On 21 March 2025, Germany amended its constitutional debt brake, which had limited its structural (cyclically adjusted) federal budget deficit to 0.35 percent of GDP and required German states to run structurally balanced budgets. Now, defence and security related spending above 1 percent of GDP is freed from the 0.35 percent limit. The amendment also created a €500 billion extrabudgetary fund for ‘additional’ infrastructure investment, and German states can now run maximum deficits of 0.35 percent of GDP ¹ Additionality under the amendment requires that investment in the federal budget must account for at least 10 percent of the budget (slightly less than its 13 percent share of the federal budget in 2024). .

From the perspective of rearmament and closing of Germany’s infrastructure gaps, the amendment is very good news. But it also creates a conflict with the European Union’s fiscal rules, assuming Germany’s extra headroom is used as intended. A satisfactory solution requires a reform of the reform rather than fudges.

Germany’s spending room under EU fiscal rules 

Following the constitutional amendment, Germany’s federal structural balance (including the use of extrabudgetary funds) must observe:

b_Fed ≥ −0.35 − (s − 1) −  f = 0.65 −  (s + f)                                              (1)

where f denotes annual spending from the extrabudgetary infrastructure fund, and s spending on security – including from an extrabudgetary fund for defence created in 2022. Security spending s is defined as defence spending plus “federal spending on civil protection, the intelligence services, cybersecurity and support for countries attacked in violation of international law.” 

Hence, under the amended debt brake, the balance of revenues and budgetary non-security spending – including interest spending – cannot be lower than 0.65 percent of GDP. But the debt brake does not restrict security spending or infrastructure fund spending. 

As pointed out by Zettelmeyer (2025), full use of the borrowing space under the new debt brake would conflict with EU fiscal rules in a very fundamental sense. For plausible long-run nominal growth rates and security spending – say, 2.5 to 3.5 percent of GDP – German debt would rise from its end-2024 level of about 63 percent of GDP to a long-run level of 100 percent or more (Table 1). This runs counter to the EU Treaty requirement for debt above 60 percent of GDP to eventually decline to 60 percent. 

Table 1: Germany, long-run debt level as a share of GDP if borrowing space under new debt brake is fully utilised

_{Source: Bruegel. Note: the table shows steady-state values of debt-to-GDP D for different combinations of security spending s and nominal growth g, using the formula D = (s + 0.65)/g . See Zettelmeyer (2025).}

However, it is unclear whether Germany would want to utilise its full constitutional borrowing space in the long term. 

In what follows, we focus on a more immediate question: what federal borrowing space would Germany have under EU fiscal rules, assuming that its national fiscal rule is binding (implying that equation (1) holds with equality) and that Germany avails itself of the national escape clause (NEC), as offered by the European Commission (2025) ² EU Regulation 2024/1263, Article 26. ? This would allow EU countries to exempt defence expenditure above 2021 levels from the permissible net expenditure path up to a maximum of 1.5 percent of GDP per year, for up to four years. 

Under the new rules, Germany will soon need to submit a four year ‘medium-term fiscal structural plan’ consistent with reaching a general government structural primary balance of just over 0.1 percent of GDP by 2031 if it requests and is granted a seven-year adjustment period, and of about 0.4 percent of GDP by 2028 if it chooses a four-year period (Darvas et al, 2024). To maximise its fiscal space, Germany will likely request the seven-year period. This means that it would be required to gradually raise its structural primary balance from -0.3 percent in 2024 in steps of about 0.4/7 = 0.06 per year. 

To translate this into a constraint on the overall structural balance of the federal government, b_Fed  , it must be borne in mind that in Germany, the general government balance also includes states (b_States), municipalities (b_Mun) and social security (b_Ssec), as well as general government interest expenditures (int). Hence, for any year between 2025 and 2028, before the application of the NEC, the federal structural balance must satisfy: 

b_Fed ≥ SPB* − int − b_States − b_Mun − b_Ssec                                                          (2) 

where SPB* is the structural primary balance floor prescribed by the EU fiscal rules. Activation of the NEC would lower this floor by min{1.5, d – d₂₀₂₁} , where d denotes defence spending according to the ‘classification of the functions of government’ (COFOG) definition. Based on provisional Eurostat data ³ See [gov_10a_exp] General government expenditure by function (COFOG) , d₂₀₂₁ = 1 percent of GDP. 

