Gini Coefficient

One number between 0 and 1 can make Sweden, India, and South Africa look comparable, while hiding whether inequality comes from wages, assets, inheritance, land, tax design, or survey method. The Gini coefficient is a compression machine: it turns a whole income distribution into the area between two curves. That compression is useful, but it is also the trap.

How it works

The Gini starts with the Lorenz curve, introduced by Max O. Lorenz in 1905. Put every household in order from poorest to richest. The x-axis is the cumulative share of people. The y-axis is the cumulative share of income or consumption. Perfect equality is a 45-degree line: the bottom 40% receive 40% of income.

The Gini is the area between that equality line and the Lorenz curve, divided by the full area under the equality line. In plain terms:

Gini = 0 means everyone has the same income.

Gini = 1 means one person has all income and everyone else has zero.

Most official datasets report the index from 0 to 100. A World Bank Gini of 35 is a coefficient of 0.35.

The useful thing is not precision. The useful thing is discipline. A country can grow GDP per person while the Lorenz curve bends further away from equality. The Gini makes that bend visible.

What the number hides

A Gini of 0.35 can describe several societies that feel nothing alike.

Same Gini, different machine	What can be happening underneath
Wage inequality	Engineers, doctors, and financiers pull away from median workers
Asset inequality	Housing, equity, and business ownership compound faster than wages
Inheritance	Position transfers through families rather than paychecks
Tax design	Market inequality is high, disposable-income inequality is lower
Survey design	Consumption data and income data produce different readings

This is why the Gini is a map symbol, not the territory. A tax-and-transfer state can have high market-income inequality and a lower disposable-income Gini. A country with weak income records may use consumption surveys instead. A country with hidden wealth may look cleaner than it is.

The Lorenz curve carries more information than the coefficient. Two curves can cross and still produce the same Gini. When that happens, the single number cannot tell whether the pain sits near the bottom, the middle, or the top.

Where it shows up

The World Bank’s SI.POV.GINI indicator tracks national Gini indexes, but years differ because surveys do not arrive on the same clock. As of the World Bank series available in 2026, Sweden sits near the high-20s to low-30s in recent observations, India sits in the low-30s in recent consumption-based readings, and South Africa sits above 60 in its latest World Bank observation. Those three numbers tell a real story. They do not tell the whole story.

South Africa’s number carries the afterlife of land, race, job access, and capital ownership. Sweden’s number carries taxes, transfers, unions, and asset-price pressure. India’s number is harder to read because consumption surveys, informal income, urban property, and household structure all bend the signal. The same instrument is measuring three different social machines.

A sharp framing line: the Gini is best at saying “look here,” not “the explanation is here.”

What's contested

The first dispute is measurement. Income Gini, consumption Gini, wealth Gini, pre-tax Gini, and post-tax Gini answer different questions. Comparing them as if they are one object creates fake certainty.

The second dispute is interpretation. A rising Gini can come from broad poverty reduction plus faster gains at the top, or from stagnation at the bottom. The number alone cannot tell whether the society is getting richer in an unequal way or merely sorting pain into one part of the distribution.

The third dispute is moral weight. A 0.40 Gini is not automatically “bad” without knowing mobility, public goods, median income, housing costs, health access, and political capture. The coefficient measures dispersion. It does not measure dignity.

Why this crosses realms

The Gini belongs next to concept power law because inequality is often not a bell curve problem. Wealth, city size, attention, firm value, and creator income frequently have heavy tails. Averages behave badly when the tail owns the room.

It also belongs near concept fermi paradox. Both are lessons in compression. The Drake equation compresses unknown cosmic terms into one expected number. The Gini compresses a national distribution into one coefficient. In both cases, the number is less useful than the argument over which term is doing the work.

An open question

If a society lowers its Gini through transfers while asset ownership keeps concentrating, has inequality fallen, or has only the visible income layer been cooled? The next page worth writing is concept wealth inequality, because income is only the annual shadow of ownership.

Key sources

Max O. Lorenz, “Methods of Measuring the Concentration of Wealth” (1905), Publications of the American Statistical Association - the curve the Gini later measures against.
Corrado Gini, Variabilità e mutabilità (1912) - the original statistical concentration measure.
Anthony B. Atkinson, “On the Measurement of Inequality” (1970), Journal of Economic Theory - why inequality measures embed social choices.
World Bank, SI.POV.GINI, accessed 2026-06-24 - national Gini index series with country-year gaps.
Our World in Data, “Income Inequality” - useful visual context for comparing inequality datasets and definitions.

Abhishek's take

What grabs me about the Gini is how polite it makes conflict look. One decimal can hide housing, inheritance, wages, credit access, and state capacity inside the same clean figure. I trust it as an alarm, not as a diagnosis.

Tags: #inequality #measurement #economics #statistics #public-policy

Abhishek