On 28 July 2026, before the joint Foreign Affairs and Defence Committees of the Chamber of Deputies and the Senate, Italian Foreign Minister, Antonio Tajani, announced that the government had “decided to use SAFE, requesting an intervention of 14.9 billion euros by the end of the year”. He corrected himself within hours: the 14.9 billion had merely been “booked”, representing a ceiling, and Italy would probably draw six, seven, eight or nine billion euros. Defence Minister, Guido Crosetto, added that the SAFE (Security Action for Europe) decision is a “technical” one, an alternative to issuing treasury bonds, and that the genuinely political choice is the budgetary deviation that the Houses of Parliament will be asked to authorise. A senator from the governing Legue party, then, recalled that it is up to the Legislature to decide whether, and when, to use those funds, while that “booking” is merely an expression of interest.

This episode sheds light on an underlying structural issue: the European Union (EU) built SAFE, its fastest defence-financing instrument, on a legal basis (Article 122 TFEU) that bypassed the European Parliament (EP). However, SAFE is now delayed by precisely the kind of domestic parliamentary budget procedures that its legal basis allowed the Union to avoid in Brussels.

Budgetary sovereignty, we argue, is not a residue of national sovereignty to be overcome on the path toward the common defence. It is the key variable that determines whether common defence is financially feasible at all and above all it is the condition of its democratic legitimacy. Where a Member State (MS) lacks suitable procedures to support a multiannual commitment, funds may formally be allocated yet never effectively committed or disbursed.

Therefore, alongside the an – whether a state may resort to armed force – and the quomodo – how it may do so, constitutional scholarship should attend to quibus opibus bellum alere, that is, with what resources military action is to be financed, and who holds the authority to decide about it.

SAFE Asks Member States To Do Something Their Budgets May Not Allow

Regulation (EU) 2025/1106 (the SAFE Regulation, hereinafter the Regulation), adopted within the ReArm Europe Plan/Readiness 2030, makes up to 150 billion euros in loans available to MSs for joint procurement and investment in the European defence industrial base. Procedurally the scheme mirrors the Recovery and Resilience Facility (RRF): MSs submit national investment plans, the Commission assesses them, and the Council adopts implementing decisions. The two instruments diverge, though, in legal basis and institutional design: RRF rests on Article 175(3) TFEU; SAFE on Article 122 TFEU, which empowers the Council to adopt emergency economic measures on a Commission proposal, but without the involvement of the EP.

The EP, albeit supporting the substance, challenged the legal basis of SAFE. On 20 August 2025 it brought an action for annulment before the Court of Justice (Case C-560/25, Parliament v Council), arguing that SAFE does not meet the exceptional circumstances that Article 122 TFEU requires, and asking the Court to annul the Regulation while maintaining its effects until a substitute measure enters into force. Commissioner Kubilius has reassured capitals that a negative ruling would not automatically suspend the scheme. The reassurance is telling: everyone expects the instrument to outlive the issue of its own legality.

Two aspects of SAFE are relevant. First, it provides loans rather than grants, on favourable terms and with a long maturity, so the amounts count as national public debt and remain subject to domestic fiscal rules. In this sense, Minister Crosetto’s description of SAFE as an alternative to issuing Italian government bonds is technically correct if one looks solely at the funding structure. It does not extend to the governance dimension, since SAFE loans are embedded in an EU‑level framework with its own timetable, conditionalities and political constraints, which domestic issuance does not replicate. Second, SAFE rests on the assumption that MSs possess the fiscal and institutional capacity to commit to a multiannual procurement path. In other words, it assumes a domestic financing architecture that falls outside the mechanism the Regulation itself establishes.

Italy’s Annual Budget Against SAFE’s Multiannual Logic

Nineteen MSs submitted national defence investment plans by 30 November 2025, Italy among them, with an allocation of 14.9 billion euros confirmed in the Commission’s January 2026 proposal. However, by the end of May 2026 only five MSs had actually signed loan agreements with the Commission: Belgium, Croatia, Lithuania, Poland and Romania. Poland, the largest beneficiary with 43.7 billion euros, had by then also become the first MS to receive money, drawing 6.6 billion in pre-financing. Italy has still not signed, and two months later its Foreign Minister was still describing the allocation as a mere reservation and its Defence Minister was still deferring the decision to the end of the year.

Brussels has made it clear that this timetable does not work. A Commission spokesperson observed that waiting until December for the Italian signature “is not a suitable timing”, because if Italy were to accept less than the 14.9 billion allocated, the Regulation sets a legal deadline requiring the Commission to redraw the plan and reallocate the funds within the year. The European instrument thus follows a timetable that does not align with the logic of the national budget cycle.

Nor is the parliamentary passage straightforward on either side. Within the governing coalition, the Ministry of Defence has pressed for the loans, whereas the Ministry of the Economy has resisted them. On the opposition side, parties are equally divided: the Democratic Party supports SAFE considering it as the instrument closest to genuine European defence coordination; by contrast, the Five Star Movement and the Greens and Left Alliance oppose defence expenditure as such.

Unlike other systems, Italy finances defence through the ordinary annual budget, within a politically fragmented parliamentary system and under EU fiscal constraints made particularly stringent by a high debt-to-GDP ratio. The Documento programmatico pluriennale per la difesa provides a planning framework, but it does not bind future financial years. According to NATO estimates, Italian defence spending stood at 1.49 per cent of GDP in 2024; meeting the agreed targets would require an additional 10-11 billion euros annually, to be sustained across successive parliamentary terms, with no legal mechanism shielding them from renegotiation in each annual budget.

