## Notes from 24 January 2026 [[2026-01-23|← Previous note]] ┃ [[2026-01-25|Next note →]] I read a press release today from [Germany's Federal Ministry for Digital and State Modernization](https://bmds.bund.de/aktuelles/pressemitteilungen/detail/digitale-verwaltung-bund-bayern-und-hessen-starten-neuen-umsetzungsansatz) that, beneath the usual ministerial optimism, describes something interesting. The federal government is launching what it's calling a new implementation approach for administrative digitization, using Bavaria and Hesse as pilot states. The underlying mechanics suggest they're trying a new approach to solve a real coordination failure that's plagued German administrative reform for years. The context is the Online Access Act (Onlinezugangsgesetz or OZG), passed in 2017, which mandated that 575 administrative services be available online by 2022. The law required federal, state, and local governments to digitize citizen and business interactions, with implementation responsibility falling primarily on municipalities. By the 2022 deadline, most services remained offline or were available only in scattered municipalities. Germany's federal system distributes implementation authority across roughly 11,000 municipalities, sixteen states, and the federal level, with each controlling different pieces of the infrastructure, regulatory frameworks, and budgets needed for digitization. The original OZG essentially told municipalities to figure it out themselves with some federal funding and technical guidelines. Most couldn't or didn't. What makes this new pilot different is the accountability structure. The federal ministry and the two state governments are contracting what they're calling a "General Contractor" who takes end-to-end responsibility for getting services operational in every municipality in both states by December 2026. Not just building the software, but technical integration with existing municipal systems, organizational onboarding of local governments, coordination across all stakeholders, and ensuring actual operational capability. The political commitment is unusually specific for German administrative announcements. Bavaria and Hesse are each guaranteeing that five high-volume services will be available statewide by year-end. Bavaria selected online address changes, driving license applications, construction permits, residence permits, and weapons permits. Hesse chose a similar set but swapped construction and weapons permits for child maintenance advances and industrial facility permits. The governance implications are more interesting than the press release suggests. Municipalities keep the legal mandate to provide these services but are effectively losing IT sovereignty. They become consumers of a centralized service rather than operators of their own infrastructure. This reverses the original OZG logic, which assumed local implementation flexibility would drive innovation and adaptation to local needs. Instead, that flexibility produced fragmentation, duplicated effort across thousands of jurisdictions, and wildly varying citizen experiences depending on where you lived. The new model treats standardization as more valuable than local variation. The technical architecture seems to follow a unified front-end, distributed back-end approach. Citizens interact with a single portal without needing to know which server processes their request. The contractor is responsible for making this work across municipal IT systems that currently run on incompatible platforms, were built by different vendors, and use varying data standards. The project is explicitly designed to be "scalable"—other German states can join starting summer 2026, presumably using the same contractor and implementation model. Germany isn't unique in having multi-level governance problems for digital service delivery—most federal systems struggle with this. What's different here is the explicit decision to purchase coordination capacity from the private sector rather than build new intergovernmental mechanisms. The OZG failed because no single level of government had the authority, resources, and technical capacity to drive implementation alone. This model acknowledges that and essentially outsources the missing coordination function to a contractor who can operate across all three levels. Whether this represents state capacity being built or merely rented is the key question. If it works, other states copy the model, and the contractor's role becomes permanent, you've effectively created a private layer of administrative infrastructure that sits between federal policy and local implementation.