Inequality and growth: Kuznets to Piketty, and India's contested evidence

Detail

1. Basic tools: how inequality is measured

• Gini coefficient: one number that shows how unequally income or spending is shared.
• 0 = everyone has the same amount (perfect equality).
• 1 = one person has everything (perfect inequality).

• It is often written on a 0–100 scale. So 0.255 is the same as 25.5.

• Lorenz curve: a graph that plots the share of total income (or spending) held by the bottom x% of people.

• If income is shared equally, the curve is a straight 45° line, the line of equality.
• Formula: Gini = A / (A + B). Here A is the area between the line of equality and the Lorenz curve, and B is the area below the Lorenz curve.

• Worked example: if A = 0.13 and A + B = 0.5, then Gini = 0.13 / 0.5 = 0.26. This is close to India's rural consumption Gini.

• Consumption Gini vs income Gini:

• A consumption Gini is built from what households spend. India's household surveys (HCES) measure spending.
• An income Gini is built from what households earn. Most rich countries use this.
• Spending is always more equal than earning, so the two are not comparable.

2. Kuznets curve (Simon Kuznets, 1955)

• Idea: an inverted-U curve. As per capita income (average income per person) rises, income inequality first rises, reaches a peak, then falls.
• Rising phase (early development):
• Workers leave agriculture, where incomes are low and fairly equal.
• They move into industry, where incomes are higher but more unequal.

• The gap between the two sectors pushes overall inequality up.

• Falling phase (mature development):

• Urbanisation (more people living in towns and cities) and mass education raise the skills of ordinary workers.
• Political pressure from workers and voters leads to redistribution: taxes, welfare and labour laws.

• The gaps narrow.

• The mechanism in plain words: workers shift from low-productivity sectors to high-productivity sectors. This first raises overall inequality a lot, and only later lowers it [6].

• Policy meaning: "grow first, and equality will come later." This idea supported the trickle-down view.

3. Challenges to Kuznets

• East Asia's "growth with equity" (South Korea, Taiwan):
• Both countries grew fast with no rising phase of inequality.
• Reasons: early land reform (land spread among small farmers) and broad education.

• Several East Asian economies went from low to middle incomes while reducing inequality [6].

• Rich countries since 1980: inequality has risen again. This is the opposite of what the falling part of the curve predicts.

• Weak data support:
• The curve seen across countries depends heavily on a few Latin American countries, which are both middle-income and very unequal.

• The rising part of the curve has disappeared in recent decades [6].

• Lesson: inequality depends on the type of growth and on institutions (land, education, taxes). It does not depend only on the stage of growth [6].

4. Piketty: Capital in the Twenty-First Century (2013 French edition / 2014 English edition)

• Core inequality: r > g
• r = the return on capital: profit, rent, interest and dividends earned on wealth each year.

• g = the growth rate of the whole economy (total income).

• Chain of cause and effect:

• When r > g, the owners of wealth earn more each year than the economy grows.
• They reinvest part of that return, so inherited wealth grows faster than wages and total output.

• Wealth therefore concentrates in a few families. Piketty calls this "patrimonial capitalism".

• Worked example:

• Suppose r = 5% and g = 1.5%.
• Over 20 years, reinvested wealth grows 1.05²⁰ ≈ 2.65 times. The economy grows 1.015²⁰ ≈ 1.35 times.

• So wealth roughly doubles relative to the economy.

• Piketty's remedy: a progressive global wealth tax and higher top income tax rates.

• Link to Kuznets: Piketty argues that the fall in inequality in the mid-20th century came from wars, inflation and high taxes. It was not an automatic stage of growth.

5. Elephant curve (Lakner–Milanovic, 2016; data 1988–2008)

• What it is: a growth incidence curve. It shows how much the real income of each percentile of the world's population grew between 1988 and 2008 [4].
• A percentile is 1% of the world's people, ranked from poorest to richest.

• The curve is "anonymous": it compares positions in the distribution, not the same people over time.

• Data: household surveys from about 120 countries. Each country was split into deciles (tenths of its population). Incomes were adjusted for inflation and for price differences using 2005 PPP dollars [4][5].

• PPP, or purchasing power parity, is an exchange rate that equalises what money can buy in different countries.

• The working paper came out in 2013 (World Bank Policy Research Working Paper 6719) [5]. The journal version followed in 2016.

• Shape:

• Hump (the elephant's back): big gains for the emerging middle classes of Asia, mainly China and India.
• Dip (the base of the trunk, around the 75th–90th percentile): little or no gain for the lower-middle classes of rich countries.

• Raised trunk tip: big gains for the global top 1%.

• Meaning: globalisation reduced inequality between countries but raised it within many countries.

6. Trickle-down vs redistribution

• Trickle-down: growth alone is enough. Jobs and higher wages slowly reach the poor.
• Redistribution: the state must deliberately move resources to the poor through taxes, transfers, land reform, and public health and education.
• Growth elasticity of poverty (GEP): the % fall in poverty for each 1% rise in mean income.
• Formula: GEP = (% change in poverty rate) / (% change in mean income). It is usually a negative number.
• Worked example: mean income rises 10%, and the poverty rate falls from 20% to 16%. That is a 20% fall in poverty, so GEP = −20 / 10 = −2.
• GEP is higher (in absolute size) when initial inequality is low. The poor then hold a bigger share of each extra rupee, so the same growth lifts more people out of poverty.
• This is the link between inequality and poverty policy: lower inequality makes growth more pro-poor.

