Named economic laws, paradoxes and fallacies

Detail

A. What a "law", a "paradox" and a "fallacy" mean

• Economic law: a pattern seen again and again in the economy. It is a tendency, not a fixed rule like a law of physics. It holds "other things being equal" (ceteris paribus).
• Paradox: a result that goes against common sense but is true. Example: using coal more efficiently led to more coal being burnt.
• Fallacy: a belief that sounds right but is wrong. Example: "a broken window is good for the economy."
• Why UPSC asks about them: each one names a policy trap. Examples: gaming targets (Goodhart), rising costs of health and education (Baumol), and failed energy savings (Jevons).

B. Baumol's cost disease (William Baumol and William Bowen, 1966)

• Definition: in some sectors productivity barely rises. Their wages still rise because they must keep pace with wages in high-productivity sectors. So their costs and prices rise steadily.
• Classic example (the string quartet):
• A Beethoven quartet needs 4 players and about 40 minutes, the same as in 1800.
• So output per musician (productivity) has not changed.

• Musicians' pay has still risen. If it had not, they would leave for factory or IT jobs, where productivity and pay have grown.

• Worked example with numbers:

• Manufacturing: output per worker doubles, from 100 to 200 units. Wages double, from ₹20,000 to ₹40,000. Cost per unit stays ₹200.
• Teaching: one teacher still teaches 40 students. The teacher's wage also rises to ₹40,000 to stay competitive. Cost per student rises from ₹500 to ₹1,000.

• Result: goods get cheaper relative to services, and services get more expensive relative to goods.

• Sectors affected: labour-intensive services such as health, education, government administration, the performing arts, and care work.

• Evidence: an IMF working paper (2015) on public education spending found that Baumol's effect does push costs up, but by less than the theory predicts [7].
• Relevance for India:
• India's growth is services-led. Services rose from 50.6% of GVA in FY14 to 55.3% in FY25 (Economic Survey 2024-25) [5]. GVA (gross value added) means the value of output minus the value of inputs used.
• The share reached 56.4% as per the FY26 First Advance Estimates [6].

• In a services-heavy economy, public health and education budgets must keep rising just to deliver the same level of service.

• The "cure" debate: AI, telemedicine and digital classrooms could raise output per worker in services. If they do, the "disease" weakens.

• Related point: it partly explains Wagner's law (see the grid below), because government services keep getting costlier.

C. Jevons paradox (W.S. Jevons, The Coal Question, 1865)

• Observation: James Watt's steam engines used coal far more efficiently. Britain's total coal use still went up.
• Mechanism, step by step:
• Better efficiency means less coal per unit of work.
• So the effective cost of each unit of work falls.
• Cheaper work leads to new uses and more users.

• Total coal use rises.

• Rebound effect: part of the expected saving is "eaten back" by extra use.

• Direct rebound: the same person uses more of the same service, e.g. drives more in an efficient car.
• Indirect rebound: the money saved is spent on other energy-using goods.

• Backfire (the true Jevons paradox): the rebound is more than 100%, so total use actually rises.

• Worked example:

• A car improves from 10 km/litre to 20 km/litre. Expected result: fuel use halves, from 100 litres to 50 litres a month.
• The owner then drives 1,500 km instead of 1,000 km. Fuel used = 1,500 ÷ 20 = 75 litres.
• Expected saving = 50 litres. Actual saving = 25 litres. Rebound = 25 ÷ 50 = 50%.

• If the owner drove more than 2,000 km, fuel use would go above 100 litres. That is backfire.

• Indian example: UJALA (Unnat Jyoti by Affordable LEDs for All):

• Launched in January 2015.
• More than 36 crore LED bulbs distributed so far [3].

• Jevons risk: when lighting is cheap, households keep lights on longer and add more fittings. Actual savings can then be smaller than the engineering estimates.

• Other examples: cheaper AI compute raises total compute demand and data-centre electricity use. Efficient ACs can lead to more AC use.

