Why the German Federal Budget Should Be Reformed – and How

1. Introduction

For years, the German Federal Audit Office (2023) has been pointing out that only around ten percent of the federal budget is available for discretionary spending – meaning that the government can only actually decide on around ten percent of annual expenditure. The rest is already committed due to existing legal obligations and personnel costs. An overly tight budget not only jeopardises the federal government’s ability to act, but also democracy: after all, political decisions made by voters should have fiscal implications. Otherwise, democracy is hollowed out and dematerialised.

In light of the special fiscal circumstances of recent years – in particular the exceptions for defence, infrastructure and climate protection – we have examined how the federal budget could develop until 2035. To this end, in this paper, we propose a new measure for the discretionary share of the budget, which we dub Fiscal Space (FS) Indicator.

The FS indicator shows: By 2035, there could be no discretionary fiscal space left in the regular federal budget. The main reasons for this are rising social and labour market expenditure on the one hand, and sharply rising interest payments on the other, as the federal budget bears the entire financing costs for both the special fund and the defence exemption.

Germany is thus approaching a fiscal situation akin to the United States: every few years, a new special fund would have to be approved – comparable to the regular raising of the debt ceiling in the US – just to maintain minimal political room for manoeuvre. Conducting fiscal policy in this way is neither sustainable nor properly democratically legitimated.

How can such a fiscal impasse be prevented? Cosmetic cuts to individual programmes are not enough. The budget needs to be restructured with a razor-sharp focus on one goal: creating unsubsidised employment. We explain below why this is the single most important goal for getting to government finances which are sustainable in the long term.

Around half of the federal budget goes to transfers to private households and firms: social benefits, subsidies, or energy price relief, summarised below under the term survival subsidies. In most cases, income subsidies are only the second-best solution. They also crowd out spending on public services that could reduce future subsidy needs (e.g., spending on education and training).

The future viability of the federal budget depends largely on whether it is possible to strengthen economic growth, tap the unused potential in the labour market, and promote sustainable business models. This is the only way to reduce the need for subsidies in the long term – and thus also ease the burden on the budget.

Our forecasts show that without reforms, the need for subsidies will continue to rise. We are therefore proposing measures to limit this and regain room for manoeuvre in the budget. This path may at first look more arduous than across-the-board cuts, but, given the magnitude of the challenge, it is the only feasible and sustainable one. The alternative is short-sighted: Sizeable across-the-board cuts invariably involve cutting investment, typically one of the few discretionary elements of the budget. Slashing investment depresses future growth, thus reducing future tax revenue, which is likely to cause budgetary issues in the future. Thus, in doubt, across-the-board cuts do not solve budget problems but kick the can down the road instead.

However, one conclusion is hard to evade when running the numbers: Even if subsidies are successfully reduced, even if the government manages to be as effective as possible with spurring growth, the scope available under the current debt brake is not sufficient in the medium term. The plasticity of the budget is by far not as large as assumed by many, sustainable reforms do take time. Therefore, a budget reform must be accompanied by a reform of the debt brake so that it once again enables sustainable, democratic but also properly rule-bound fiscal policy.

2. Projection: Discretionary fiscal space is dwindling

What is discretionary fiscal space?

A key issue in fiscal policy is to determine the level of discretionary expenditure. Discretionary expenditure refers to expenditure that policy makers can effectively decide on each year. A large part of the federal budget is not subject to such discretion, e.g., because it is needed to pay for legal entitlements such as pension payments or salaries. In recent years, based on estimates by the Federal Audit Office, it was assumed that approximately ten percent of the budget was discretionary. With a budget volume of around 500 billion euros, the room for manoeuvre thus amounts to approximately 50 billion euros, an average of just under three billion per federal ministry.

In view of the changed circumstances – the coronavirus crisis, the energy crisis, Russia’s attack on Ukraine, the election of Trump – and the changed fiscal framework – in particular the extensive exemption of defence spending from the debt brake and the special fund for infrastructure and climate protection – we have reassessed the level of discretionary fiscal space, examined its historical development, and projected its future evolution. We propose a new fiscal space indicator as a measure of the political and budgetary leeway a government has at its disposal.

