## Notes from 22 February 2026 [[2026-02-21|← Previous note]] ┃ [[2026-02-23|Next note →]] I’ve really enjoyed [[Rainer Kattel]]’s [current Substack series](https://rainerkattel.substack.com/p/week-4-the-long-shadow-of-npm), a week-by-week module on public administration and public-sector innovation. Each post builds a small set of concepts (especially the idea that reform tools never arrive on a blank slate) and then uses those concepts to interpret both Western and non-Western administrative systems. In the week on [[New Public Management (NPM)]] (“[the long shadow of NPM](https://rainerkattel.substack.com/p/week-4-the-long-shadow-of-npm)”), for example, he argues that NPM mattered not only as a reform agenda but as a shift in organizational routines: it imported private-sector logics (targets, competition, contracting, performance incentives) and then built accountability machinery around them, changing what governments reward, notice, and learn. That structure makes the series unusually teachable, and I can easily imagine it helping people who teach public administration. At the same time, I struggle with the way he often writes about NPM. When he engages with NPM ideas or with implementation experiences, he rarely seems interested in the most generous interpretation of what NPM was trying to solve. The tone can feel settled in advance... less like “here is why this family of reforms appealed, what problems it aimed at, what it enabled in some contexts,” and more like “here is why it was misguided”, full stop. That may be cathartic (and sometimes fair), but it can flatten the story, and it risks missing why NPM travelled so widely, and why it sometimes produced outcomes that were not simply pathological. One reason a more charitable lens matters is that NPM didn’t land in a vacuum. Rainer himself makes that point well: administrative traditions and existing routines shape what gets imported, what gets distorted, and what becomes durable. But there’s also a capacity dimension that goes beyond “traditions”. In parts of Brazil, for instance, NPM-style ideas arrived in states and municipalities with limited administrative capacity, yet they were adopted because they read as “modern”... not because those governments had anything like the institutional conditions of the UK, Northern Europe, or the US. In some of these cases, my lived impression is that the attempt to “do NPM” forced a kind of basic state-building: setting up procedures, organizing data, creating minimal reporting routines, building a bureaucracy that could at least run procurement and manage contracts (even imperfectly). You can criticize outsourcing, the metrics, the distortions (often rightly!) but it’s also plausible that the reform impulse created enabling conditions that were better than the counterfactual of doing nothing, especially in places where there wasn’t a competing modernization ideology with similar traction. Relatedly, it can be helpful to treat NPM less as a single doctrine and more as a toolkit that got adapted (sometimes opportunistically, sometimes strategically) producing very different results depending on political coalitions and institutional choices. Chile is a useful illustration: “NPM-ish” principles under Pinochet and under the _[[Concertación de Partidos por la Democracia|Concertación]]_ were not the same project in practice, and the latter’s adaptations interacted with social policy goals in ways that arguably reduced segregation and inequality. You see something similar in the UK: Thatcher-era uses of the toolkit and Blair-era uses were not identical, even if they shared some vocabulary. And in South Africa, a number of these ideas were taken up in an [[African National Congress (ANC)|ANC]] discourse that wanted to break with a more traditional bureaucratic ideas tied to apartheid... again, not because the principles were neutral, but because the same set of managerial concepts can function differently when the political meaning of “modern administration” is contested. None of this is a defence of NPM as a default operating model. It’s more a plea for a slightly different historical sensibility: if we want students (and practitioners) to learn from the last forty years, we need to explain not only how NPM rewired routines in damaging ways, but also why it sometimes served as a bridge (however imperfect!) toward basic administrative capabilities in low-capacity settings, and why its effects varied so much across contexts. That richer picture doesn’t soften the critique; it makes it more useful.