• Policy lesson: efficiency on its own may not cut total use. Pair it with carbon pricing, taxes or caps.
• Trap: this is the same W.S. Jevons who was one of the founders of marginalism in 1871 (Section 3 of this topic).

D. Goodhart's law (Charles Goodhart, 1975)

• Statement: "When a measure becomes a target, it ceases to be a good measure."
• Origin: the Bank of England targeted money-supply numbers. Once these became targets, the link between money supply and inflation broke down.
• Why it happens: people change their behaviour to hit the number, not to reach the real goal behind it.
• Indian and global examples:
• GDP and state rankings: states or officials may push up the measured indicator without real improvement.
• World Bank Doing Business report:
  • Data irregularities were first reported inside the World Bank in June 2020. The report was paused and investigated.
  • It was discontinued in September 2021 [2].
  • It has been replaced by Business Ready (B-READY), first released in 2024 [2].
  • B-READY rates economies on three pillars: Regulatory Framework, Public Services and Operational Efficiency [8].
  • Lesson: when rankings are targeted this hard, the ranking itself gets corrupted.
• NPA targets and evergreening:
  • An NPA (non-performing asset) is a loan whose interest or principal has been unpaid for more than 90 days.
  • Suppose banks are judged on keeping NPAs low.
  • They may give a fresh loan to a weak borrower so the borrower can repay the old loan. This is evergreening.
  • Reported NPAs look low, but the hidden bad debt keeps growing.

E. Campbell's law (Donald Campbell, 1976)

• Statement: the more a number is used to make social decisions, the more it is corrupted. It also distorts the social process it was meant to measure.
• Sibling of Goodhart's law: Goodhart came from monetary economics. Campbell came from social science and education.
• Examples:
• Teaching to the test: schools drill for exam scores, so real learning suffers.
• Crime statistics that fall because cases are not registered.
• Hospitals judged on death rates that avoid taking serious patients.

F. Parkinson's law (C. Northcote Parkinson, 1955, The Economist)

• Statement: work expands to fill the time available for its completion.
• Bureaucracy version: the number of staff grows by itself, whatever the actual workload.
• Officials want more subordinates, not rivals.

• Officials create work for each other.

• Link to India: the motto "minimum government, maximum governance". Related steps include cutting old laws, merging departments, and digital delivery that cuts paperwork.

G. Broken window fallacy (Frédéric Bastiat, "What Is Seen and What Is Not Seen", 1850)

• Story: a boy breaks a shopkeeper's window. People say this is good, because the glazier (glass-fitter) now gets work.
• Bastiat's answer:
• Seen: the glazier earns ₹6,000.
• Not seen: the shopkeeper would have spent that ₹6,000 on new shoes. So the shoemaker loses the sale.

• Net result: society has one window, the same as before, but no new shoes. Society is poorer by one pair of shoes.

• Meaning: destruction is not a net gain. It only moves spending around and uses up wealth.

• GDP link:
• GDP is a flow measure: it counts production during a period.
• So rebuilding after a flood or cyclone adds to GDP.
• The houses, roads and assets destroyed (the stock of wealth) are not subtracted.

• So disaster years can show a misleading GDP "boost".

• Wider lesson: always ask about the opportunity cost (the next best use of the same money). The same logic applies to wars, "make-work" schemes and scrappage subsidies.

H. Lump of labour fallacy

• Definition: the false belief that an economy has a fixed amount of work. On this view, more workers or more machines must take jobs away from others.
• Why it is a fallacy:
• New workers also earn and spend, which creates new demand.

• Machines raise productivity and lower prices. Lower prices raise real incomes, and new industries and jobs appear.

• Where it appears:

• Automation and AI: fear of permanent mass joblessness.
• Migrants: the belief that they "steal" local jobs.
• Retirement age: the belief that older workers who stay on block young people's jobs.

• Work-sharing proposals: shorter weeks to "spread" jobs.

• Caution for Mains: the total number of jobs is not fixed. But individual workers can still lose out during the transition. The answer is reskilling and safety nets, not blocking technology.