7. India's evidence splits by data source

(a) Consumption surveys show falling inequality

• HCES (Household Consumption Expenditure Survey) Gini:
• rural 0.266 → 0.237
• urban 0.314 → 0.284
• both between 2022-23 and 2023-24

• Consumption inequality fell in most major states [2].

• World Bank consumption Gini: 28.8 (2011-12) → 25.5 (2022-23) [2][3].

• "Fourth most equal country" (PIB, July 2025):
• India ranks after the Slovak Republic, Slovenia and Belarus [2].
• China's Gini is 35.7 and the USA's is 41.8 [2].

• India is more equal than every G7 and G20 country on this measure [2].

• Poverty context:

• Extreme poverty (below $2.15 a day) fell from 16.2% (2011-12) to 2.3% (2022-23).

• This lifted about 171 million people above the line [3].

• World Bank's own warning: inequality "may be understated due to data limitations" [3]. Also, survey method changes in 2022-23 make comparisons over time harder [3].

(b) Income and wealth data show rising concentration

• World Inequality Lab (WIL): Bharti, Chancel, Piketty and Somanchi (2024), "Billionaire Raj". In 2022-23:
• the top 1% held about 22.6% of national income
• the top 1% held about 40.1% of national wealth

• this is the highest income share since 1922, higher than under the British Raj.

• World Inequality Database income Gini: 52 (2005) → 61 (2023) (World Bank brief, citing WID) [3].

• Wage gaps: the median earnings of the top 10% were 13 times those of the bottom 10% in 2023-24 [3].

(c) Why the two stories differ

• Surveys under-capture the rich:
• Rich households often refuse to take part or under-report.

• So the top of the distribution is "missing" from the survey.

• Consumption is smoother than income:

• Families spread spending over good and bad years.

• The rich save a large part of their income, so their spending understates their income.

• Different methods:

• WIL combines income-tax data, national accounts and rich lists (for example, billionaire lists) to fill in the top of the distribution.

• Surveys do not do this.

• Comparison trap: India's consumption Gini (25.5) cannot fairly be ranked against other countries' income Ginis, such as the USA's 41.8.

• Wealth vs income: wealth builds up over generations (Piketty's r > g), so it is always far more concentrated than income. Compare 40.1% of wealth with 22.6% of income for India's top 1%.

(d) Spatial inequality (between regions)

• State per capita income (Class 10, Development):
• Haryana ₹3,25,759 vs Bihar ₹60,337 (2023-24).

• Ratio: 3,25,759 ÷ 60,337 ≈ 5.4.

• Multidimensional poverty also varies widely: below 1% in Kerala vs about 35% in Bihar [3].

• Multidimensional poverty looks at deprivation in health, education and living standards, not just income.

• Rural-urban gap: the gap in MPCE (monthly per capita consumption expenditure) is narrowing (Section 4).

• Newer figures are in the World Inequality Report 2026 (verify current).

Prelims Hooks

• Kuznets curve (1955) has an inverted-U shape. Inequality first rises, then falls as per capita income rises. Trap: it is not U-shaped.
• Gini = A / (A + B) on the Lorenz curve. 0 means perfect equality and 1 means perfect inequality. A lower Gini means more equal.
• Piketty's r > g: the return on capital is greater than the growth rate of the economy, so wealth concentrates. Book: Capital in the Twenty-First Century (2013/2014).
• Elephant curve is by Lakner–Milanovic, using data from 1988–2008. Winners: Asia's middle class and the global top 1%. Losers: rich-country lower-middle classes.
• Growth elasticity of poverty is the % fall in poverty per 1% rise in mean income. It is higher when initial inequality is low.
• India's consumption Gini (World Bank): 28.8 (2011-12) → 25.5 (2022-23). India is the "4th most equal", after the Slovak Republic, Slovenia and Belarus [2].
• HCES Gini 2022-23 → 2023-24: rural 0.266 → 0.237 and urban 0.314 → 0.284 [2].
• WIL "Billionaire Raj" (2024): the top 1% held 22.6% of income and 40.1% of wealth (2022-23). This is the highest income share since 1922.
• Trap: India's official Gini is consumption-based, not income-based. So it cannot be directly compared with the USA's 41.8 income Gini.
• Extreme poverty ($2.15 a day): 16.2% (2011-12) → 2.3% (2022-23) [3].

Mains Points

• "Most equal" vs "Billionaire Raj":
• Both claims are "correct" within their own data.
• Consumption surveys miss the rich and smooth out income differences. Tax and wealth data capture the rich.

• Policy needs both lenses. India also needs better survey coverage of the rich and an income distribution survey, with the World Bank's "understated" caveat noted [3].

• Kuznets is not destiny:

• East Asia shows that land reform and mass education can give growth without a rise in inequality.

• For India this means human capital and asset access (land, credit) are the levers, rather than waiting for trickle-down (GS-III inclusive growth).

• Inequality reduces the poverty payoff of growth:

• A lower Gini raises the growth elasticity of poverty.

• So redistribution through DBT, public health and education, and progressive taxes helps both equity and poverty reduction. It is not a pure trade-off with growth.

• Spatial and wealth concentration is a federal and fiscal issue:

• Haryana's per capita income is about 5.4 times Bihar's (2023-24), and multidimensional poverty is below 1% in Kerala vs 35% in Bihar [3].
• This supports equalising transfers by the Finance Commission.
• Piketty's r > g feeds the debate on wealth or inheritance taxes.