence is that disputes which are, in substance, disputes about the limits of conferred competence, now present themselves as disputes about institutional balance, and the conferral case-law of the coming decades will be found in legal basis case-law.
IX. Conclusion
The choice of legal basis of EU legal acts is, as this article has sought to demonstrate, a question of EU constitutional law of the first order. It determines the speed at which the EU moves, the procedure that governs the adoption of EU legal acts, the voting rule applied within the Council, the institutional position of the Parliament, and the standard against which the resulting EU legal act is reviewed by the Court of Justice. That the choice of legal basis must reflect the aim and content of the EU legal act, that where two or more purposes are present the predominant one must be identified, and that recourse to a dual legal basis is excluded where the procedures attached to the potential legal bases are incompatible, was all settled in Generalised Tariff Preferences, refined in Titanium Dioxide, and given substantive expression in Tobacco Advertising I. This case-law is settled, and theoretically it continues to structure every EU legal act adopted within the EU legal order. What has changed in the past decade, as the preceding sections have demonstrated, is the use to which the provisions of EU primary law are being put. The breadth with which the legal bases are construed is long-standing. What is new is the externally generated impetus of the EU legal acts now adopted upon them.
The first proposition of this article concerned Article 114 TFEU.Footnote 159 The provision on internal market harmonisation has historically been the workhorse of EU secondary law, and it has long been the provision most often tested against the conferral-based limits laid down in Tobacco Advertising I. As demonstrated, Article 114 TFEU is now being deployed as the legal basis for EU legal acts whose centre of gravity lies, on a plausible reading of their recitals and operative provisions, in the management of common external dependencies of the EU Member States, rather than in the removal of obstacles to intra-EU trade or the elimination of appreciable distortions of intra-EU competition. The FSR Regulation reframes state aid by third states as an internal market distortion, although EU Member States had no divergent national law on the matter to be harmonised. The CRMA addresses a common external supply vulnerability through internal market harmonisation, although the divergence said to be harmonised is difficult to identify. The IMERA establishes a standing framework for EU-level emergency management of the internal market, triggered by events that arise from outside the EU. Each of these EU legal acts raises, to varying degrees, the question whether the conditions stated in Tobacco Advertising I, that the EU legal act genuinely respond to obstacles or distortions, have in fact been satisfied. None has been challenged before the Court of Justice, and thus none has been authoritatively assessed against those conditions.
The second proposition of the article concerned Article 207 TFEU.Footnote 160 The CCP, an exclusive competence of the EU,Footnote 161 widened by the Treaty of Lisbon, was for most of its history exercised through WTO-consistent, non-discriminatory legal acts, oriented to the promotion of liberal trade. The foregoing demonstrates that the provision has been reoriented, in a relatively short period, into a platform for defensive geoeconomic response. The ACI and the FDI Screening Regulation, and the recast proposal for the latter, have each, in different ways, mobilised Article 207 TFEU to address the conduct of specific third states, rather than to set the terms on which the EU engages with all third states alike. The legal basis questions raised are far from uniform. The FDI Screening Regulation rests on Article 207(2) TFEU alone, and the recast proposal adds Article 114 TFEU. The ACI is grounded in Article 207 TFEU, and the question is whether the response measures it authorises remain within the scope of the CCP. In other words, the CCP is being asked to do work different in kind from what it was historically asked to do, and the legal basis question naturally follows.
The third proposition of the article concerned Article 122 TFEU.Footnote 162 The earlier discussion demonstrates that the emergency clause, used sparingly before 2020, has migrated into one of the principal channels for major commitments of the EU. The institution that EU primary law positions as co-legislator on most files of consequence in the ordinary course, the Parliament, has been routed around on several of the largest commitments of the period, totalling expenditure in the hundreds of billions of euros. The Parliament’s response, litigation, should produce the first major ruling on the structural limits of Article 122 TFEU.Footnote 163 If the Court of Justice upholds the Council’s use, the provision will be deployed in still further cases. If it annuls the SAFE Regulation, there will be a need to find legal bases commensurate with the extent of the EU’s ambitions.
Given this analysis, numerous further questions follow. The first is whether the case-law of the Court of Justice will, in the years ahead, accommodate such uses of the legal bases concerned, for the EU legal acts already adopted and for those to come, or whether the Court of Justice will police such uses in the vein of Tobacco Advertising I. The second is whether Article 122 TFEU can sustain EU legal acts whose substance is fiscal policy, defence policy, or industrial policy, and whether the frequency of its invocation is itself an argument against the continued characterisation of any given invocation as exceptional. The third is whether the EU’s increasing reliance on dual and multiple legal bases can continue to satisfy the compatibility of procedures test stated in Titanium Dioxide. The fourth is whether the principle of institutional balance in Article 13(2) TEU furnishes a meaningful constraint on the formal exclusion of the Parliament under Article 122 TFEU. The fifth is whether Article 40 TEU, the provision that explicitly provides the Court of Justice with the power to ensure mutual non-encroachment between CFSP matters and non-CFSP matters, will be invoked to discipline the use of non-CFSP legal bases for substantively defence-oriented EU legal acts. The Court of Justice has, in recent years, equipped itself with a richer toolkit for addressing such questions than was available even a decade ago, including a sharper conception of the EU’s constitutional identity,Footnote 164 and a willingness to give the values of Article 2 TEU operational legal effect.Footnote 165
The wider question, which underlies all of this, is whether the body of EU legal acts that has emerged can be sustained on its current legal foundations. There are two paths. The first is continued accretion, whereby the institutions continue to rely on legal bases in the EU Treaties for acts that are functionally external in orientation, litigation over legal basis persists at the margins, and the Court of Justice further refines the centre of gravity and institutional balance doctrines. The second is more daring, namely, treaty revision, necessitating a formal alignment of EU primary law with the new competences and suitable legal bases that have, in fact, developed through the EU legal acts of the past decade. The arguments for revision, legal certainty, democratic legitimacy, a cleaner division of competence, are familiar. The political obstacles are equally familiar. In the absence of revision, however, the Court of Justice will continue to carry a substantial share of the constitutional adjustment, and the choice of legal basis of EU legal acts will remain the principal vehicle through which that adjustment is carried out. Above all, the legal basis of EU legal acts now requires sustained constitutional attention.