17/11/2026

Dutch researcher mobility: what the 2020 numbers can and cannot tell us

 Summary

  • Churn dominates. Dutch universities exchanged about 6,500 researchers with the world in 2020 (3,353 in, 3,191 out), roughly 6% of the research workforce, so the net balance is the smallest number in the picture.
  • The net turned positive, but only slightly. Scopus shows a deficit of 434 in 2015, +108 in 2019 and +162 in 2020, which is 0.15% of the workforce.
  • One corridor explains 2020. The US corridor swung by 160 (from -65 to +95). Without it, the Dutch net fell from +173 in 2019 to +67 in 2020, and a pandemic effect is a plausible explanation.
  • Italy and Switzerland deserve more attention. Italy was a net donor every year, but the surplus shrank from +62 to +27. Switzerland pulled a growing net of researchers away, from -16 to -55.
  • Two databases disagree on size. For 2018, Scopus shows a net of -35 and OpenAlex shows -221, so the “brain gain” claim rests on one source’s reading of two years.
  • Policy is moving but not settled. The Tulip Fund funded 34 researchers, 29 from the US (including returning Europeans), and the bill dropping the English-language test has yet to pass.
  • What leaders should do. Target outreach at Italy and the US, fix housing, partner careers and start-up support, track retention by cohort, and watch Germany and Switzerland as competitors.

Background

Dutch universities are among the most open research systems in the world, and the data show it. In 2020, Scopus records 3,353 researchers arriving in the Netherlands and 3,191 leaving, out of about 108,000 active publishing researchers. That is roughly 6,500 moves in a single year, or about 6 percent of the workforce changing country of affiliation.

A tempting headline follows. For years the Netherlands lost more researchers than it gained, with a net deficit of 434 in 2015. In 2019 the balance turned positive (+108), and in 2020 it grew to +162. Add the US research funding cuts, the Dutch Tulip Fund, and a new coalition that has reversed course on English-taught programmes, and the story writes itself: the Netherlands moved from brain drain to brain gain.


I think that story is too neat. This post sets out what the data support, what they do not, and what I would do about it as someone running or advising an institution.

01/11/2026

Twenty years after the Stern Review the world looked away, and the economics got better

 Blog post, based on the LSE launch of The Growth Story of the 21st Century (5 November 2025).

In a nutshell. Twenty years on, Lord Stern's verdict is that the climate is changing faster and more severely than we thought in 2006, and that the Review's core argument has only grown stronger: preventing climate damage costs far less than living with its effects. His own summary at the launch was that the original conclusions were right, and are stronger today.


1. Background: how the world lost the thread

Climate change did not disappear from the news. It disappeared from the agenda. Those are different things.

Every few years a new shock takes the room. A financial crisis. A pandemic. A war in Europe. Artificial Intelligence. An election that changes who holds the budget. Each of these is serious, and none of them is an excuse that anyone should have to defend. 

But Nicola Ranger, speaking at the LSE launch, described the pattern plainly: when shocks hit, political capacity and fiscal space shrink, and the climate agenda rolls back. The transition she and her colleagues expected twenty years ago to be smooth has turned out to be a process of shocks.

17/10/2026

Intelligenza artificiale e futuro del lavoro: cosa ci insegna la storia dei grandi sistemi tecnici

Lezione di Albert Schram, PhD

Abstract

Questa lezione esamina l'impatto dell'intelligenza artificiale (IA) sul lavoro e sulle imprese dal punto di vista della storia economica dei grandi sistemi tecnici. Parte da un paradosso. Le organizzazioni hanno buone ragioni per essere prudenti: secondo la RAND Corporation oltre l'80% dei progetti di IA fallisce, e la rapidità dei progressi tecnologici rende presto superati gli investimenti. Le persone, invece, non possono permettersi di aspettare, perché le competenze richieste dai datori di lavoro stanno già cambiando. 

