The state is growing. We pay the bill.

Germany ought to ask itself a simple question: how much state do we need, and how much of it actually makes our lives better? Anyone demanding new taxes, additional posts and further rules has to explain what measurable benefit citizens and companies get for them. A good purpose is not yet proof of a good instrument.

This debate is conducted with many figures that are wrong. For this piece I have recalculated four indicators against their sources. One of them I had to correct against my own draft.

Above half: a spending ratio, not a performance record

In 2025 the state spending ratio stood at 50.3 per cent and thus rose above 50 per cent for the first time since the coronavirus years 2020 and 2021.^[1] What is meant is the total spending of the federal government, the Länder, the municipalities and social insurance in relation to gross domestic product^[3] – not the share of people who work for the state, and not the share the state takes from each individual’s income. In euros that is 2,249 billion, 5.1 per cent more than the year before.^[2]

This is the point at which I had to correct my own draft. It relied on a data series obtained via a financial data provider and gave 50.5 per cent for 2025, 49.4 for 2024 and 48.1 for 2023. The official result is 50.3, 49.5 and 48.4 per cent. The deviations are small. But anyone who demands accuracy of politicians has to demand it of themselves first.

Such a ratio is not in itself proof that money is being spent badly. It is a good occasion to reverse the burden of proof. It is not only citizens and companies who have to explain how they are to finance still more. The state has to explain why it cannot manage better with what it already claims.

What the ratio does not say: it does not rise in every year – after the high value of 2020 it fell for three years. It does not measure what the state itself produces. And it cannot be raised by adding in numbers of people: adding railway employees does not turn a spending ratio of over 50 per cent into a ratio of over 60 per cent. Employees and spending measure different things. The largest single item, at around 52 per cent of state spending, is social benefits; their share has hardly changed in ten years.^[2]

More staff is not yet a better state

As at 30 June 2025, according to provisional official results, 5,475,100 people were employed in the public service, around 5.5 million. Against the previous year 95,100 were added, 1.8 per cent more.^[4] Against 2015 it is 829,600 more, an increase of 17.9 per cent.^[5] That is a development across several governments, not a decision of the present federal government.

Anyone criticising this development ought to know who they mean by it. I have recalculated where the increase since 2015 has gone – and half of it is accounted for by four areas that nobody talks about in the bureaucracy debate:

The four areas account for 413,000 of the 829,600 additional employees, that is 49.8 per cent. They are subsets of the total and do not together make up the overall sum. Note: staff in police training in Hesse have been assigned to universities from 2022; the two series are not fully comparable individually, though their sum is.^[5]

Teachers, police officers and childcare workers are no counter-world to the people who produce. They are the part of the state that most citizens actually want. My criticism is aimed at a political reflex: building up staff must not be the standard answer to badly organised processes. Before new administrative posts arise, tasks, responsibilities and processes belong under examination.

The burden does not come only through the tax assessment

Anyone talking about relief while pointing only at the income tax schedule is not describing the whole bill. The national CO₂ price on fuels is the clearest example: it appears on no tax assessment and nevertheless on every fuel bill and every heating bill.

On 12 August 2026 the cabinet adopted an amendment intended to continue the corridor of 55 to 65 euros in 2027 as well. That is not yet an act.^[7] What the CO₂ price means in practice for a single lorry load is set out in the commentary CO₂-Abgaben: Preistreiber mit Ansage. On the planned tax on sugary drinks and the planned plastic packaging tax there are separate reports; both are plans and not taxes already levied in 2026.

Health, climate and environmental goals deserve a serious debate. But even a legitimate goal does not release politics from the duty to disclose effect, distribution and the cost of enforcement. A new burden does not become relief by carrying a name that sounds good.

Reducing bureaucracy: 45 million for business

The national regulatory control council presented its annual report on 8 October 2026. The headline from it was: in the reporting period July 2025 to June 2026 the annual compliance cost fell by around 6.8 billion euros, to the lowest level since 2020.^[8] That sounds like a turning point. Two sentences later it says how this relief is distributed.

Of the 6.8 billion euros, around 6.4 billion falls to citizens, around 370 million to the administration and around 45 million to business. In the same period the bureaucracy costs for business rose by 1.3 billion euros, above all through the implementation of the European sustainability reporting obligations.^[8] A business that waited for relief in 2026 got 45 million euros spread across every company in the country – and almost thirty times that in new obligations on top.

The year’s biggest relief consists in an obligation being scrapped that nobody had yet had to meet.

And the 6.4 billion for citizens is not what it sounds like either. Almost the entire relief – 7.7 billion euros – comes from a single act, the buildings modernisation act, above all from the scrapping of the obligation to run newly installed heating systems on 65 per cent renewable energy. The regulatory control council itself writes about this: considerable future costs would thereby no longer arise, “but no directly noticeable relief arises for those addressed by the rule either, since these costs had not yet been incurred”.^[8]

A future burden avoided on paper is not the same as a form that has disappeared today. For a business what counts is not how many billions are saved in a presentation. What counts is whether less working time goes into documentation, whether permits come faster and whether the next investment gets simpler rather than more complicated.

For an assessment in both directions: the regulatory control council nevertheless sees a positive trend. In the reporting period the annual compliance cost has halved across all those affected and still lies 6.4 billion euros above the level of 2011; at the end of the previous reporting period it was 13.2 billion. One-off compliance cost of 8.9 billion euros arose in the same period, however, 6.8 billion of it from the renovation requirements of the same buildings modernisation act.^[8] These 6.8 billion are not the 6.8 billion from the headline.

Protect performance rather than inventing new burdens

I do not want a weak state. I want a state that masters its core tasks and respects the people who finance it and work for it. Social protection and economic performance must not be played off against each other.

For that, four standards are needed. First: examine the task before building up staff. New administrative posts only with a comprehensible need, examined processes and measurable goals. Second: a full account rather than a label. Every new levy with a transparent presentation of the tax burden, enforcement, changes to data processing and ongoing effort. Third: relief with proof. Measure the reduction of bureaucracy by which obligations actually fall away and what time comes back in everyday life – not by which future costs have been avoided on paper. Fourth: core tasks first. Education, security, working infrastructure and reliable administration before everything else.

Germany needs no blanket disparagement of its public servants and no invented figures. It needs political leadership that lets itself be measured against a simple rule: before the state demands more, it has to show that it manages better with what it has. A capable state earns trust through results, not through the size of its budget.

Andreas W. HerbFounder and CEO MBG Group, Paderborn

What I expressly do not claim

I do not claim that every public euro is badly spent, that the state spending ratio has risen in every year, or that public service employees do too much or too little. Employee numbers alone prove neither the one nor the other. Three claims circulating in this debate are expressly not supported by the checked data: a state spending ratio of over 60 per cent including railway employees, more than 5.6 million employees in the public service, and a state spending ratio rising without interruption. The political demands in this text are my position, not the statement of the statistics.