Two questions hiding in one word
Across this rung we asked where a single person's pay comes from. We traced it to the value a worker adds, to their human capital, to the supply of and demand for their particular skill. Each story explained one wage. But step back and a different question appears: not "why does this person earn that?" but "how is the whole pie split across everyone?" That is the question of inequality, and it is the hardest one in the rung — partly because the facts are slippery, and partly because reasonable people genuinely disagree about what to do.
Before any number, fix one distinction that trips up almost every public argument: income versus wealth. Income is a *flow* — money arriving over time, like salary, interest, or a pension, measured per year. Wealth is a *stock* — what you own at a moment, the value of your house, savings, and shares minus your debts. The gap matters because the two are far from the same. A young doctor can have a high income yet near-zero or negative wealth after student loans; a retired farmer can have modest income but own land worth a fortune. This is the income versus wealth distinction, and wealth is almost always distributed *more* unequally than income.
Drawing the distribution: the Lorenz curve
Suppose we want to *see* how unequal a country is, not just feel it. The classic picture is the Lorenz curve, and the recipe is simple. Line everyone up from poorest to richest. Then plot, for the bottom 10%, 20%, 30%, and so on, the share of *total* income they collectively receive. The horizontal axis is the cumulative share of *people*; the vertical axis is the cumulative share of *income*. Both run from 0% to 100%.
Now the trick that makes the curve readable. Imagine a society of *perfect* equality, where everyone earns exactly the same. Then the bottom 20% of people earn 20% of income, the bottom 50% earn 50%, and so on — every point sits on the 45-degree diagonal. That diagonal is the line of perfect equality, our benchmark. Any real society's Lorenz curve sags *below* that line: because the poorest are lined up first and they earn less than their head-count share, the curve starts shallow and only catches up at the far right, where the rich finally get counted. The deeper the sag, the more unequal the society. A curve that hugged the bottom and right edges would be near-total inequality — one person with everything.
Cumulative income share, bottom X% of people
Bottom % Perfect equality A real-ish country
-------- ---------------- ------------------
20% 20% 7%
40% 40% 18%
60% 60% 34%
80% 80% 57%
100% 100% 100%
The right-hand column always lies below the middle one:
that downward sag, the gap from the 45-degree line,
IS the inequality the Lorenz curve makes visible.Squeezing the picture into one number: the Gini
A picture is hard to put in a headline, so economists boil the Lorenz curve down to one figure: the Gini coefficient. The idea is geometric and intuitive. Look at the crescent-shaped area trapped *between* the line of perfect equality and the actual Lorenz curve. The Gini is that gap area divided by the whole triangle beneath the diagonal. If the curve sits right on the diagonal, the gap is zero and the Gini is 0 — perfect equality. If one person held everything, the gap fills the whole triangle and the Gini is 1 — perfect inequality. Every real country lands in between.
To anchor the scale: most rich countries sit somewhere from about 0.25 (very equal, several Nordic states after taxes) to roughly 0.40 (the United States). Some highly unequal countries reach the high 0.50s. A jump of "just" 0.05 sounds tiny but represents a large shift in who holds what. The Gini's great virtue is comparability — one number lets you rank countries or watch a country change over time. That is also its trap, which we will face squarely in a moment.
What one number cannot say
The Gini's compression is also its blindness. Because it summarises the *whole* curve into one figure, very different societies can share the same Gini. One country might be unequal because its poorest are desperately poor; another might have a comfortable bottom but a runaway top 1%. The Gini cannot tell them apart, yet the two call for completely different responses. This is why serious analysts pair it with other tools: the share of income held by the top 1% or 10%, ratios like the 90/10 (the income at the 90th percentile divided by that at the 10th), and direct poverty measures that count how many fall below a fixed line.
Two deeper limits matter even more. First, a *snapshot* hides *mobility*. The same households are not stuck in the bottom forever — students, the newly arrived, and the retired drift through low-income years that they climb out of, so a frozen photo can overstate how permanent the ranking is. A society with high yearly inequality but easy movement between rungs is not the same as one where birth fixes your station. Second, the numbers describe *money*, not lives: they say nothing about whether the gaps were earned by effort, inherited, or won by luck — and that *moral* question, not the measurement, is what most arguments are really about.
Where the gaps come from
If wages flow from the value a worker adds, then anything that lifts some people's productivity above others' will spread incomes apart. The most discussed engine is skill-biased technical change: new technology — computers, then software, now automation — multiplies what a *skilled* worker can produce while doing little for, or even replacing, routine work. That raises the relative demand for human capital and pulls the educated and the rest apart. Globalisation pushes the same way, by exposing routine domestic jobs to lower-wage competition abroad. Neither is a villain with intent; both are the supply-and-demand logic of this rung, applied to different kinds of labour.
But labour-market forces are only half the tale, and they fade as we move from income to wealth. Returns to capital — rent, dividends, capital gains — flow to whoever already *owns* assets, which is why the split of income between labour and capital shifts the picture even when no wage changes. Layer on inheritance: wealth passed across generations gives some a head start that has nothing to do with their own productivity, and a slice of top incomes is really economic rent — payment above what was needed to call forth the work, captured through scarcity, market power, or position rather than added value. Add market power in pay-setting at the very top, and discrimination that depresses pay for some groups, and you have a tangle of causes that no single policy lever can address.
The equity–efficiency trade-off, honestly
Whether to shrink these gaps is the central fight, and it turns on the equity–efficiency trade-off — the tension between efficiency and equity that has shadowed this whole subject. The case for redistribution is partly moral (a fairer split, a floor below which nobody falls) and partly economic: an extra hundred dollars plainly does more for someone with little than for someone with millions, the diminishing marginal benefit you met early in the consumer track. The classic worry on the other side is that taxing high earners and topping up low ones blunts incentives — to work, to study, to take risks — and so quietly shrinks the very pie we are dividing.
The honest position is that this trade-off is real but its *size* is genuinely contested — one of the live disputes of the field, not a settled number. Some redistribution may cost almost nothing, or even help, if it pays for things that raise productivity: childhood nutrition, schooling, health, the human capital of people who could not otherwise afford it. Other redistribution may bite hard on effort. Which is which depends on the exact tax, the exact transfer, and behaviour economists measure imperfectly. Beware the metaphor of "a leaky bucket" carrying water from rich to poor: it is a vivid warning that some is spilled, but it quietly assumes the bucket only ever loses — and sometimes the same dollar, spent on a child's schooling, comes back larger.
So where does that leave you? With tools, not a verdict. Economics can draw the Lorenz curve, compute the Gini, trace the causes, and estimate — within wide error bars — what a given policy would do to the pie and to its slices. What it *cannot* do is tell you how much equality is worth, or whose claim to a dollar is more just. Those are value judgements, the normative questions a free society settles through politics, not through a model. The discipline's job, done well, is to make the trade-offs visible and the false certainties impossible — to ensure that when we argue, we argue about real magnitudes and honest values, not about made-up facts.