La storia offre tre chiavi di lettura: le ferrovie italiane dell'Ottocento, i cui benefici arrivarono con decenni di ritardo per carenze istituzionali; l'elettrificazione delle fabbriche americane, che divenne produttiva solo quando l'organizzazione fu ridisegnata; e il concetto di "saliente rovesciato" di Thomas P. Hughes. I dati del World Economic Forum, di Eurostat, dell'Istat e del Politecnico di Milano mostrano che in Italia l'uso dell'IA è raddoppiato in un anno. Resta però sotto la media europea, è poco strutturato e il divario tra grandi e piccole imprese è in crescita. Il principale ostacolo all'adozione non è la tecnologia, ma la mancanza di competenze adeguate. 

La lezione confronta il modello della sostituzione con quello del potenziamento e discute la "frontiera tecnologica frastagliata" e il passaggio agli agenti di IA. Propone inoltre un metodo per valutare gli investimenti basato su due filtri e tre lenti (efficienza, crescita e valore strategico). Conclude che oggi il saliente rovesciato dell'IA sono le competenze e l'organizzazione. Per gli studenti questo vuoto è un'occasione: unire una solida competenza disciplinare alla capacità di usare, verificare e governare l'IA.




Executive summary in 15 punti

  1. Il paradosso centrale. Le organizzazioni possono permettersi di aspettare sull'infrastruttura. Nessuno può permettersi di aspettare sulle competenze.

  2. La lezione delle ferrovie. Nell'Ottocento le ferrovie italiane contribuirono poco all'unificazione economica, e i loro benefici arrivarono soprattutto nel Novecento. Il limite non era tecnico ma istituzionale.

12/10/2026

Rebuild, Don't Restore: The World Order After the Storm Warnings

Margaret MacMillan on the end of the rules-based order, and where I disagree


This is a conversation with Margaret MacMillan, a historian and emeritus professor at the University of Oxford. Her latest work focuses on the collapse of the rules-based order we lived in until recently, and its replacement with something much more chaotic, unstable, and dangerous. They discuss why today is as significant as the biggest milestones of the 20th century, whether current American foreign policy is here to stay, and what the long-term consequences of the war in Ukraine might be. [Link to the full episode: https://www.youtube.com/watch?v=zOyW_XgEvpI&t=1239s]

11/10/2026

From the Moon Landing to AI: A Boomer's Front-Row Seat to Change.

The Last Boomers

I was born in 1964, the last year of the baby boom, a latecomer to my own generation. My parents were the Silent Generation, the war children, who looked up to the Greatest Generation, the people who had won the war. They were loving parents, and my youth was as good as I could have imagined. 

They taught me three things I still live by: how to look after myself, how to live by my core values and stand up for them, and how to find real joy in people and in simple things. The first shows in my life. I have never been unemployed or dependent on others, and I have no debts. Soon most of the boomers will be gone, and I want to say what it was like before the story goes quiet.


05/10/2026

China's Education Miracle — and Its Blind Spot: Rankings, Robots, and the Missing Humanities

Two recent analyses — one dissecting the THE 2027 World University Rankings [1] and another reporting from the ground at WorldDidac Asia 2026 in Hangzhou [2] — document the extraordinary rise of Chinese education. But read together, and against a broader critical lens, they also raise a harder question: what kind of graduate is this system designed to produce, and what is being left out?




04/10/2026

Don’t Ban the Bot, Protect the Pedagogy: A Practical AI Governance Framework for Secondary Schools

Adapting Higher Education Validity Models to Meet the Realities of Secondary Classrooms

When Eindhoven University of Technology (TU/e) overhauled its assessment strategy around generative AI, it demonstrated that governance is not about catching cheaters—it is about validity protection: ensuring that assessments measure real student understanding (Sadowski & Oliveira, 2026).



However, secondary schools operate under vastly different constraints than universities. While higher education institutions enjoy institutional autonomy and research budgets, secondary schools deal with rigid exam board syllabi, strict safeguarding duties for minors, severe teacher workload constraints, and an environment that a recent report described as "chaotic" and "crying out" for national guidance (Clarke, 2026; Seldon & Bunting, 2026).