Opinions

AI in Keynes’s theory

 (AdobeStock)

4' min read

Translated by AI
Versione italiana

4' min read

Translated by AI
Versione italiana

It has been 80 years since the economist John Maynard Keynes passed away, but his ideas, preserved in his works, are still relevant today.

In his 1930 essay, *Economic Possibilities for our Grandchildren* (Keynes, 2010), Keynes makes an interesting prediction: the increase in productivity linked to technological progress would mean that, within 100 years (and therefore by 2030), the economic problem – understood as the struggle to meet basic needs – would no longer exist.

Loading...

In particular, according to Keynes, workers would have worked very little – around 15 hours a week – to meet their needs; consequently, the problem would no longer have been work, but rather how to spend their free time.

We are now approaching 2030; Keynes’s prediction has not come to pass – in fact, working hours have not been significantly reduced, but the issue he raised remains highly relevant today, as Artificial Intelligence (AI) is the subject of debate due to its potential negative effects on the labour market, leading to the much-feared notion of ‘technological unemployment’.

Indeed, the debate in the literature regarding the impact of innovation on the labour market is quite lively (Vivarelli, 1995): on the one hand, innovation can make it possible to achieve the same level of output with less labour, leading to the negative effect known as the displacement effect; on the other hand, reducing the labour input means incurring lower labour costs (measured by wages), and this reduction could allow for a reduction in the selling price of the final good in a competitive market, which might stimulate demand, thereby having a positive effect on the labour factor – the ‘compensation effect’. Even in a non-competitive market, we might observe the compensation effect: in fact, in this case, the entrepreneur does not pass on the cost reduction to prices, but the higher profits could be invested in further innovation in the future. However, the ultimate effect of innovation on employment is unclear, as it depends on various factors such as the degree of competition in the market and the elasticity of demand.

In recent years, there has been increasing discussion about the introduction of robots and AI into the economy. With regard to the process of robotisation, a recent study (Caselli et al., 2025) analyses which occupations are most at risk from robots and what the overall effect of automation has been on the labour market. Building on other studies (Acemoglu and Restrepo, 2020), this paper examines the potential overlap between the tasks performed by robots and those required by occupations. The results show that robots have not had a negative impact on employment, confirming the findings of other similar studies (Dottori, 2021). However, the number of workers employed in the design and maintenance of robots has increased by around 50 per cent in less than ten years. This demonstrates that when firms invest in robotisation, the number of workers performing complementary tasks increases (Acemoglu and Restrepo, 2019). Furthermore, there is an increase in routine jobs in general, and in cognitive routine jobs in particular, where investment in robotisation is highest. Unfortunately, there has been a decline in jobs requiring physical effort from workers. As we can see, therefore, robotisation can boost productivity but may also exacerbate income inequality in the labour market.

Another recent study (Bonfiglioli et al., 2025) shows that AI can either automate jobs or complement workers’ tasks. To measure the degree of exposure to AI, the study examines employment growth in AI-related occupations. The results show that the impact is negative for low-skilled workers and production workers, whilst the impact is positive for workers at the top of the wage distribution and for those working in STEM (Science, Technology, Engineering, Mathematics) professions. It is therefore demonstrated that AI has contributed to the automation of jobs and to the widening of inequalities.

The previous discussion highlights the importance of robotisation and AI for the economy, as they help to boost productivity; however, the benefits are not felt by everyone.

It is therefore important to assess the economic reasons that have prevented the reduction in working hours, as predicted by Keynes; at the same time, re-reading the economist’s essay could help to identify the measures needed to capitalise on the opportunities generated by AI and mitigate its potential negative effects on the labour market. We are, in fact, confident that new jobs can be created, but there is a significant risk that these posts may remain unfilled due to a mismatch between labour supply and demand in terms of skills.

It is therefore clear how important economic policy measures are: on the one hand, to improve access to education and ensure that people possess the skills required by the labour market; and on the other hand, to implement the redistribution of wealth through the tax system, so as to tackle the problem of income inequality, which may increase as a result of the distortions caused by AI in the labour market.

Loading...

Bibliography

Acemoglu, D., Restrepo, P. (2020). Robots and jobs: Evidence from US labour markets. Journal of Political Economy, 128: 2188–2244. https://doi.org/10.1086/705716

Bonfiglioli, A., Crinò, R., Gancia, G., Papadakis, I. (2025). Artificial Intelligence and Jobs: Evidence from US Commuting Zones. Economic Policy, 40: 145–194. https://doi.org/10.1093/epolic/eiae059

Caselli, M., Fracasso, A., Scicchittano, S., Traverso, S., Tundis, E. (2025). What workers and robots do: An activity-based analysis of the impact of robotisation on changes in local employment. Research Policy, 105135. https://doi.org/10.1016/j.respol.2024.105135

(*) University of Salerno

Copyright reserved ©
Loading...

Brand connect

Loading...

Newsletter

Notizie e approfondimenti sugli avvenimenti politici, economici e finanziari.

Iscriviti