Skip to content
Back to search results

Lorenz Curve

Introduced by
Max Otto Lorenz
Year
1905
Type
Graphical representation of distributional inequality
Family
Inequality measurement
Font size

What does it do?

The Lorenz curve is a graphical device that displays the full shape of inequality in a distribution by plotting the cumulative share of a quantity (such as income) held by the cumulative share of the population, ranked from poorest to richest.

Introduced by Max Lorenz in 1905, it underlies the Gini coefficient and provides the basis for ranking distributions by inequality when one curve lies entirely above another.

The intuition behind the method

Line everyone up from the lowest income to the highest, then ask: what share of total income do the poorest 10 percent hold? the poorest 20 percent? and so on, up to 100 percent. Plotting these cumulative answers traces a curve that bows below the diagonal. If income were perfectly equal, the poorest x percent would hold exactly x percent of income and the curve would be the straight 45-degree line. The more income is concentrated at the top, the deeper the curve sags away from that line. The picture of the sag is the inequality.

History

Max Otto Lorenz, then a young American economist, introduced the curve in his 1905 paper in Publications of the American Statistical Association as a graphical method for comparing the concentration of wealth. Corrado Gini built his 1912 concentration coefficient directly on the area between the curve and the equality line. Anthony Atkinson's 1970 paper established the deep link between Lorenz dominance and social-welfare rankings, showing that Lorenz-dominance is equivalent to agreement among all inequality-averse welfare functions, cementing the curve's central place in inequality theory.

When to use it

Use the Lorenz curve to visualize the complete inequality structure of a distribution, to compare distributions when one may dominate another, and as the foundation for computing the Gini coefficient and related measures. It is invaluable for seeing where in the distribution inequality is concentrated rather than collapsing it to one number. It is descriptive and graphical, so when distributions' curves cross it cannot by itself rank them, and it conveys relative shares only — two populations with identical Lorenz curves can have vastly different absolute living standards. It is defined for nonnegative quantities; negative incomes pull the curve below the axis and complicate interpretation.

Limitations and considerations

  • Cannot rank distributions whose curves cross; the picture alone is then indecisive.
  • Shows relative shares only, so it ignores absolute levels and overall growth.
  • Reading precise quantities off the curve is imprecise compared with summary indices.
  • Sensitive to negative or zero values, which distort the convex shape.

Source conditions do not replace independent scientific review or validation by an analysis engine.

Strengths

  • Displays the entire inequality structure, showing exactly where shares are concentrated.
  • Provides an unambiguous inequality ranking when one curve dominates another (Lorenz dominance).
  • Forms the geometric basis of the Gini coefficient and the Atkinson and generalized-entropy measures.
  • Intuitive and visually compelling for communicating inequality to non-specialists.

Applications

  • Visualizing income and wealth inequality within and across countries.
  • Comparing pre- and post-tax-and-transfer income distributions to show redistribution.
  • Assessing concentration in land ownership, firm size, or market share.
  • Constructing concentration curves to study the distribution of public spending or health outcomes by income rank.
  • Teaching the foundations of inequality measurement and the Gini coefficient.

How it works

How it works

1

Units are sorted in ascending order of the quantity of interest, so that the poorest unit comes first and the richest last. This ordering is what makes the resulting curve convex and bounded below the diagonal.

Formula
x_{(1)} \le x_{(2)} \le \dots \le x_{(n)}\ \text{(ascending order)}
2

For each cumulative population fraction p (e.g., the bottom k of n units), the curve plots L(p), the cumulative share of total income held by that bottom fraction. The horizontal axis is the cumulative population share and the vertical axis the cumulative income share, both running from 0 to 1.

Formula
p_k = \frac{k}{n},\qquad L(p_k) = \frac{\sum_{i=1}^{k} x_{(i)}}{\sum_{i=1}^{n} x_{(i)}}
3

The diagonal L(p) = p represents perfect equality, where every population fraction holds an equal income fraction. Because units are ordered from poorest to richest, the Lorenz curve always lies on or below this line; its endpoints are fixed at (0,0) and (1,1).

Formula
L(p) = p\ \text{(line of perfect equality)};\quad L(p) \le p\ \text{always}
4

When one distribution's Lorenz curve lies everywhere on or above another's, it Lorenz-dominates the second and is unambiguously more equal — a ranking on which all standard inequality measures and a wide class of social-welfare functions agree (Atkinson's theorem). When two Lorenz curves cross, no such unambiguous ranking exists and different inequality indices may disagree.

Formula
L_1(p) \ge L_2(p)\ \forall p \Rightarrow \text{distribution 1 Lorenz-dominates 2}

Common pitfalls

  • Concluding that one distribution is more equal when the two Lorenz curves actually cross.
  • Forgetting that the curve depicts relative shares, so a poorer society can have a higher (more equal) curve than a richer one.
  • Comparing curves built on different income concepts or units without noting the difference.
  • Treating sample Lorenz curves from small samples as exact, ignoring sampling variability at the tails.

Frequently asked questions

How is the Lorenz curve related to the Gini coefficient?

The Gini coefficient is exactly twice the area between the Lorenz curve and the line of perfect equality. A curve hugging the diagonal (little area) gives a Gini near zero, and a deeply bowed curve gives a Gini near one. The Lorenz curve thus contains all the information the Gini summarizes plus the shape detail the single number discards.

What does it mean when two Lorenz curves cross?

Crossing means one distribution is more equal in one part of the range (say, the bottom) but less equal in another (say, the top). In that case there is no unambiguous inequality ranking: different inequality indices, which weight parts of the distribution differently, can legitimately disagree about which distribution is more unequal. Only when one curve lies everywhere above the other (Lorenz dominance) do all standard measures agree.

Can the Lorenz curve compare living standards between countries?

Only their relative inequality, not their absolute prosperity. Because the curve plots shares, it is scale-invariant: a poor and a rich country with identically shaped distributions have the same Lorenz curve. To compare living standards you need absolute figures (mean income, generalized Lorenz curves that scale by the mean), not the ordinary Lorenz curve alone.

Related methods

Sources

  1. Lorenz, M. O. (1905). Methods of measuring the concentration of wealth. Publications of the American Statistical Association, 9(70), 209–219.

    Open DOI reference
    Copied.
  2. Atkinson, A. B. (1970). On the measurement of inequality. Journal of Economic Theory, 2(3), 244–263.

    Open DOI reference
    Copied.