## Notes from 18 January 2026 [[2026-01-17|← Previous note]] ┃ [[2026-01-19|Next note →]] ## Whitehall Monitor 2026 and the civil service workforce debate I read the [[Institute for Government (IfG)]]’s *[Whitehall Monitor 2026](https://www.instituteforgovernment.org.uk/publication/whitehall-monitor-2026)* (released January 13). It’s basically an annual X-ray of the UK civil service: headcount and grades, churn, geography, morale, digital capability, plus whether the government’s reform story matches what’s happening on the ground. To be more precise, it's not really an annual X-ray, but rather the best of its kind you can find in the whole world. What I like about the IfG’s style is that it treats “workforce” as a **governing system** (incentives, skills, management, accountability), not just as staffing levels. The tone is blunt, almost impatient: mission-led government might be a useful frame, but right now it’s “_so broad as to be functionally useless for driving workforce reform_”. ## Some workforce findings (what stood out to me) A few data points really anchor the debate: - **Scale + composition:** the civil service is up **35%** from the 2016 low to **520,440** (Q3 2025), even after most Brexit/pandemic surge work should have tailed off. The **policy profession is up 116%** since 2016. The implicit question is: are ministers actually better served by doubling policy capacity, or did the system just expand the easiest thing to expand? - **Grade inflation:** Grades **6/7 up 132% since 2010**, junior admin grades down nearly half. Automation explains some of it, but the story that feels most convincing is this: grade drift + faster promotion as a workaround for pay restraint. That’s not “workforce strategy”; it’s a quiet compensation mechanism. - **Churn as the hidden tax:** internal moves spiked in 2016/17 and never normalized. The 2024/25 drop seems driven by recruitment freezes rather than healthier incentives. This looks like a system that can slow churn temporarily, but can’t govern it. - **Geography is stuck:** senior civil service still **65% London-based** (target 50% by 2030). London has **24% more** civil servants than in 2010 while several regions and Scotland still have fewer.  ## Workforce reduction issues The fiscal frame matters because it forces the delivery question. Admin budgets are set to fall **11% (2028/29)** and **16% (2029/30)**. If staff costs are >80% of these budgets, then headcount cuts are going to happen. The IFS (Institute for Fiscal Studies) estimates (via the IfG) that the 16% target implies around **18% headcount reduction** (roughly **29k–40k**). What worries me is that the tools on the table to make it happen look blunt: - voluntary exits: **8,586 applications** by Aug 2025; - freezes + exits: easy to administer, hard to target (you lose high performers and block new talent); - performance management: the report keeps circling the same problem... government wants the option of removing poor performers but does not appear to have a machinery that actually does it at scale (the “mutually agreed exits” pilot had **30 applications** by Aug 2025). The Strategic Workforce Plan being repeatedly delayed (now “early 2026”) feels like a tell: the state is trying to cut without having built the managerial infrastructure to cut *intelligently*. ## Capabilities and performance gaps This is where the IfG is most useful for me: it insists that capability is the constraint, not just headcount. - Operational delivery is **54%** of staff but only **13%** of senior civil service, versus **26%** policy. So the “front line” is disconnected from who holds senior power and voice. - Digital: DSIT (the department responsible for digital government stuff) may diagnose correctly, but the question is leverage. Only **20%** of senior civil servants consider themselves digitally skilled; the policy–digital culture gap persists. About AI: you can’t magic productivity out of poor data and legacy IT. And then pay: senior civil service pay described as “broken,” **24% real-terms fall since 2010**, and the government responds (so far) with principles, not a plan.  ## Mission-led government in practice Reading the report, I think it's fair to say that “missions” as they currently operate are more as a **branding layer** than as a governing mechanism. The spending review tried to align structures around missions, but the IfG doesn’t see clear evidence the missions changed allocation decisions in a meaningful way (health still dominates; education stays middling; everything fits inside missions because missions are enormous and vague). The one thing that *does* look like a real operating model experiment is the *[Test, Learn and Grow](https://www.gov.uk/government/news/communities-across-the-country-to-benefit-from-innovation-squads-to-re-build-public-services)* program: central teams embedded with local delivery, iterating service design. But it reads like a promising pilot sitting next to a machine that still runs on old incentives. ## Contrast with the German and Brazilian administrative debate This comparison keeps nagging me after [[2026-01-17]]. The UK has problems, but it has a public ecosystem that *lets you argue about them* in capability terms. The Whitehall Monitor (plus NAO, committees, open-ish data like the People Survey) creates a standing platform for saying: “This is what the workforce looks like; this is what incentives are doing; this is where delivery capacity is thin; here is who is accountable”. Germany and Brazil, by contrast, are not missing this because we lack money, talent, or technical capacity. Germany is at least as wealthy and administratively sophisticated as the UK. The absence feels political/institutional: **who gets to define the workforce story in public, and what questions are considered legitimate to ask.** In practice, the dbb’s **[[2026-01-17#The Monitor öffentlicher Dienst|Monitor öffentlicher Dienst]]** becomes the closest thing to an integrated annual workforce narrative in Germany—useful on data, but structurally shaped toward what would be useful for a union to argue about: external pressures and protection. That isn’t “wrong”, but it crowds out a different conversation: performance, capability gaps, management quality, productivity, and what delivery actually requires. I don’t buy “federalism makes it impossible” as the full explanation. Coordination is hard, sure—but fiscal data consolidates across levels despite comparable complexity. My guess is that the blocker is not capacity; it’s **political economy**: a reform debate architecture that incentivizes avoidance of performance discourse (because it quickly reads as an attack on the public service) and lacks an independent institution with the mandate, data access, and credibility to keep capability questions on the agenda year after year. The UK shows you can build a capability-and-performance debate outside government (IfG-style). Germany’s (and Brazil's!) gap looks less like a technical omission and more like a choice—an equilibrium—about what the system prefers to argue about (and what it prefers to keep off the table).