Microeconomics › Section O · Poverty and economic inequality › Topic O.4

Measuring income inequality and its causes

Poverty counts the people below a line; inequality describes how the whole income pie is shared. Economists measure that sharing with quintiles, the Lorenz curve and the Gini coefficient, and then use supply and demand to ask why US inequality has risen since about 1980.

Efficiency and equity

Earlier topics judged markets by efficiency: whether they make total surplus as large as possible. Equity asks a different question: whether the way that surplus is shared is fair. A competitive market can be perfectly efficient and still leave some people very poor, because wages reflect the value of what people produce, not what they need. Economics can measure the distribution and trace why it changes; deciding how much inequality is too much involves value judgements it cannot settle alone.

Quintiles: slicing the pie

The most common method ranks all households from lowest to highest income and splits them into five equal groups, called quintiles. Then you ask what share of total income each quintile receives. If incomes were equal, each would get 20%. In Census Bureau data for 2023, the top fifth of households received a little over half of all money income, while the bottom fifth received about 3%. Around 1980 the top fifth's share was about 44%. Much of the rise went to the very top: the top 5% alone received about 16% of income around 1980 and about 23% in 2023.

The Lorenz curve

A Lorenz curve turns quintile shares into one picture. The horizontal axis shows the cumulative share of households, from poorest to richest; the vertical axis shows the cumulative share of income they receive. To plot it, add the shares up: the bottom 40% receives the first plus the second quintile's share, and so on, until 100% of households receive 100% of income. Perfect equality is the 45-degree line, where the bottom 20% gets 20% and the bottom 60% gets 60%. The further the curve sags below that line, the more unequal the distribution.

Worked example: from shares to a Gini

A hypothetical country's quintiles receive 4%, 9%, 15%, 23% and 49% of income. 1. Cumulative shares: 4%, 13%, 28%, 51%, 100%. These are the Lorenz points at 20%, 40%, 60%, 80% and 100% of households. 2. Area under the curve, using trapezoids of width 0.2:

B = 0.2 \times \frac{(0 + 4) + (4 + 13) + (13 + 28) + (28 + 51) + (51 + 100)}{2 \times 100} = 0.292

3. The area under the equality line is 0.5, so the Gini coefficient is \(G = 1 - 2B = 1 - 0.584 = 0.416\). 4. Check: equal 20% shares give \(B = 0.5\) and \(G = 0\). Quintile data hide inequality within each fifth, so this slightly understates the true Gini.

The Gini coefficient

The Gini coefficient condenses the Lorenz curve into one number: the area between the equality line and the curve, divided by the whole area under the equality line. A Gini of 0 means everyone has the same income; a Gini of 1 means one household has everything. Comparisons need care about which income is measured. The Census Bureau's Gini for US pre-tax money income was about 0.485 in 2023. After taxes and transfers, recent OECD figures put the US at about 0.38 to 0.40, near the high end for wealthy countries, against roughly 0.26 to 0.28 in the Nordic countries.

Income, wealth and snapshots

Income is a flow received over a period, such as wages, interest and transfers in a year. Wealth is a stock: the value of all assets, such as savings, shares, pensions and a home, minus debts. Because income differences accumulate, wealth is far more unequal: Federal Reserve data for 2024 show the top 1% of households holding about 30% of all household wealth. A one-year snapshot also exaggerates lifetime inequality. Most people earn little when young, most in middle age and less after retiring, so a society of mixed ages would show some inequality even if everyone's lifetime earnings were identical.

Cause one: changing households

Part of the rise in household inequality comes from how households are formed. Far more married women now work for pay than in 1970, so more households have two earners. It is also more common for high earners to marry each other, which economists call assortative mating: two doctors rather than a doctor and a lower-paid partner. At the other end, single-parent households, which usually have one modest income, have also become more common. More high-earning couples at the top and more one-earner families at the bottom widen the gap between households even if no one's wage changes.

Cause two: a shift in wages

Since about 1980, pay gaps between workers have widened, especially between workers with and without a college degree. Supply and demand explain why. The supply of college graduates has grown strongly, which by itself would push their relative wage down. Yet their wage premium rose. The only way both can happen is if demand for skilled workers grew even faster. The main suspects are information technology, which raises the productivity of skilled workers most and replaces many routine middle-skill tasks, and globalisation, which widens markets for skilled services while exposing less-skilled workers to more competition.

Worked example: the race between education and technology

In a hypothetical market for college-educated labour, demand is \(Q_d = 100 - 2W\) and supply is \(Q_s = 3W - 50\), with \(W\) the hourly wage and quantities in thousands of workers. 1. Start: \(100 - 2W = 3W - 50\) gives \(W = \$30\) and \(Q = 40\). 2. More graduates shift supply to \(Q_s = 3W - 40\). Alone, this gives \(W = 140 \div 5 = \$28\): the wage falls. 3. New technology also shifts demand to \(Q_d = 125 - 2W\). Now \(125 - 2W = 3W - 40\) gives \(W = \$33\) and \(Q = 59\). 4. Read it: employment and the wage both rose, so demand outran supply.

Other forces at work

Several other changes push in the same direction, and economists debate their weights. Union membership has fallen, removing a force that compressed wages. The federal minimum wage, $7.25 an hour since 2009, has lost much of its real value. Winner-take-all markets let the best performers, from software firms to entertainers, sell to global audiences, so small differences in talent bring huge differences in pay. Market power also matters: firms with little competition earn profits that flow to owners and top managers. Changes in top tax rates affect how much of these gains people keep.

Common mistakes to avoid

First, confusing inequality with poverty: if every income doubled, shares and the Gini would not change, but poverty would fall. Second, adding up quintile shares wrongly when plotting a Lorenz curve: the points are cumulative, ending at 100%. Third, comparing Gini figures measured differently: pre-tax money income gives a higher Gini than income after taxes and transfers. Fourth, mixing up income and wealth: wealth is a stock and far more unequal. Fifth, concluding that a rising college premium means fewer graduates: supply rose, but demand rose faster.

Key takeaways

Quintile shares, the Lorenz curve and the Gini coefficient measure how income is shared; a Gini of 0 is perfect equality and 1 is total concentration. In the worked example, shares of 4, 9, 15, 23 and 49% give a Gini of about 0.42. The US Census Gini for pre-tax money income was about 0.485 in 2023, and the top fifth's share has risen from about 44% around 1980 to more than half. Two main causes are changing households and a demand for skills that has outpaced supply, helped by weaker unions, a lower real minimum wage and winner-take-all markets.

Flashcards

Flashcard

Efficiency vs equity

Size of the surplus vs fairness of how it is shared

Efficient markets can still leave people poor.

Test your understanding

Question 1 of 5

Quintile shares are 5%, 10%, 15%, 25% and 45%. What are the Lorenz curve points for the bottom 40% and bottom 80%?

Question 2 of 5

What does a Gini coefficient of 0 mean?

Question 3 of 5

If every household's income doubled overnight, what would happen to the Gini coefficient?

Question 4 of 5

The supply of US college graduates has grown strongly, yet the college wage premium rose. What best explains this?

Question 5 of 5

Why does a one-year snapshot of income overstate lifetime inequality?

← Previous topicNext topic →