Inequality · US · 1948–2025
Who Gets the Growth?
Aggregate output has grown without interruption for seventy years. Its distribution has not kept pace.
For much of the twentieth century, economics carried an implicit assurance: that inequality might widen while a country was still developing, but that sustained growth would eventually narrow it again. Simon Kuznets described the relationship as an inverted arc, inequality rising and then falling as national income climbed. The proposition was influential, and its corollary was reassuring. Expand the aggregate, and its distribution would resolve itself.
Subsequent data did not support it. When later researchers assembled the long time series Kuznets never had, the arc did not appear; the mid-century compression he observed is now attributed largely to the Depression, two world wars, and steeply progressive taxation eroding concentrated wealth, rather than to any self-correcting property of growth. The five measures below trace what the distribution of American income has actually done since then.
Productivity and pay diverged after 1979.
Cumulative growth since 1979
Net productivity vs. real hourly pay of production and nonsupervisory workers (roughly 80% of the workforce), cumulative from 1979. Source: Economic Policy Institute.
From the postwar years through the late 1970s, productivity and compensation advanced in step: as output per hour rose, so did the typical paycheck. The two series separated around 1979. Net productivity has since grown by roughly 90%. The pay of a typical worker has grown by about 33%, close to one-third as fast.
The top one percent's income share passed the bottom half's in 1997.
Share of pre-tax national income
Pre-tax national income, equal-split adults. Hover for any year. Source: World Inequality Database (variable sptinc992j).
In 1980, the highest-earning one percent received about a tenth of national income and the bottom half received roughly twice that. The two shares converged through the 1980s and, in 1997, crossed: for the first time the top one percent took a larger share than the entire bottom half. The gap has since widened. The top one percent now receives about 21% of pre-tax national income; the bottom fifty percent, about 13%.
The cost of essential goods outpaced earnings.
Price growth, indexed to 1980 = 100
Each series indexed to its 1980 level. Hover for any year. Source: BLS CPI components and median weekly earnings, via FRED.
Even the pay that workers did receive purchased less of what matters most. Since 1980, college tuition has risen roughly thirteenfold and medical care about eightfold. Nominal median earnings rose about fivefold, barely ahead of overall inflation, which means the typical real wage sits close to where it stood four decades ago while the price of a degree or a hospital stay moved well beyond it.
Growth per person did not slow.
Real GDP per capita, chained 2017 dollars
Output per person, chained 2017 dollars. Hover for any year. Source: BEA via FRED (A939RX0Q048SBEA).
Real output per person more than quadrupled over the same period, from about $15,000 in 1948 to over $70,000 today, rising through every decade in which pay stalled and shares diverged. The preceding patterns are therefore not attributable to a shortage of growth. The output was generated; it was not broadly distributed.
Measured against the economy's own growth, most of the distribution fell behind.
Real income growth by group, 1980–2014, minus the +61% economy-wide average
Each group's cumulative real pre-tax income growth minus the population-wide average of +61%. A positive bar outgrew the economy; a negative bar fell behind it. The top 1% is the apex slice of the top 10%. Source: Piketty, Saez & Zucman (2018), Table II.
A cleaner way to see the distribution is to subtract the economy from it. Between 1980 and 2014, average real income per adult grew about 61%. Charted as the deviation from that average, the picture is stark: the bottom half grew some sixty points below the economy as a whole, close to flat in absolute terms, while the top one percent grew roughly one hundred and forty points above it. The gains did not merely favor the top; for most of the distribution, income growth ran behind the growth of the country that produced it.
Which side captures the surplus is not, in the first instance, a question of who earned it. It is a question of who can decline the terms and walk away.
Introductory economics establishes that the burden of a tax settles not on whoever is legally charged with it, but on whichever side of the market is least able to move. The same principle operates on the gains from growth. Capital has become comparatively mobile, able to relocate, automate, or substitute one input for another; most labor has not. When the two negotiate over a rising surplus, the more mobile party captures the larger share, largely independent of who performed the work. This is a mechanism rather than a moral claim, which is why it has proven difficult to argue away.
Intergenerational mobility declined as the gains concentrated.
The distributional shift carries a generational consequence. Of children born in 1940, about ninety percent went on to earn more than their parents; among those born in the 1980s, about half did. When the authors of that finding modeled the remedy, the result was pointed: applying the 1940 distribution to today's growth recovers mobility to 80%, whereas restoring postwar growth rates to today's distribution recovers it only to 62%. Distribution, not the rate of growth, accounts for most of the decline.
A demonstration
In the agent-based model known as Sugarscape, identical agents gather a scattered resource under identical rules, with no differences in ability and randomized starting positions. Within a few hundred iterations, a small minority holds nearly everything. Extreme concentration emerges from compounding alone, from the fact that having more assists in acquiring more, absent any difference in effort or talent.
A precedent
The board game Monopoly began in 1903 as The Landlord's Game, designed by Elizabeth Magie to illustrate the harm of concentrated ownership. It shipped with two rule sets: one in which all players prospered as property changed hands, and one in which a single player accumulated everything. When Parker Brothers acquired it decades later, only the winner-take-all rules were kept.
Inequality as design, not necessity.
The economist Kate Raworth characterizes inequality as a design failure rather than an economic law, and the distinction organizes everything above. An economy optimized for the size of its output, on the assumption that distribution follows, will not reliably produce a broadly shared result; the preceding five decades are the evidence. Growth is genuine, but it is not self-distributing. Restoring broad mobility is a matter of designing for it deliberately, as the postwar arrangements briefly did.
Do not wait for economic growth to reduce inequality, because it will not. Instead, create an economy that is distributive by design.Kate Raworth · Doughnut Economics