Given that d is expected to be at most 1.5 percent of GDP in 2025, we focus on the case d ≤ 2.5 and min {1.5, d – 1} = d – 1  ⁴ The value of d reported for Germany was 1.1 percent in 2023. In the same year, the German defence budget (including spending from the 2022 extrabudgetary defence fund) was 1.36 percent of GDP. Defence spending planned for 2025 (including the extrabudgetary fund) is €75.21 billion, 1.72 percent of GDP (using 2025 GDP of €4377.1 billion, see ). .  This implies that after NEC activation, the federal structural balance must satisfy:

b_Fed = SPB* − (d − 1) − int − b_States − b_Mun − b_Ssec                                   (3) 

Assuming that the state-level debt brake binds (d) and condition (1) holds with equality, one obtains:

b_Fed = SPB* − d + 1.35 − int − b_Mun − b_Ssec = 0.65 − (s + f)

or equivalently:

(s − d) + f = −SPB* + int + b_Mun + b_Ssec − 0.7                                          (4)

Equation (4) defines a constraint for the sum of spending f from the extrabudgetary fund and the excess of security spending in the definition of the German budget over defence spending in the COFOG definition, (s – d). Table 2 projects the terms on the right-hand side of (4) for 2025-2028, as follows:

• SPB*, the presumed reference path for the general government structural primary balance in the case of a seven-year adjustment period, rises from -0.3 in 2024 to -0.06 in 2028, in steps of 0.06 (as discussed above);

• int are projections taken from the April 2025 IMF World Economic Outlook (WEO) ⁵ Computed as the difference between primary net lending/borrowing and net lending/borrowing.

• b_Ssec is set to –0.25 percent of GDP, its 2024 value according to Bundesbank data. According to Büttner and Werding (2021), maintaining the current level of pension claims (as confirmed under the incoming government parties’ coalition agreement) will require additional federal grants of at least 0.2 percent between 2024 and 2028. This corresponds to the current social security deficit. Hence, any reduction of the 2024 social security deficit would not create fiscal space for s + f, as it would be offset by a correspondingly higher non-security federal spending need.  

• Figure 1 shows historical data for b_Mun since 1991, together with IMF output gap estimates. With the notable exception of the 2020 COVID-19 recession, when municipalities received massive federal support, b_Mun is quite cyclical. We use this fact to project b_Mun up to 2028 using WEO forecasts for the output gap and general government revenue (dark blue dotted line in Figure 1 and row 4 in Table 1). Municipal deficits decline but take time to disappear, in line with assessments that the medium-term fiscal outlook for municipalities in Germany remains weak (KfW, 2024). 

• To obtain an upper bound for the available space for s – d + f , we make the optimistic assumption that municipal deficits disappear immediately, perhaps because states use their additional borrowing space under the amended debt brake to bail out municipalities (row ‘b_Mun �DZ�پ��������پ���’).&�Բ�����;&�Բ�����;

Table 2: Projected space for German security spending and spending from the infrastructure fund under EU fiscal rules, 2025-2028 (% of GDP)

^{Source: Bruegel based on Spring 2025 IMF World Economic Outlook (WEO) projections and Darvas et al (2024). Note: row (6) computes s - d + f using equation (4) with b_Mun  = b_Mun (baseline), while row (7) computes s - d + f with b_Mun  =  b_Mun (optimistic).}

Rows (6) and (7) show that the maximum that EU fiscal rules allow Germany to spend from the infrastructure fund and on security, in excess of COFOG defence spending, s – d +f , is substantially below what it currently spends in excess security spending: 2025 s – d is budgeted at 0.57 percent of GDP of which 0.16 percent corresponds to aid to Ukraine ⁶ According to Henze (2025), total security spending including aid to Ukraine and federal spending on civil protection, intelligence services and cybersecurity amounted to €88.7 billion, or 2.03 percent of GDP, 0.57 percent of GDP higher than our estimate of d of 1.46 percent for 2025 (see footnote 3). Aid to Ukraine amounts to €7 billion, or 0.16 percent of GDP . In the baseline, 2025 spending room is even negative (-0.21), which implies that to meet the EU fiscal constraint. Germany would have to run a non-security primary surplus that is at least 0.21 percent higher than it is allowed under the reformed debt brake. 

This has two implications. First, unless Germany cuts most of s – d and/or runs a higher non-security primary surplus than required under its national fiscal rule, it will not be able to spend anything from its infrastructure fund. Second, even if it does not spend from the infrastructure fund, it would need to reduce s – d . While use of the NEC would give Germany room to raise COFOG-definition defence spending d by more than 1 percent of GDP, Germany could not maintain its current non-COFOG security spending unless it generates larger surpluses in the non-defence budget.

Figure 1: Fiscal balance of municipalities and German output gap, 1991-2028

(% of GDP, values for 2025 to 2028 are projections)

^{Source: Bruegel based on Deutsche Bundesbank, series , and WEO. Note: output gap projections are taken from the Spring 2025 WEO. Fiscal balance is projected using a regression of the municipal fiscal balance on the lagged municipal fiscal balance, change in the output gap, change in general government revenue, and a dummy variable assigning the value 1 to 2020 and 0 elsewhere.}

A dilemma

In its coalition agreement, the incoming German government pledges to spend €150 billion from its infrastructure fund by 2029 at the federal level alone, or about €37.5 billion per year, corresponding to about 0.9 percent of GDP ⁷ See , page 52. . Additional spending from the fund is supposed to benefit states and municipalities. But according to the calculations above, the current EU fiscal rules, even after the activation of the NEC offered by the European Commission, will not allow any of this spending, unless it is offset by fiscal savings elsewhere. EU fiscal rules also constrain the new German golden rule for security spending. While it can benefit from European Commission flexibility for increased core defence spending, it will need to cut additional security spending (or offset it with savings).

