Fact: Automation will replace tasks, not entire jobs

Number: 5
Year: 2024

A common perception is that “robots will take our jobs”. While for many, this is a very unsettling idea, others predict that the automation of work will solve one of our biggest challenges: the declining workforce in our ageing societies. The reality is, however, that both the worries and the hopes are mostly unfounded.

A young person is taking notes at an office desk which is cluttered with office supplies

Automation will not “save” us, and neither will it “take our jobs” any time soon. It will, however, replace certain tasks. Most employers do not appear to expect layoffs because of automation. In a survey published in 2018 by the staffing firm ManpowerGroup, firms in the US were asked how they expect automation and digitisation will affect jobs in the short term. Results indicated that only 4% of firms expect job losses, with 91% planning to maintain or even increase the number of workers [1].

Firms in most countries in Europe, also expected an increase in workforce according to the survey, especially in the continent’s largest economies. However, many stated that digitisation or automation may require re-skilling or on-the-job training to adapt to modifications of the jobs.

Why are all these companies so confident in maintaining or increasing employment levels? Well, how likely it is for a task to be replaced by automation depends on the kind of task – more specifically on the worker’s skill level. Researchers identify three skill levels: high, medium and low.

High-skill tasks require extensive and specific knowledge – think of pilots or surgeons. They have robots to help them, but to programme their very specialised knowledge into robots would be incredibly impractical. Also, people are suspicious of leaving high skilled tasks to robots alone. Would you accept a machine to fly your plane or to operate on you all alone?

Then there are low-skill tasks – equally unlikely to be automated, but for a different reason: they aren't worth the cost of it.

It is the medium-skill jobs which are considered most at risk. These are jobs that require post-secondary school training, but not necessarily a university degree or similarly advanced training. Think construction work or information and support services. Medium-skill workers usually make up the largest of the three groups. Many tasks here could be taken over by machines, especially with the current advances in artificial intelligence.

However, it is important to note that the replacement of certain tasks does not equal the replacement of entire jobs. Here, the retraining and social safety nets are important. Existing examples of automation and digitisation have shown that maintaining the labour force is important for an economy, for example humans are often needed to operate machines or software.

What we can certainly say: advances in robot technology allow for increases in productivity of labour - whether packing boxes in an Amazon warehouse, helping with car manufacturing or data processing. But automation will not simply put people out of a living – provided the right policies of labour market resilience are implemented. The FutuRes project develops recommendations for such policies, based on data-driven research.

Emily Barker, research fellow with focus on theoretical and empirical macroeconomics, University of Southampton


[1] See the findings of the automation survey by ManpowerGroup (2018): Skills Revolution 2.0

Photo: Anna Tarazevich/Pexels