Unemployment in macroeconomic theory

Detail

1. The basic measure

• Labour force: people who are working plus people who are not working but are looking for work or available for it.
• Unemployment rate (UR):
• UR = (Unemployed ÷ Labour force) × 100.

• Worked example: labour force = 50 crore, unemployed = 1.6 crore → UR = 1.6 ÷ 50 × 100 = 3.2%.

• Latest official data:

• The UR under usual status (ps+ss) was 3.2% for persons aged 15 years and above in PLFS 2023-24 (July 2023–June 2024). Usual status means the person's main activity over the last 365 days, plus any side activity [2].
• Male UR fell from 3.3% to 3.2%. Female UR rose from 2.9% to 3.2% (2022-23 → 2023-24) [2].
• Youth UR (15-29 years) was 10.2% (2023-24), about three times the overall rate [3].
• The overall UR (15+) fell from 6.0% (2017-18) to 3.2% (2023-24) [4].

• Net new EPFO subscriptions rose from 61 lakh (FY19) to 131 lakh (FY24). This is a sign that more jobs are becoming formal [4].

• What the low number hides: open UR in India is low because most poor people cannot afford to stay unemployed. They take any work, however poorly paid. So a low UR does not mean good jobs.

2. Taxonomy of unemployment

2.1 Voluntary unemployment

• Definition: people choose not to work at the current wage, or they wait for a better job.
• Example: a graduate who spends years preparing for a government job and turns down private-sector work.
• Classical economists said all unemployment is either voluntary or frictional. Wages would fall until everyone who wanted a job got one.

2.2 Involuntary unemployment

• Definition (Keynes): people are willing to work at the current wage, but no jobs are available.
• Why it happens: in a slump, firms sell less, so they hire fewer workers. Wages do not fall fast enough to clear the labour market, because wages are "sticky".
• Example: factory workers laid off in a recession.
• This idea underlies the Keynesian case for government action to raise demand.

2.3 Frictional unemployment

• Definition: short spells without work while people search for or switch jobs.
• It exists even at full employment, because matching workers to jobs always takes some time.
• Cure: better job information.
• The National Career Service (NCS) portal of the Ministry of Labour & Employment was launched in July 2015 [5].
• By 20 November 2025 it had 6.02 crore jobseekers, 54.27 lakh employers and 8.17 crore vacancies mobilised since launch. All its services are free [5].
• Vacancies on NCS rose from about 13 lakh (2021-22) to 35.7 lakh (2022-23) [6].

2.4 Structural unemployment

• Definition: a long-term mismatch between the skills or location of workers and the jobs available. It is caused by changes in technology or in the structure of the economy.
• Indian examples:
• Ahmedabad textile mill workers lost jobs when the mills closed in the 1980s.

• Handloom weavers lost work when powerlooms replaced them.

• Cure: reskilling, apprenticeships and help to move to new locations. Demand stimulus alone does not fix it.

2.5 Cyclical unemployment

• Definition: unemployment caused by deficient aggregate demand (total spending in the economy is too low) during a recession.
• It falls when the economy recovers.
• Link to the deflationary gap: when planned spending is below the full-employment level of output, output and jobs fall (see income-determination-keynes).
• Cure: fiscal stimulus (more government spending or tax cuts) or monetary stimulus (repo rate cuts).
• Stimulus → more spending → firms produce more → firms hire more workers.

2.6 Technological unemployment

• Definition: workers are replaced by machines, automation or AI.
• Gandhi's epigraph in the NCERT chapter reflects his fear that machines would displace labour in a labour-rich country.
• Economic Survey 2024-25:
• It says 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 [7].
• Its chapter on the AI ecosystem says AI may complement (work alongside) labour in the near term, while firms learn to use it. But the productivity gain from this has a ceiling. It calls AI and labour one of the biggest uncertainties India faces [8].
• India's advantage, according to the Survey, is its young, tech-savvy workforce, which can use AI to augment (add to) its productivity [7].

Quick comparison table

Type	Main cause	Duration	Right cure
Frictional	Job search	Short	Information (NCS)
Structural	Skill or location mismatch	Long	Reskilling
Cyclical	Low aggregate demand	Lasts as long as the recession	Fiscal/monetary stimulus
Technological	Machines, AI	Long	Reskilling, social protection
Voluntary	Worker's choice	Varies	Not a policy problem in itself
Involuntary	No jobs at current wage	Varies	Demand management

3. Natural rate of unemployment and full employment

• Natural rate of unemployment = frictional + structural unemployment.
• It is the rate that is consistent with full employment. In the long run, the economy tends to move back to it. This is the view of Friedman and Phelps, who developed it in the late 1960s.
• NAIRU (non-accelerating-inflation rate of unemployment) is the UR at which inflation stays stable:
• If UR is pushed below NAIRU → labour is scarce → wages rise → inflation speeds up.

• If UR is above NAIRU → inflation slows.

• Key point: full employment ≠ zero unemployment. Frictional and structural unemployment remain even at full employment.

• Worked example: frictional = 2%, structural = 3% → natural rate = 5%. If actual UR = 7%, then cyclical unemployment = 7 − 5 = 2%.

