Few complaints are as widespread as the one about German bureaucracy. Applications take forever, responsibilities are unclear, digitalisation stalls.
The complaint is justified. The explanations usually offered — lazy officials, absent will — are not. The actual causes are more interesting and harder to fix.
Federalism as the starting point
The most important structural factor. Competences are distributed between federation, states and municipalities, and that distribution is anchored in the constitution.
Concretely: the federation can pass a law in many areas, but execution lies with the states and practical implementation often with municipalities. Each level has its own procedures, its own software, its own staffing.
A nationally uniform digital process therefore requires coordination between sixteen states and thousands of municipalities. That isn't a failure, it's the design.
The design has a purpose — distributing power, proximity to citizens, protection against centralised error. Its price is slowness, and we pay that price deliberately, even if it's rarely put that way.
Legal protection costs time
The second large factor, and the least understood.
German administrative law grants extensive rights of participation and challenge. For larger projects, authorities are consulted, associations heard, objections processed, and decisions can be litigated.
Every one of those steps exists for a reason. Participation prevents decisions being taken over the heads of those affected. Rights of challenge prevent arbitrary administrative decisions.
And every one of those steps takes time. A planning procedure respecting all participation rights can run for years before anything gets built.
Anyone wanting faster procedures must say which of these rights should be curtailed. That's usually avoided in the debate because it's unpopular.
The written-form requirement
A legal detail with enormous practical consequence.
Numerous provisions require written form, traditionally meaning a handwritten signature on paper. As long as that stands in the law, no authority can digitalise the procedure even if it wanted to.
Considerable effort has gone into removing such requirements in many areas and creating digital alternatives. But the stock of provisions is vast, spread across federal and state law, and reviewing it is laborious detail work with no political visibility.
Digitalisation isn't a software problem
The most common error of thinking in this debate.
Digitalising an administrative procedure doesn't mean putting a form online as a PDF. It means rethinking the whole process: which data are needed, where do they come from, who checks what, how are decisions documented, how are deadlines observed.
Frequently it emerges that the existing process grew over decades and nobody has a complete overview. Digitalising it without first understanding it produces digital renderings of nonsense.
Then there's the data silo problem. Authorities often may not exchange data for data protection reasons, which is why citizens must supply the same information repeatedly. The so-called once-only principle is meant to change this and is legally demanding.
Staffing
A factor that will worsen matters in coming years. A substantial share of public sector employees will retire within a foreseeable period.
At the same time the public sector competes for skilled workers, particularly in IT, at a disadvantage given pay structures.
This meets a moment when tasks are increasing rather than decreasing. It doesn't take a prophet to see problems ahead.
What would actually help
It is worth being fair about what the system does well, because the complaint tends to crowd it out.
German administration is, by international comparison, remarkably free of petty corruption. You do not pay anyone to move a file. Decisions are documented, reasoned and challengeable. Records are accurate and durable. Registration systems mean that a great many things which require active effort elsewhere simply happen — you are enrolled, notified, and counted without doing anything.
Those properties are not free. Much of the procedural weight people find maddening is what produces them. A system that decided quickly and without documentation would be faster and would also be one in which outcomes depended more on who you knew.
Which is not an argument against reform. It is an argument for being specific about which delays are protecting something and which are simply the residue of processes nobody has revisited in forty years. Those two categories look identical from a waiting room, and telling them apart is most of the actual work.
A few points on which specialists broadly agree.
Standards rather than individual solutions: shared interfaces and components built once and reused, instead of parallel development.
Legal simplification: consistent removal of form requirements and proof obligations that make no sense in a digital context.
Procedural reform: deadlines for authority consultations, concentration of competence for major projects, deemed approval where deadlines lapse.
And the most honest measure: fewer regulations. Every new provision creates enforcement work, and that work is rarely counted when the provision is adopted.
What I don't expect is rapid improvement. Administrative reform is a topic without campaign glamour, with long horizons and no ribbon to cut. That's the real reason it happens so seldom.