The concept of discretionary expenditure is intuitively easy to understand but difficult to define precisely. Following Streeck & Mertens (2010), we consider all expenditure over which the government has no discretion for legal or other reasons as principally non-discretionary. However, governments with a sufficiently long time horizon can, of course, decide on many things, such as the level of pension subsidies. We therefore refine the definition by considering only those expenditures as discretionary that are decided upon with a maximum of one year’s notice. This means, for example, that housing benefits, which must generally be reapplied for each year, are discretionary expenditures, but student loans, which are granted for the entire period of study, are not. As in Streeck & Mertens (2010), German defence spending is not considered discretionary, as it is largely determined by NATO plans.

Refining the approach of Streeck & Mertens (2010), we classify expenditure at the functional level. These functions include, for example, spending on “general public services and administration”, “defence”, or “childhood support”. In addition, we account for future commitments in expenditure categories that count as discretionary otherwise. Even though road investment is generally classified as discretionary, for instance, commitments towards funding a specific, already contracted road project are not.

As we are trying to understand how much of the budget is available for discretionary expenditures under the debt brake, we exclude expenditure exempted from the debt brake. This includes emergency loans granted in recent years and all defence expenditure exceeding one percent of GDP, reflecting the revised debt brake’s provisions. We also exclude special funds, both those exempted from the debt brake (Bundeswehr, infrastructure and climate protection), as they only provide additional funds on a temporary basis, and those that are not (e. g. the climate and transformation fund), in order to focus on discretionary fiscal space within the regular budget framework.

The crux of the federal budget: interest payments and survival subsidies

Figure 1 shows how the discretionary budget share has developed since 1952. Over the last 30 years, it has fallen from above 40 percent to around 16 percent today. The two biggest drivers behind this development are, on the one hand, interest expenditure, which rose sharply from the 1990s until the financial crisis and again since 2022. On the other hand, the share of social transfers has increased significantly, particularly those related to social security and unemployment benefits.

We refer to some social transfers in the federal budget as survival subsidies. These transfers are used by the state to subsidise people who do not earn enough money to finance their own living. There may be good reasons for the state to subsidise income – no one wants poverty or unstable pension and health insurance – but this is only the second-best solution. The ideal situation would be for the economy to work in such a way that everyone could live on their own income.

Survival subsidies in Figure 1 include federal social security contributions, unemployment benefits (now Bürgergeld and accommodation costs), student loans (Bafög), social compensation and benefits for survivors and victims of political events, as well as social aid and protection (including benefits for poor pensioners and asylum seekers). Survival subsidies reflect the fundamental problem that many people in Germany cannot get along without state support, thus tying up the federal budget for more than a year.

Our analysis is limited to interest and social transfers, but companies also receive a variety of transfers from the federal budget that could be classified as survival subsidies – primarily to ease the burden of energy prices. In the 2025 federal budget, subsidies for electricity and gas total just under 25 billion euros, and the trend is rising. More and more business models are only viable if they are subsidised by the state. There may be good reasons for energy price relief, just as there are for providing assistance to households. However, the aim of fiscal policy should be to enable as many business models as possible to be self-sustaining.

Below, we estimate how interest expenditure, survival subsidies and, with them, discretionary fiscal space could develop over the next ten years.

Projection of the federal budget: the FS indicator

Our projection of the total budget volume is based on expected federal revenue. Forecasts for tax revenue until 2029 are taken from the government’s medium-term financial plan and projected based on its trend during that period. Other revenue is assumed to be constant and at the average level of the past ten years. Lastly, the admissible amount of borrowing given the current fiscal framework – consisting of the constitutional debt brake and the European fiscal rules – is added (and assumed to be fully exhausted).

On the expenditure side, most non-discretionary spending items are projected based on their historical ten-year trend. That applies to expenditure on general public services and administration, public order and safety, student support, childhood benefits, social compensation, social aid, nuclear security, and legal commitments in otherwise discretionary spending categories made in the past. We calculate unemployment benefits using the federal government’s unemployment forecast and the historical elasticity of this expenditure with respect to the unemployment rate. Defence spending is truncated at one percent of GDP, according to the recent amendment of the constitutional debt brake.