4. Dynamics and linked ideas

4.1 Hysteresis

• Definition: high unemployment persists even after the recession that caused it has ended.
• How it happens:
• The long-term unemployed lose skills and job contacts.
• Employers see them as less employable.

• Cyclical unemployment turns into structural unemployment → the natural rate rises.

• Example: the post-COVID "scarring" debate, about whether the pandemic caused lasting damage to jobs and output.

4.2 Okun's law

• Definition (Arthur Okun, 1962): when unemployment rises, output falls below its potential (the maximum output the economy can sustain without rising inflation) by a larger proportion.
• Rule of thumb: each 1 percentage point of UR above the natural rate goes with an output gap of about 2-3%.
• Formula: Output gap (%) ≈ −β × (UR − natural rate), with β ≈ 2 to 3.
• Worked example: natural rate = 5%, actual UR = 7%, β = 2.5 → output gap ≈ −2.5 × 2 = −5%. So actual output is about 5% below potential.
• Evidence:
• An IMF study, Okun's Law: Fit at 50? (2013), found the law strong and stable in most advanced economies, including during the Great Recession. It estimated a coefficient of about −0.4: each 1% output gap moves UR by about 0.4 percentage point. That matches the 2-3% rule of thumb above [9].

• Later IMF research (2021) found that UR responds less to output swings in developing economies than in advanced ones [10].

• Why it fits India poorly:

• A slowdown pushes people into low-productivity self-employment, farm work or casual work, not into open unemployment.
• So output falls, but the measured UR barely moves.
• An example is the move of workers back to agriculture after COVID-19.

4.3 Efficiency wage

• Definition: a wage set above the market-clearing level (the wage at which labour supply equals labour demand). Firms pay it to raise effort and loyalty and to reduce turnover (workers leaving).
• Examples:
• Henry Ford's $5 day (1914): Ford roughly doubled wages, and quit rates fell sharply.

• Shapiro-Stiglitz shirking model: a high wage makes losing your job costly, so workers do not shirk (avoid work).

• Consequence: wages do not fall to clear the market → more people want jobs than there are jobs → involuntary unemployment.

• Indian angle: the large gap between government pay and private pay helps explain the long "waiting" for government jobs, which is a form of voluntary unemployment.

4.4 Labour arbitrage

• Definition: moving work to places where labour is cheaper, for example through offshoring.
• India's IT-BPM sector grew on this cost gap with the US and Europe.
• Risks now:
• Reshoring: rich countries move work back home.
• AI: routine coding and back-office tasks can be automated, which hits exactly the middle- and lower-wage jobs that the Economic Survey flags [7].

5. Conclusion: how far these categories fit India

• These textbook categories fit India only partly.
• Three forms matter more than open unemployment:
• Disguised unemployment: more people work on a job, often a family farm, than are needed. Their marginal productivity is close to zero.
• Seasonal unemployment: no work in the months between farm seasons.

• Underemployment: people work fewer hours or at lower productivity than they want or could.

• Even a UR as low as 3.2% (2023-24) [2] sits alongside a youth UR of 10.2% [3] and a large informal workforce.

• The core problem is the quality of jobs, not just their number.

Prelims Hooks

• Natural rate of unemployment = frictional + structural. It excludes cyclical unemployment.
• Full employment ≠ zero unemployment. Frictional unemployment exists even at full employment.
• Involuntary unemployment is Keynes's idea: people are willing to work at the current wage but find no job.
• NAIRU: if UR is pushed below it, inflation speeds up. The idea is linked to Friedman-Phelps.
• Okun's law (1962): 1 percentage point of excess UR goes with about a 2-3% output gap. It fits developing economies like India poorly.
• Efficiency wage: a wage above market-clearing that causes involuntary unemployment. Examples: Ford's $5 day (1914) and the Shapiro-Stiglitz model.
• Hysteresis: a recession leaves unemployment high, which raises the natural rate.
• PLFS 2023-24: UR (usual status, 15+) was 3.2%; youth UR (15-29) was 10.2%; PLFS is conducted by NSO under MoSPI [2][3].
• National Career Service portal: launched July 2015 by the Ministry of Labour & Employment. It tackles frictional unemployment, not structural [5].
• Trap: reskilling cures structural unemployment, while a fiscal stimulus cures cyclical unemployment. Do not mix them up.

Mains Points

• Low UR, weak job quality: UR fell from 6.0% (2017-18) to 3.2% (2023-24) [4]. But Okun's law barely works in India, because shocks push people into low-productivity informal work instead of open unemployment. Policy should therefore track underemployment, earnings and formalisation (such as EPFO additions [4]), not just the UR.
• Match the cure to the type: demand stimulus helps with cyclical unemployment. Structural and technological unemployment need Skill India-type reskilling, apprenticeships and labour mobility. Frictional unemployment needs job-matching platforms such as NCS [5]. A single tool cannot fix all of them.
• AI and labour arbitrage: India's IT-BPM model was built on cheap skilled labour. Reshoring and AI now threaten mid- and low-wage tasks [7]. India needs policy that uses AI to raise worker productivity rather than simply replace workers, along with social protection for displaced workers [8].
• Hysteresis and the post-COVID recovery: long spells without work cause lasting "scarring", especially for youth, whose UR was 10.2% in 2023-24 [3]. This supports quick counter-cyclical support such as MGNREGA and early-career employment programmes.