I. Say's law (J.B. Say, 1803) — recap

• Statement: "supply creates its own demand."
• Producing goods gives people incomes.

• Those incomes are spent, so there can be no general glut (too many unsold goods across the whole economy).

• Classical view: savings become investment through the interest rate, so full employment returns by itself.

• Keynes (General Theory, 1936) rejected it:
• Demand can fall short of supply.
• People may hoard money.

• Then output and employment fall. This is the effective demand idea (Section 4).

• Link: the paradox of thrift (see the grid) follows from Keynes's rejection of Say's law.

J. Recap grid of laws homed in other topics

Law / idea	One-line statement	Home topic
Gresham's law	Bad money drives good money out of circulation	Money
Engel's law	As income rises, the share spent on food falls	Demand
Veblen goods	Demand rises with price because of status value	Demand
Wagner's law	Public spending grows faster than national income as economies develop	Budget
Okun's law	Unemployment rises about 1 point for every ~2% shortfall of GDP below potential	Employment
Phillips curve	Short-run inverse link between inflation and unemployment	Inflation
Laffer curve	Beyond some rate, higher tax rates reduce revenue	Taxation
Kuznets curve	Inequality first rises, then falls with development (inverted U)	Poverty
Paradox of thrift	More saving by all can reduce total income and saving	Income determination
Dutch disease	A resource boom raises the currency and hurts other exports and manufacturing	BoP

• Okun's law formula: change in unemployment rate ≈ −½ × (GDP gap %).
• The GDP gap (output gap) = (actual GDP − potential GDP) ÷ potential GDP. It is negative when the economy produces less than it could.
• Example: GDP is 4% below potential → gap = −4% → unemployment rises by about 2 percentage points.

Prelims Hooks

• Jevons paradox: efficiency gains raise total resource use. It comes from The Coal Question (1865), by the same Jevons who was a founder of marginalism.
• Baumol's cost disease (Baumol-Bowen, 1966): wages rise in low-productivity services, so those services grow costlier. The string quartet is the classic example.
• Goodhart's law (1975) came from Bank of England monetary targeting. Campbell's law (1976) is the social-science version ("teaching to the test").
• Doing Business was discontinued in September 2021 after data irregularities. It was replaced by B-READY (first report 2024), which has three pillars: Regulatory Framework, Public Services, Operational Efficiency [2][8].
• Parkinson's law (1955): work expands to fill the time available. Trap: it is not about prices or money.
• Broken window fallacy is Bastiat (1850), "What Is Seen and What Is Not Seen". It is not Say or Mill.
• Say's law: "supply creates its own demand". Keynes rejected it in 1936.
• Lump of labour is a fallacy: the amount of work in an economy is not fixed.
• UJALA (launched January 2015) has distributed more than 36 crore LED bulbs. It is a textbook setting for the rebound effect [3].
• Services share of GVA: 55.3% (FY25) [5], and 56.4% (FY26 First Advance Estimates) [6].

Mains Points

• Efficiency is not enough for climate goals (GS-III):
• Missions such as UJALA save energy per unit, but the Jevons rebound can wipe out part of the saving [3].

• So pair efficiency with carbon pricing, the carbon credit trading scheme, and demand-side caps.

• Targets vs outcomes in governance (GS-II/III):

• Goodhart's and Campbell's laws explain several failures: gamed rankings (Doing Business, discontinued 2021 [2]), NPA evergreening, and schools "teaching to the test".

• Remedies: use many indicators, have them audited by independent bodies, and measure outcomes rather than outputs.

• Services-led growth and fiscal pressure (GS-III):

• Services are 55-56% of GVA [5][6].
• Baumol's cost disease means health and education costs will keep rising, and Wagner's law pressure on budgets will grow with them.

• Productivity tools (digital public infrastructure, AI in services) are the long-run cure.

• Automation and jobs (GS-III):

• Arguments to ban AI or migration often rest on the lump of labour fallacy.
• A better answer is to reskill workers, give social security to gig workers, and support the transition, not to fix the number of jobs.