We argue that robots are not destroying jobs in a significant way, computers are not polarising labour markets, and AI tends to complement rather than substitute labour. In our view, the main impact of digital technologies has been on the nature and organisation of work, not on employment numbers.

In the research summarised in this report, we try to confront dominant narratives with carefully compiled evidence. In the last few years there has been a kind of "automation anxiety" which in our view is completely unsupported by data.

Empirical application across 24 EU countries (2011-2022) suggests (misleadingly) very little change in the share of green jobs in the last 15 years. Furthermore, no correlation is found with key environmental indicators. 6/7

The main conceptual problem is that the reliance on occupational titles leads to false positives (misclassification of non-green jobs) and false negatives (exclusion of essential contributors). This jeopardizes accurate assessment. 3/7

A seismic shift

This is pretty cool! A word cloud based on my google scholar profile. You can make your own here: scholargoggler.com