What inequality measures can—and cannot—tell us
A single number can describe a distribution, but it cannot explain the lives, policies, and trade-offs behind that distribution. Reading inequality means knowing what is being measured first.
“Inequality is rising” sounds like a complete statement, but it leaves several questions unanswered. Inequality of what? Among whom? Before or after taxes and transfers? Measured at one moment or across a lifetime?
Start with the object
Income is a flow: wages, pensions, profits, and transfers received over a period. Wealth is a stock: assets minus debts at a point in time. Consumption describes what households use. Wages describe payment for work. Poverty measures compare resources with a threshold. These measures overlap, but none can stand in for all the others.
A country can have a large wealth gap and a smaller disposable-income gap. A household can have low income this year but substantial assets. A person can earn a high salary in a city where housing costs absorb much of it. The number changes when the object changes.
What the Gini does
The Gini coefficient summarises how evenly a distribution is shared. A lower value indicates a more equal distribution; a higher value indicates a more unequal one. It is useful because it compresses a complex distribution into a comparable indicator.
It is limited for the same reason. Two societies can have the same Gini while placing the difference in different parts of the population. The coefficient does not show whether inequality is concentrated at the top, whether poverty is deep, or how much mobility people experience over time.
Definitions change the story
Disposable income after taxes and transfers is not the same as market income before them. Equivalised household income adjusts for household size. A relative poverty line moves with the distribution; an absolute threshold does not. Regional data can reveal differences that a national average hides.
The OECD’s income database is valuable partly because it documents these choices. A responsible comparison reads the metadata before reading the ranking. Otherwise a chart can create the appearance of disagreement when researchers are actually measuring different things.
From measurement to policy
Statistics do not decide what counts as fair. They help identify where a policy may be working, who is missing from an average, and which trade-offs deserve scrutiny. The next question is causal: did wages change, did taxes change, did household composition change, or did the data capture a different population?
Good inequality reporting keeps both parts visible. It gives the number enough context to be useful, and it gives the people behind the number enough dignity not to become a decorative trend line.
Sources & methodology
The sources below anchor the explanation. They are starting points for verification, not decoration.
- 01 OECD — Income Distribution Database
Comparable data on income inequality and poverty, including definitions and methodology.
- 02 World Inequality Database
Research database and documentation for the distribution of income and wealth.
- 03 OECD — Distributional national accounts
Context for connecting household-level distributions with national accounts.
