Technological unemployment

Topic: Employment, Unemployment and Informalisation · NCERT: Beyond NCERT

Meaning

Technological unemployment is unemployment that happens when machines, automation or artificial intelligence (AI) do work that people used to do, so those workers lose their jobs.

It matters because the jobs lost may never come back in the same form. The displaced worker needs new skills, not just a recovery in demand. The Economic Survey 2024-25 calls the link between AI and labour one of the biggest uncertainties India faces [5].

Explanation

How it works

• Basic chain:
• A firm adopts a new machine, software or AI tool.
• The tool does a task faster or more cheaply than a worker.

• The firm needs fewer workers for that task, and some lose their jobs.

• The job loss comes from the supply side, meaning a change in how goods are made. It does not come from a fall in total spending.

• So it can happen even when the economy is growing.
• It usually lasts a long time. The old skill is no longer wanted, so the worker must learn a new one or move to a new place.
• Textbooks often treat it as a special case of structural unemployment (a long-term mismatch between workers' skills and the jobs available). Our notes list it separately because of its link to AI.

Substitute or complement: the key question

• Substitute: the machine replaces the worker, so jobs are lost.
• Complement: the machine works alongside the worker and makes the worker more productive.

• The worker's output per hour rises → the worker becomes more valuable → wages and jobs can grow.

• The Economic Survey 2024-25 says AI may complement labour in the near term while firms learn to use it. But the productivity gain from this has a ceiling [5].

• In the longer run, the Survey warns that AI may automate a large part of economically valuable work. This could cause large-scale labour displacement, especially of workers in the middle and lower parts of the wage distribution [4].

What makes it rise or fall

• It rises when:
• jobs are made up of routine, repeatable tasks such as data entry, basic coding or back-office work, because these are easy to automate
• machines become cheaper than the workers they replace

• technology changes faster than workers can retrain

• It falls when:

• workers are reskilled, so they move into the new jobs that technology creates (for example, jobs building, running and repairing the machines)
• cheaper production lowers prices → people buy more → firms need more workers elsewhere in the economy

• policy pushes firms to use AI to augment workers (add to what they can do) rather than replace them

• Right cure (from our comparison table): reskilling and social protection (income support for workers who lose jobs). A demand stimulus alone does not help, because the problem is not low spending.

In India

• The historical fear: the Gandhi epigraph in the NCERT employment chapter reflects Gandhi's fear that machines would displace labour in a labour-rich country like India.
• A classic Indian example: handloom weavers lost work when powerlooms replaced them. This is technology causing long-term job loss in a whole trade.
• The AI threat to IT-BPM:
• India's IT-BPM (information technology and business process management) sector grew through labour arbitrage. This means rich countries moved work to India because skilled labour here was cheaper.

• AI can now automate routine coding and back-office tasks. These are exactly the middle- and lower-wage jobs that the Economic Survey flags [4].

• Economic Survey 2024-25 view:

• The risk is large-scale displacement of middle- and lower-wage workers [4].
• India's advantage is its young, tech-savvy workforce, which can use AI to augment its productivity [4].

• It calls AI and labour one of the biggest uncertainties India faces [5].

• Why youth matter here: youth unemployment (15-29 years) was 10.2% in 2023-24, about three times the overall rate [2]. Young entrants to IT and back-office work are the group most exposed to automation of entry-level tasks.

• Policy tools: Skill India-type reskilling, apprenticeships and social protection for displaced workers.

Don't confuse with

• Structural unemployment: the wider category. It covers any long-term mismatch of skills or location, for example Ahmedabad textile mill workers after the mills closed in the 1980s. Technological unemployment is the part of it caused specifically by machines, automation or AI.
• Cyclical unemployment: caused by low aggregate demand in a recession, and it ends when the economy recovers. Technological unemployment can happen in a boom, and a fiscal or monetary stimulus does not cure it.
• Frictional unemployment: short gaps while people search for or switch jobs. Better job information fixes it, for example through the National Career Service portal [3]. Technological unemployment is long-term and needs reskilling.
• Disguised unemployment: too many people doing work that fewer could do, as on a family farm, so their extra output is close to zero. They still look employed. A technologically unemployed worker has actually lost the job.

Prelims Hooks

• Definition: unemployment caused when workers are replaced by machines, automation or AI. It is long-term, and textbooks usually treat it as a type of structural unemployment.
• Cure: reskilling and social protection, not demand stimulus. Trap: a fiscal stimulus cures cyclical unemployment, not technological or structural unemployment.
• Economic Survey 2024-25: AI may cause large-scale displacement of workers in the middle and lower parts of the wage distribution [4]. It sees India's advantage in a young, tech-savvy workforce that can use AI to augment productivity [4].
• Economic Survey 2024-25 (AI ecosystem chapter): AI may complement labour in the near term, but the productivity gain has a ceiling. It calls AI and labour one of the biggest uncertainties India faces [5].
• Classic Indian example: handloom weavers displaced by powerlooms.
• Natural rate of unemployment = frictional + structural. Because technological unemployment is a form of structural unemployment, faster automation can raise the natural rate.

Mains Points

• AI and India's services model: India's IT-BPM growth was built on labour arbitrage (cheap skilled labour). AI can automate routine coding and back-office tasks, so it threatens the middle- and low-wage jobs this model created [4]. India needs a policy mix that pushes AI to augment workers rather than replace them, along with social protection for those who are displaced [5].
• Match the cure to the cause: technological unemployment is a supply-side, long-term problem. The answer is Skill India-type reskilling, apprenticeships, labour mobility and income support, not a demand stimulus. A low headline unemployment rate of 3.2% (PLFS 2023-24) [1] can hide this, because displaced workers often move into low-paid informal work instead of showing up as unemployed.
• Gandhi's concern, revisited: in a labour-rich country with a youth unemployment rate of 10.2% (2023-24) [2], technology that saves labour carries a real social cost. The debate is how to gain productivity from AI while keeping growth rich in jobs. India's young, tech-savvy workforce is both its biggest exposure and its biggest advantage [4].

Related concepts

• Voluntary unemployment
• Involuntary unemployment
• Frictional unemployment
• Structural unemployment
• Cyclical unemployment
• Natural rate of unemployment
• Hysteresis
• Okun's law
• Efficiency wage
• Labour arbitrage

Read more

• Unemployment in macroeconomic theory