This is why, in Italy, the SAFE issue shifts from defence policy to fiscal governance, and the crucial political moment is the parliamentary vote authorising the budgetary deviation rather than the submission of the loan request. From this perspective, the insistence on the prerogatives of the Houses is a restatement of ordinary constitutional arrangements rather than a tactical position: Article 81 of the Italian Constitution routes every euro through parliamentary authorisation, and no European allocation displaces that requirement. What SAFE suggests is that a state lacking a dedicated multiannual financing instrument cannot translate a European allocation into a national commitment at the pace expected by the EU. Instead, it must do so one budget cycle at a time.

Six Models, Three Configurations, One Recurring Tension

The comparison we develop in a forthcoming study covers six systems and yields six distinct models of defence financing, arranged along a continuum of decreasing legal and temporal rigidity.

France sits at the rigid end. Its loi de programmation militaire sets a 413.3 billion euros spending trajectory for 2024-2030 in an ordinary statute that binds no subsequent finance act but makes departure politically costly. Moreover, a centralised procurement agency gives Brussels a single interlocutor. Germany moved the issue into the Constitution, first exempting the 100 billion euros Sondervermögen from the debt brake in 2022, and then, in March 2025, removing defence spending above 1 per cent of GDP from the brake. On both occasions, the amendment required the highest parliamentary consensus, a two-thirds majority, and paired the new fiscal space with dedicated Bundestag oversight. Poland reached the highest spending level within the Alliance, above 4 per cent of GDP, through the ordinary budget and a statutory support fund administered by the state development bank, without any constitutional amendment. Denmark secures a ten-year planning horizon through the Forsvarsforlig, a flexible and renegotiable cross-party agreement. Italy and Austria close the series relying on the ordinary annual budget; Austria holding expenditure structurally low under constitutional neutrality.

These six models can be further reduced to three configurations.

The first group is characterised by stable multiannual programming, as in France and Denmark, which replicate domestically the long-horizon logic assumed by SAFE and can engage with it without institutional friction.

The second group includes systems that have passed constitutional amendments for defence, with Germany as the only complete example, where fiscal space was obtained at the highest parliamentary consensus required for constitutional amendment.

The third group encompasses the systems lacking both dedicated programming and constitutional amendments. In these cases the speed of adjustment depends not on legal instruments but on contingent political factors: threat perception in Poland, coalition cohesion in Italy, the constitutional significance of neutrality in Austria.

Poland is the case that challenges the obvious inference. The highest level of defence spending in the Alliance was reached without constitutional amendments, but on the basis of an ordinary statute, the 2022 Homeland Defence Act, a fund of statutory rank, and a cross-party consensus driven by the perception of the Russian threat. Therefore, this case suggests that dedicated financial instruments do not produce speed. What they produce is predictability of the trajectory and stability of multiannual commitments within a coherent framework of decision and oversight. The financial variable bears less on the intensity of adjustment than on its sustainability over time and its legitimation.

The Same Friction, Twice

Strategic urgency acts as a vector of power concentration: invoked as a condition of decision-making efficiency, it shifts prerogatives towards the executive and compresses those of representative institutions. The pattern recurs at both levels of the European legal space: within the MSs in the relationship between executives and legislatures over the allocation of resources, and within the Union in the choice of Article 122 TFEU and in the resulting litigation before the Court of Justice.

None of the examined systems removes defence financing from representative bodies. Germany acted by constitutional amendment, through the highest parliamentary consensus that the procedure demands. France legislates its trajectory. Denmark negotiates it across political parties. The Italian Houses of Parliament vote on it annually. Ultimately, even where executives enjoy the widest margins, military expenditure remains tied to parliamentary approval as a condition of its validity. Our claim is that there is an irreducible core of parliamentary legitimation of defence spending, and that no model of common defence built for maximum decisional speed can render it irrelevant.

The EU has not yet found the equilibrium point. The recourse to Article 122, the EP’s action for annulment, and the continued reliance on off-budget instruments such as the European Peace Facility in supporting Ukraine all reproduce, at the Union level, the friction between executive responsiveness and representative legitimation that is replicated across the MSs. Italy’s unsigned loan agreement is an example of what this friction looks like when it reaches the final stage.

Coordination Rather Than Uniformity

None of these models can be transplanted wholesale since each reflects its own legal culture and constitutional framework. The French programming statute presupposes a semi-presidential system built around executive primacy. The German special fund presupposes a constitutional culture in which amending the Basic Law is a genuinely available instrument of adaptation. The Danish agreements presuppose a consensual tradition that fragmented parliamentary systems cannot manufacture on demand.

Therefore, if a common European defence is to emerge, it will look like a mosaic rather than a uniform institutional form. What is needed is not the harmonisation of national financing regimes but their coordination around common objectives, enabling each MS to undertake commitments that are predictable, verifiable and compatible with its constitutional architecture. The central issue is how to reconcile strategic responsiveness, parliamentary accountability and constitutional pluralism. In its current configuration, SAFE addresses only the first of these concerns.

Seen from this angle, the 14.9 billion euros allocated to Italy and not yet drawn down should not be read as evidence of a country failing to keep up. The Italian experience exemplifies a broader constitutional point: budgetary sovereignty is not an impediment to European defence integration; conversely, it is one of the foundations of its legitimacy. Recognising it as such, rather than treating it as a residual obstacle to be bypassed, is essential for a defence union that is not only operationally effective but also constitutionally sustainable.