## Notes from 13 December 2025 [[2025-12-13|← Previous note]] ┃ [[2025-12-15|Next note →]] The [[Konrad-Adenauer-Stiftung (KAS)]] released a paper on public administration reform ("[More Performance – Less State?](https://www.kas.de/documents/d/guest/mehr-leistung-weniger-staat-1)") that stands out from the usual administrative reform rhetoric in Germany. What makes this paper interesting is that it moves past the generic "we need more interministerial cooperation" framing into specific operational proposals. The core argument centers on two main challenges: bundling central cross-sectional tasks (Personnel, IT, Procurement, Funding, Data Protection) at existing federal agencies, and reorganizing the subordinate administrative area which currently contains 946 (!) authorities. The most interesting strategic insight is about changing decision-making procedures from "_Einvernehmen_" (consensus/agreement) to "_Benehmen_" (consultation). This is about [[executive decisiveness]] – ending the model where collegial decision means everyone must agree, which gives individual ministries veto power. Under their proposal, centralization and standardization of tasks cannot be blocked against the lead department. Exceptions only come through Cabinet-level justification. This flips the default: instead of individualized ministerial work being standard and shared work requiring approval, shared services become mandatory default and exceptions require approval. The personnel focus is particularly welcome given that most German administrative reform discourse fixates on IT and digital government. They identify that personnel expenses doubled in ten years while trust in state capacity sank, and that management and central administration now make up almost 20% of federal employees (excluding Bundeswehr, the army). The diagnosis: many tasks performed in parallel up to 14 times, coordination overhead, demographic pressure binding personnel away from technical expertise and so on. Their proposed solution: create a _Federal Personnel Agency_ (_Bundespersonalagentur_) within an existing superior authority, subordinate to a specialized ministry. Initially governed by a steering group under a State Secretary with whole-of-government mandate, reporting to Cabinet, then transitioning to normal departmental management after two years. **Observations and concerns:** First, a definitional clarification since I read this through AI translation: when the paper emphasizes "non-political service delivery," they're likely distinguishing between policy (the vertical, substantive work of each ministry – education policy, health policy, defense policy) and operations (the horizontal administrative functions – HR, IT, procurement). In this framing, a central personnel agency isn't "political" because it's not making policy decisions about education or health, it's providing operational support that any ministry needs. This is a valid distinction. Personnel functions should indeed be separate from policy verticals – you don't want the Education Ministry running its own HR system differently from the Health Ministry just because they do different substantive work. The whole point is that these operational functions should be standardized horizontally. But here's my concern with how they frame it: even within the "operations" category, there's a difference between service delivery and strategic intelligence. The paper leans heavily into service delivery language – the Federal Personnel Agency would "make standardized service and consulting services of high quality available to all departments" and "supports as a service provider, not as a decision-making authority". This framing is politically convenient but undersells what a central personnel function actually needs to do. Yes, it should deliver operational services (payroll, benefits administration, recruitment processing). But it also needs to produce strategic workforce intelligence: analyzing demographic trends, identifying skill gaps, developing competency frameworks, setting standards for assessment and development, managing talent pipelines, conducting workforce research. This work isn't policy in the vertical sense, but it's also not just service delivery – it's strategic operations. Creating standards is inherently strategic work. Deciding what competencies matter for senior civil service positions, how performance should be assessed, what training priorities are, how to balance specialist versus generalist development – these are strategic choices about state capacity that shape what government can do. You cannot have a central personnel function that "just delivers services" without it becoming a center of strategic intelligence about the workforce. Unless the real ambition is genuinely just a shared services center handling transactional HR – payroll, benefits, basic recruitment logistics. That's useful but significantly smaller than what successful models actually do. The british model under the Cabinet Office, the American [[US Office of Personnel Management (OPM)]] paired with Office of Management and Budget (OBM, or even the [[Australian Public Service Commission (APSC)]] – these all do strategic workforce planning, not just service delivery. Second, if a central IT agency (_Bundesdigitalagentur_) would absorb existing digital authorities (as they propose with ITZBund under BMDS), a personnel agency needs to absorb things like the _Bundesakademie für öffentliche Verwaltung_ ([BAköV](https://www.bakoev.bund.de)). They mention this peripherally but don't emphasize it. The civil service college needs fundamental reform anyway – it's not particularly innovative in its current form. Folding it into a broader Federal Personnel Agency with clearer mandate and resources could be transformative. Third, placing this under Interior Ministry seems politically odd. The most effective models I know put central personnel functions either under Finance (because Finance has real clout – people respect the ministry that controls budgets) or under a Cabinet Office equivalent (the _Bundeskanzleramt_?). The British model has problems – the [[Public Sector Reform in the UK#Public Sector Reform in the UK#The Maude Review (2023)|Maude Review]] identified many of them – but the basic architecture of Cabinet Office ownership makes sense for maintaining cross-government perspective. The American model of OPM linked to Executive Office of the President, working closely with OMB, creates useful pairing: one handles budget/strategy, other handles people/talent. Those functions need to be tightly connected. Interior Ministry in Germany handles federalism, internal security, some digital infrastructure – it's not obviously the natural home for strategic workforce planning. Though I understand the practical politics: Interior already manages BVA, has administrative expertise, and might face less ministerial resistance than putting it somewhere more powerful. Fourth, their proposed steering committee model – representatives from all ministries helping make decisions in the new agencies – reminds me of the [[Délégation Interministérielle à l’Encadrement Supérieur de l’État (DIESE)]] delegates system in France. Every ministry has a DES (_Délégations à l'encadrement supérieur_) who participates in senior personnel decisions. This creates cross-ministry input but also multiplies coordination costs. A smaller committee might work better – easier to make decisions, less prone to lowest-common-denominator outcomes. Could be appointed by government and approved by parliament to guarantee political neutrality, similar to how regulatory commissions work. Or simply a smaller technical group not requiring representation from every single ministry. The goal should be effectiveness, not representational legitimacy. ___ This paper is interesting because it provides operational detail that most German reform discourse lacks. The consultation-not-consensus principle and the default-to-shared-services architecture are both correct moves. The distinction between policy verticals and operational horizontals is also important. My concern is that by emphasizing "service delivery" so heavily, they may be setting up a Federal Personnel Agency that's too limited in scope. Yes, it should serve departments, not dictate to them. But it also needs to build strategic intelligence about the federal workforce, set meaningful standards, and think seriously about long-term talent policies. That's not policy-making in the vertical sense, but it's also not just processing benefits and transactions.