This creates a conundrum. From a European perspective, it is good news that Germany has finally thrown off the shackles of its debt brake. But the Commission has just successfully negotiated medium-term structural plans with 22 EU countries, all of which ended up complying with the new rules (Boivin and Darvas, 2025). Half a dozen countries – including France and Italy – have committed to significant budgetary consolidation. In light of this, the Commission cannot make an exception for Germany beyond the exception for defence spending it has already offered ⁸ Pench (2025) argued that limiting additional borrowing allowed under the national escape clause to military spending cannot be justified under the new EU fiscal rules. However, the proposed limitation has not been challenged by any member state. .

Meanwhile, Germany’s new government cannot afford politically to send a medium-term fiscal plan to Brussels – as it must do soon after it takes office – that envisages significantly lower use of the infrastructure fund than it has just promised.

One potential way out might be for Germany to argue that its new willingness to spend will boost growth much beyond what the Commission assumed in June 2024. Another would be for the Commission define the 2021 defence expenditure baseline in nominal values (rather than in percent of GDP). But none of these options are likely to lead to a satisfactory outcome. The constraint on spending from the infrastructure fund might then become less tight, by a few decimals of GDP.  But it still would be very tight. 

Another (cosmetic) fudge might be to use the infrastructure fund – at the federal level or that of the ��ä�Ի��� – to replace expenditure that is currently undertaken within the budget. This would make it look like Germany is using the extra fiscal space provided by the new debt brake, but without raising the general government deficit. 

But this strategy does not look promising either. Whether or not spending from the fund replaces current expenditure for consumption or investment purposes, it would violate the new constitutional rule, which requires only additional investment to be funded (ie in addition to investment of 10 percent of the annual federal budget).

A further potential fudge would be to reclassify some of the defence spending that is currently reported to NATO but falls outside the COFOG definition as defence spending within the meaning of COFOG. But even if this were possible – COFOG standard is a UN-level standard and cannot be changed unilaterally – it would merely narrow the gap between s and d. But Table 2 implies that even if s – d = 0, the room for spending under the infrastructure fund would be much smaller than what the coalition has pledged. 

A better way out

The clean solution would be to reform the EU fiscal rules yet again – for the benefit not just of Germany but of all other EU countries, particularly those that currently plan harsh cuts to public investment (Boivin and Darvas, 2025). For example, such a reform might take the form of excluding EU-approved increases in infrastructure spending from the initial application of the rules, up to some limit. 

An alternative, more radical approach might be to raise the EU Treaty debt reference value from 60 percent to 90 percent of GDP, while eliminating the ‘deficit resilience safeguard’ (created at the insistence of the then German government), which requires a minimum reduction of the structural primary deficit of 0.25 percent of GDP when the structural deficit exceeds 1.5 percent of GDP and the debt ratio exceeds 60 percent of GDP. From a German perspective, the benefit of such a reform would be that it could allow meaningful borrowing under its new constitutional debt rule even beyond the next four years (Zettelmeyer, 2025). Making this change would require unanimity of EU countries. 

The fact that a reform along these lines would happen so soon after the last reform is not pleasant. But the alternatives – placing an economically unjustifiable constraint on German spending, or letting Germany emerge as the only country that is allowed to break the rules – are much worse. And the benefit of a well-executed reform would extend beyond Germany, allowing higher public investment across the EU, while still maintaining debt sustainability.

The authors thank Zsolt Darvas, Lucio Pench, Florian Schuster-Johnson and Stavros Zenios for their comments.

References

Boivin, N. and Z. Darvas (2025) ‘The European Union’s new fiscal framework: a good start, but challenges loom’, Policy Brief 06/2025, Bruegel, available at /policy-brief/european-unions-new-fiscal-framework-good-start-challenges-loom"

Büttner, T. and M. Werding (2021) ‘Optionen zur Stabilisierung der Einnahmesituation der Sozialversicherungen’, WIP-Analyse, December, Wissenschaftliches Institut der PKV: Köln, available at   

Darvas, Z, L. Welslau and J. Zettelmeyer (2024) ‘The implications of the European Union’s new fiscal rules’, Policy Brief 10/24, Bruegel, available at /policy-brief/implications-european-unions-new-fiscal-rules

European Commission (2025) ‘Accommodating increased defence expenditure within the Stability and Growth Pact’, C(2025) 2000 final, available at

Henze T (2025) ‘Ökonomische Restriktionen für die Umsetzung des Finanzpakets’, IW Policy Paper 6/2025, Institut der deutschen Wirtschaft Köln, available at

KfW (2024) Kfw-Kommunalpanel 2024, KfW Bankengruppe, available at 

Pench, L. (2025) ‘Dilemmas for the EU in deficit-financing of defence expenditure and maintenance of fiscal discipline’, Working Paper 03/2025, Bruegel, available at /sites/default/files/2025-04/WP%2003%202025.pdf

Zettelmeyer J (2025) ‘Can Germany afford to take most defence spending out of its debt brake?’ Analysis, 11 March, Bruegel, available at /analysis/can-germany-afford-take-most-defence-spending-out-its-debt-bra