For the amount of federal subsidies for the social security system, we only account for transfers reflecting the current political consensus within the government. Regarding the pension system, we build upon the official projections of federal subsidies, as those have been very accurate in recent years, and consider the costs of an improved recognition of child-raising periods (“Mütterrente III”), which is planned for 2027. However, although healthcare, long-term care, or unemployment insurance are in need for more financial support, we do not assume further federal subsidies in these categories, as they have not yet been politically agreed upon. Instead, the government has just decided to stabilize these pillars of social security by means of bridging loans, which do not weigh on the regular budget as they are exempted from the debt brake. For the years after 2029, we forecast social security spending based on its medium-term trend. Our approach is conservative in the sense that more stabilization measures are looming, which could well reduce discretionary spending even further.

Ultimately, we need to project interest expenditure. We assume German government bond yields to increase gradually to a level of 2.7 percent in 2030 and beyond. As we understand it, both the interest payments for all expenditures from the special fund and for debts incurred for defence are covered by the federal budget. Even with moderate interest rates, this leads to a significant increase in interest expenditure. While necessary investments are made possible by ad hoc rules, the federal budget is becoming increasingly fragmented.

Using this projection, we construct the FS indicator – our forward-looking estimate of fiscal space (see Figure 2).^[1] It measures the share of discretionary federal spending and how it will evolve in the ten years ahead. The indicator consists of three metrics: discretionary fiscal space in the current year (FS0), at the end of the four-year financial planning period (FS1), and after ten years, describing the evolution of fiscal space if current trends were to persist (FS2). Moreover, we specify a range for FS1 and FS2 by considering two scenarios of economic growth being 0.5 percentage points above or below official forecasts in each year of the projection period.

Our FS indicator points to a worrisome development: the government’s fiscal space is waning. Currently, discretionary fiscal space is at 16 percent (FS0), roughly as much as the federal government spends on defence in 2025 (80 billion euros). Despite our rather conservative projection, in 2029, only five percent of expenditure (just under 25 billion euros) could be available (FS1). In 2035, discretionary fiscal space could be zero (FS2). It would then be impossible for any government to pursue a proactive policy. Democratic elections would be rendered fiscally and politically meaningless.

A government that aspires to shape the future by its policies should therefore start reforming the budget as soon as possible. This demand is old hat in its generality. In the following, we outline an analytical lens with which this can actually be achieved and estimate the additional fiscal space that could be created in the coming years with appropriate policies.

3. Breaking the trend: how to bring the federal budget back on track

When the budget becomes tight, German fiscal policy usually responds following a (small) chainsaw strategy, i.e. across-the-board cuts in all available expenditure (and, if that is not enough, VAT increases). Each ministry is then given savings targets to meet. It does not matter whether cuts make economic sense or not. This usually comes at the expense of investment and the financing of public services, as those expenditures tend to be amongst the few that can be cut easily. A look at the budgets of recent years shows that investment programmes – especially for daycare centres and all-day schools – come and go depending on the budget situation. If the need for subsidies continues to rise in the coming years, funds for many public services, such as spending on daycare centres, schools or research, may be short. This weighs on the growth potential of the German economy. Lower growth, in turn, increases the need for subsidies and reduces discretionary fiscal space – a self-reinforcing negative spiral.

The key to sustainable federal finances is higher economic growth. It addresses the two main drivers behind the declining discretionary share. Firstly, more growth ensures rising tax revenues, which are necessary to finance interest expenditure. In order for the interest costs associated with the massive increase in borrowing from 2025 onwards through special funds and the defence exemption to be sustainable, the additional funds must be spent in a targeted manner to boost growth.

