Gross Domestic Product
The number every argument about the economy starts from was built in three years by one borrowed statistician, who set out its limits in chapter one of his own report.

A report of several hundred pages reached the Senate on 4 January 1934, answering a resolution the chamber had passed two years earlier. Its central finding sat in a table near the front: the income paid out to Americans had fallen by forty percent between 1929 and 1932. Nobody in Washington had known that figure. The Hoover administration had governed the worst collapse in the country’s history on freight-car loadings, steel tonnage and share prices, none of which tells you how much poorer a nation has become.
The Senate had asked for the count in Resolution 220 of the Seventy-second Congress, and the Commerce Department gave the work to its Division of Economic Research, which borrowed a man from the National Bureau of Economic Research in New York. He was Simon Kuznets, thirty-two years old, who had come out of Pinsk to the United States in 1922 and taken a Columbia doctorate four years later under Wesley Mitchell, the Bureau’s founder. The acknowledgment page is unusually blunt about how much of it was his: he was retained to plan and supervise the study, was in full charge of it, and wrote the text. What he assembled is what the newspapers now report every quarter as gross domestic product.
The tables did what no amount of argument had managed. Net income paid out to individuals had shrunk by forty percent in three years, and Kuznets set that beside the only earlier slump covered by a reliable long series, Willford King’s estimates back to 1909, where the fall from 1920 to 1921 came to 14.4 percent. Business saving told the same story from the other side, positive by 1.9 billion dollars in 1929 and by 1932 a loss of 9.5 billion. A government that had been guessing now had the size of the hole in dollars.
Chapter one is the part of the report that has outlasted its tables. There Kuznets spent several pages on what his own measure could not see, and named the gaps one at a time instead of hedging in general. Work that never reaches a market is missing by construction, and he picked housework as the example. Prices depend on who holds the income, so the same service is dearer in an unequal country and arrives in the totals looking like more product. Nothing in the accounts registers what the earning cost the earner.
“A student of social affairs who is interested in the total productivity of the nation, including those efforts which, like housewives’ services, do not appear on the market, can therefore use our measures only with some qualifications. …
Economic welfare cannot be adequately measured unless the personal distribution of income is known. And no income measurement undertakes to estimate the reverse side of income, that is, the intensity and unpleasantness of effort going into the earning of income. The welfare of a nation can, therefore, scarcely be inferred from a measurement of national income as defined above.
The abuses of national income estimates arise largely from a failure to take into account the precise definition of income and the methods of its evaluation which the estimator assumes in arriving at his final figures. … As a result we find all too commonly such inferences that a decline of 30 percent in the national income (in terms of ‘constant’ dollars) means a 30 percent decline in the total productivity of the nation, and a corresponding decline in its welfare. … Such statements can obviously be true only when qualified by a host of ‘ifs.’”
National Income, 1929–32, Senate Document No. 124, 73rd Congress, 2nd Session (1934), pp. 6–7.
The accounts earned their standing in the war. Robert Nathan, who had worked on the 1934 estimates as a junior analyst, was running the Planning Committee of the War Production Board by 1942, and he brought Kuznets in to head its analysis and research section. In August of that year Kuznets delivered a feasibility study on the munitions program the President had approved: the targets were larger than the economy could carry, and chasing all of them at once would end in fewer finished weapons than a smaller program. General Brehon Somervell, the Army’s chief supply officer, asked the board to reject the study and suppress it. The board sided with Kuznets, the schedule was cut back to what the numbers allowed, and a question the Army had been answering by exhortation became one answered by arithmetic.

Afterwards the method travelled into the United Nations’ System of National Accounts and from there into nearly every finance ministry on earth. The United States kept gross national product, which counts output by Americans wherever in the world they are, as its headline measure until 1991, when the Bureau of Economic Analysis switched to gross domestic product, which counts output inside American borders. The reasons were housekeeping: the international standard had gone that way, and the domestic version lines up with the employment and productivity series beside it. At a press conference on 7 December 1999 the Commerce Department named the national income and product accounts its achievement of the century.
Which leaves the thing worth carrying away. A release carries several growth rates for the same three months, and a headline quotes one. In the second quarter of 2026 the Bureau reported real GDP up 1.5 percent at an annual rate and current-dollar GDP up 8.0 percent. The distance between those is prices: the index for gross domestic purchases rose 5.8 percent over the same quarter, so most of the eight is the dollar shrinking rather than the country producing. The release also carries gross domestic income, which adds up the identical activity from the earnings side and came in at 2.2 percent, and the Bureau now publishes the average of the two measures, 1.8 percent, because two sides of an identity that must balance in principle routinely refuse to in practice.
| Real GDP, output with price change stripped out | +1.5% |
| Current-dollar GDP, the same output at the prices paid | +8.0% |
| Gross domestic purchases price index, the gap between them | +5.8% |
| Real gross domestic income, the same activity counted as earnings | +2.2% |
United States, second quarter of 2026, seasonally adjusted annual rates, from the Bureau of Economic Analysis second estimate of 26 August 2026. Every one of these is a true statement about April, May and June.
Thirty-four years later the same objection came back in a form more people remember. Robert Kennedy, three days into his campaign for the presidency, stood at the University of Kansas on 18 March 1968 and read out what the gross national product counted: cigarette advertising, ambulances to clear the highways of carnage, special locks for the doors and jails for the people who broke them. Then he turned to what it left out, the health of children and the quality of their education and the joy of their play, and closed on a line that has outlived the speech around it. It measures everything, he said, “except that which makes life worthwhile.”

Kennedy was making a political argument and Kuznets a technical one, and underneath they are the same argument. Kuznets took the Nobel in 1971 for the study of economic growth his own measurement had made possible, and the warning stayed where he put it, ahead of the tables, legible to anyone who opened the book. The measure does what it claims: it adds up what the market paid for, in dollars, inside a border, over a span of months. Every interesting question about whether a country is doing well starts one step past that, and the man who built it said so in the same report.