Thanks for this Julian. I loved the point about Studs Terkel finding people felt a greater lack of meaning in their jobs in the 1950s, especially in light of Keynes's pro-leisure essay 'The Economic Opportunities of our Grandchildren'.
I also agree that AI could be an opportunity to improve the quality of re-training. I'd be interested to know what Multiverse argue about this and if there are any studies into their approach as I gather they've been quite successful and mainly do online learning. But I'm also animated by the David Autor essay from 2024 about how mid-skilled workers can use AI to upskill, which I feel compliments your piece.
Thanks, Phil. I agree that a particularly important follow up question from this piece is what steps employers, workers, and governments can take to more effectively retrain. AI may be useful as a tool to support retraining, but the evidence seems to show the single most important predictor of successful retraining is employer engagement in the process. In other words, if you have an employer - or a potential employer - engaged throughout various stages of the retraining, that can be a good way to create the relationships and tailored skillsets that make a person a good candidate for employment.
What I'm curious about is whether this sort of retraining works at a national population level. We know there are examples of effective programs with a few thousand participants. But we don't know if that kind of programming works at a larger scale. It could be that those programs are useful because they reduce labor market friction and create more 'good jobs' by hastening the process of re-employment for people who experience some sort of displacement. It could also be that there are a limited number of 'good jobs' and those programs simply shift who has access to them.
Interesting that's a great point about the employer engagement. I did some research into vocational education and training in Switzerland a few years ago, which I think is an interesting case study as their labour market is very flexible (unlike Germany) so there aren't as clear employer incentives to engage with retraining, but still the outcomes seem very strong. I think the employer engagement in the Swiss case comes through companies actually helping design the apprenticeships/ lifelong learning programmes, and I think they actually put in more of the funding for training than in the UK, for example. So perhaps Switzerland is a good case study for your point about employer engagement being key.
This was a super interesting read. Thanks Julian!
Thanks for this Julian. I loved the point about Studs Terkel finding people felt a greater lack of meaning in their jobs in the 1950s, especially in light of Keynes's pro-leisure essay 'The Economic Opportunities of our Grandchildren'.
I also agree that AI could be an opportunity to improve the quality of re-training. I'd be interested to know what Multiverse argue about this and if there are any studies into their approach as I gather they've been quite successful and mainly do online learning. But I'm also animated by the David Autor essay from 2024 about how mid-skilled workers can use AI to upskill, which I feel compliments your piece.
Thanks, Phil. I agree that a particularly important follow up question from this piece is what steps employers, workers, and governments can take to more effectively retrain. AI may be useful as a tool to support retraining, but the evidence seems to show the single most important predictor of successful retraining is employer engagement in the process. In other words, if you have an employer - or a potential employer - engaged throughout various stages of the retraining, that can be a good way to create the relationships and tailored skillsets that make a person a good candidate for employment.
What I'm curious about is whether this sort of retraining works at a national population level. We know there are examples of effective programs with a few thousand participants. But we don't know if that kind of programming works at a larger scale. It could be that those programs are useful because they reduce labor market friction and create more 'good jobs' by hastening the process of re-employment for people who experience some sort of displacement. It could also be that there are a limited number of 'good jobs' and those programs simply shift who has access to them.
Interesting that's a great point about the employer engagement. I did some research into vocational education and training in Switzerland a few years ago, which I think is an interesting case study as their labour market is very flexible (unlike Germany) so there aren't as clear employer incentives to engage with retraining, but still the outcomes seem very strong. I think the employer engagement in the Swiss case comes through companies actually helping design the apprenticeships/ lifelong learning programmes, and I think they actually put in more of the funding for training than in the UK, for example. So perhaps Switzerland is a good case study for your point about employer engagement being key.