Technological unemployment

Technological unemployment occurs when developments in technology and working practices cause some workers to lose their jobs.

Technological unemployment is considered to be part of a wider concept known as structural unemployment.

Example of technological unemployment

When labour-saving machines are introduced into the productive process, a firm can get rid of workers and produce the same amount of goods than before. Therefore, some workers can lose their job (and may take time to find a new one).

Overall Impact on Unemployment

  • Technological change doesn’t have to increase overall unemployment, even though some types of workers may temporarily lose their jobs.
  • For example, in 1800, the majority of British workers were employed in agriculture. Labour-saving technology meant that food could be produced with fewer workers and so some agricultural labourers lost their jobs as farms used more machines.
  • However, as jobs were lost in agriculture, new jobs were created in producing machines.
  • Similarly, advances in computers and robots meant that firms could produce manufactured goods with fewer workers. The increased productivity in manufactured goods meant that the relative cost fell, giving more opportunities for people to work in the service sector.

Why economists are sceptical of mass permanent unemployment

Historically, waves of automation (steam power, electricity, computers, the internet) displaced jobs but did not reduce long-run employment. The key reason is that productivity gains create new demand and new tasks:

  • Lower costs → lower prices → higher demand
  • Higher productivity → higher real incomes → spending on new goods and services
  • New technologies create entirely new occupations (often hard to predict in advance)  For example, growth in AI, may lead to more things like personal trainers, life coaches and musicians.

From this perspective, AI fits the standard Luddite fallacy critique: machines replace tasks, not humans as a whole.

Why does technological change not cause unemployment?

  • Let us suppose, technological change means we can produce food with fewer workers. Therefore, it is cheaper to produce food and the price of food should fall.
  • This means that people can spend a smaller percentage of their income on buying food. Therefore, people have more money to spend on other goods and services.
  • This increased demand for manufactured goods causes higher demand, and therefore there will be a higher demand for workers.
  • Technological innovation merely changes the types of jobs that occur in the economy. If labour productivity increases, we can enjoy a greater range of goods and services.

uk-employment-coal.

  • In 1920, there were 1.2 million coal miners in the UK. Technological change was a factor in the number employed fall to less than 5,000 in 2012. (decline of UK Coal)

typists-and-word-processors

A more modern example, in the US in 1978, there were one million employed as typists. The wordprocessor and personal computer saw these jobs largely vanish by 2020. (professionals end up doing their own typing). But, these jobs have been shifted elsewhere.

UK Unemployment during technological change

uk-unemployment-79-24

Why Technological change can increase unemployment

  • If labour markets are flexible, then technological change will not cause unemployment. However, if there are labour market inflexibilities, then it can cause unemployment – at least, for a certain time period.
  • For example, due to technological change, coal miners may lose their jobs. However, due to occupational and geographical immobilities, they may be unable to take new jobs in the service sector. (e.g. a miner may not have skills to work in computers; he may find it hard to relocate).
  • In this case, technological change can cause a temporary increase in unemployment – which will last until the coal miners develop greater skills and ability to move.

Evidence of technological unemployment in the US?

technologocical-unemployment-labour-prod

Since 2000, productivity growth has become detached from employment growth. During the early 2000s employment grew at a slower rate than productivity. Since the end of the great depression, employment growth has picked up (though in a flexible labour market – many new jobs are low paid). But, this might indicate the gains in productivity from automation are leading to lower job growth (though there could be other factors too)

us-corporate-profit-jsn-25

In this period from 2000, there was a sharp jump in corporate profits, which suggests companies are gaining higher profit from increased productivity.

Labour share of GDP

labour-share-usThis shows that labour (salaries, wages) are taking a smaller share of GDP since 1990. This also doesn’t reflect the rise in wage inequality and growth in salaries of top 1%

Will AI Cause Technological unemployment?

The technology everyone is thinking about at the moment is AI. AI will affect jobs in areas previously untouched by technological change.

What is different about AI

AI is more disruptive than past technologies because it:

  • Automates cognitive and professional tasks, not just manual labour
  • Affects white-collar, graduate, and middle-income jobs
  • Scales extremely quickly (software diffuses faster than physical capital)

This makes short- to medium-term technological unemployment more likely, especially in:

  • Routine office work (admin, basic accounting, paralegal tasks)
  • Entry-level graduate roles
  • Content production, customer support, and basic analysis

So while total employment may recover, specific groups may face persistent adjustment costs.

The key constraint: demand and human scarcity

Even the most technological advanced AI still faces limits:

  • Many services require human trust, accountability, empathy, and legal responsibility
  • Rising productivity does not eliminate demand — it often expands it (e.g. healthcare, education, leisure, personal services)
  • Humans remain scarce in areas like care work, creative direction, leadership, and complex judgement.

An economy rarely runs out of things people want — it runs out of productive capacity.

AI inequality?

A bigger risk with AI is inequality in distribution of proceeds of AI profit. Will it be captured by big tech monopolies, or will we have a universal basic income providing everyone with income

Related

 

About The Author

Item added to cart.
0 items - £0.00