Secondly, in order to reduce the need for survival subsidies, incomes must rise so that fewer people are dependent on government assistance. Half of all working women in Germany work part-time, a significant proportion of them not voluntarily, but because of a lack of full-time jobs, childcare facilities, or care infrastructure. Four million people receiving unemployment benefits have no job or work and earn so little that their income has to be supplemented by the state. This has consequences: those who work little or not at all receive low incomes, pay hardly any social security contributions, and have problems covering their living expenses – the federal government then has to step in. It also pays the costs when employees take early retirement without any reductions because it is financially advantageous for them.

To change this, a policy is needed that taps the unused potential in the labour market. We analyse how such a policy would affect the share of survival subsidies in the federal budget and, consequently, fiscal space. To this end, we assume that, by 2035, it will be possible to (a) move one million women from part-time to full-time employment, (b) putting one million unemployed to work, and (c) encourage one million pensioners to delay their retirement. This will result in more hours worked per year, which will both boost economic growth and reduce the need for subsidies.

To estimate the effect of the higher volume of work on social benefits, we use the elasticity between the two variables: on average, federal social spending has fallen by 0.9 percent in the past when the volume of work has increased by one percent. This figure is in line with the elasticities between government spending and GDP estimated by the European Commission (Federal Ministry of Finance 2019; Mourre et al. 2019).^[2]

The result of our analysis: a policy that consistently addresses the causes of survival subsidies can significantly reduce their share of the federal budget. The blue hatched area in the first panel of Figure 3 shows the additional discretionary fiscal space that such a policy could gain, significantly driven by all three policy measures (see the second panel). This would halt the trend towards ever-decreasing room for manoeuvre in the federal budget. The FS2 indicator – discretionary fiscal space available in 2035 – would still amount to 14 percent – instead of disappearing altogether.

However, we should not draw the wrong conclusions: a policy aimed at increasing the volume of work and raising incomes helps the federal budget, but even a discretionary expenditure share of 14 percent is low. The scope for future democratically elected governments to implement their programmes would remain severely limited.

Thus, more than just growth-oriented policies are needed: the federal budget should be thoroughly analysed and restructured, and the debt brake should be reformed in order to create more discretionary fiscal space in the long term. We set out concrete proposals on how this can be achieved below.

4. Three proposals for a sustainable federal budget that adds up

Based on our analysis, we derive three recommendations for a strategic realignment of the federal budget:

References

• Bach, S. / Buslei, H. / Geyer, J. / Haan, P. / Pieper, J. (2025): “Aktivrente entlastet vor allem besserverdienende Rentner*innen – mit unsicheren Beschäftigungseffekten”, Deutsches Institut für Wirtschaftsforschung, 92, S. S. 395402
• Bundesministerium der Finanzen (2019): “Analysen und Berichte: Aufteilung der 2018 neu berechneten Budgetsemielastizität auf Bund, Länder, Gemeinden und Sozialversicherungen”, Monatsbericht April 2019
• Bundesministerium für Arbeit und Soziales (2023): “Zur Reform der Transferentzugsraten und Verbesserung der Erwerbsanreize”, Webseite des Bundesministerium für Arbeit und Soziales
• Bundesrechnungshof (2023): “Bemerkungen 2023 zur Haushalts- und Wirtschaftsführung des Bundes (einschließlich der Feststellungen zur Haushaltsrechnung und zur Vermögensrechnung 2022)”,
• Deutsche Bundesbank (2025): “Früher, regulär, später: Wann Versicherte in Rente gehen und wie Ab- und Zuschläge ausgestaltet werden könnten”, Deutsche Bundesbank
• Mourre, G. / Poissonnier, A. / Lausegger, M. (2019): “The semi-elasticities underlying the cyclically-adjusted budget balance: an update & further analysis”, Europäische Kommission
• Schuster-Johnson, F. / Sigl-Glöckner, P. (2025): “Eine Agenda für die Expertenkommission zur Schuldenbremse. Von Regelchaos zum Anker für nachhaltige Staatsfinanzen in Europa”, Dezernat Zukunft,
• Streeck, W. / Mertens, D. (2010): “Politik im Defizit. Austerität als fiskalpolitisches Regime”, dms – der moderne staat – Zeitschrift für Public Policy, Recht und Management