🔍 Read the full analysis: What Measuring AI Can—and Cannot—Show on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
A November 11 report revisits how economist Simon Kuznets’ 1934 national-income estimates measured the scale of the Great Depression while warning that income alone could not show a nation’s welfare. The article applies that limitation to unpaid household work and free digital services, including AI, whose consumer value may exceed the revenue they generate. The figures show what economic measures capture—and where their boundaries remain.
A November 11 report from ThorstenMeyerAI.com revisits the origins and limits of national economic measures, arguing that GDP records market production but does not measure overall welfare. The distinction matters as unpaid household work and free digital services—including AI chatbots—can generate substantial value without appearing fully in the market transactions GDP counts.
The article traces national income accounting to a request from the U.S. Senate in 1932, after the Depression left policymakers without comprehensive estimates of the economy’s decline. Economist Simon Kuznets, working with the National Bureau of Economic Research, led the analysis. A report delivered to the Senate on Jan. 4, 1934, estimated that income produced fell from $83 billion in 1929 to $39.4 billion in 1932—a 53% decline. Income paid to individuals fell 40% over the same period.
The article emphasizes that Kuznets cautioned against treating national income as a direct measure of social well-being. His estimates focused mainly on economic activity whose results appeared in markets, leaving out much unpaid household work. They also did not show how income was distributed or how much effort and hardship went into earning it. Kuznets wrote that “the welfare of a nation can, therefore, scarcely be inferred from a measurement of national income.”
The piece connects that warning to two present-day measurement gaps. It cites a German estimate that unpaid household work was worth €987 billion in 2013 when valued like paid work, equivalent to 39% of the value added counted in GDP. It also cites a 2017 experiment in which the median participant would have required $17,530 to give up search engines for a year, despite paying no direct price to use them. A Stanford working paper, the article says, estimates $172 billion a year in consumer value from AI chatbots in the United States; that is an estimate of value, not chatbot revenue.
Infographic 01 · The sector clock, by output
China: how GDP changed shape
Share of GDP (value added at current prices) by sector, 1978 to 2025. In 2023 farming still held 22.8% of all jobs but produced 7.1% of GDP.
- Farming27.7%
- Industry47.7%
- Services24.6%
Reform starts
- Farming14.7%
- Industry45.5%
- Services39.8%
- Farming9.1%
- Industry45.4%
- Services45.5%
Services overtake industry
- Farming6.7%
- Industry35.6%
- Services57.7%
Communiqué 2025
Measured by output, not by jobs: the sector clocks of earlier episodes count workers. Jobs vs. output for farming, 2023: 22.8% of employment (year-end) vs. 7.1% of GDP.
Sources: NBS China, China Statistical Yearbook 2024, tables 3-2 and 4-2; NBS, Statistical Communiqué 2025 (28 Feb 2026).
Powered by Thorsten Meyer AI
GDP’s Market Boundary
GDP remains useful because it offers a standardized way to track production and compare economic activity over time. During the Depression, even an imperfect national-income estimate gave lawmakers a clearer account of the scale of the contraction than scattered indicators could provide. Today, governments and analysts use GDP to describe output, follow changes across sectors and inform economic decisions.
But the measure’s usefulness depends on understanding what it counts. Market output is not the same as total value or well-being: unpaid care can support families and the broader economy without being sold, while a free service can benefit users without generating an equivalent payment. This distinction becomes more visible as digital tools and AI provide services at little or no direct cost. The figures cited in the article suggest a gap between consumer benefit and recorded sales, but do not by themselves establish how much these tools improve welfare across society.
The article also uses China’s changing output mix to show what GDP can track well. Services accounted for 24.6% of Chinese value added in 1978 and 57.7% in 2025, according to the figures presented. Yet farming still employed a much larger share of workers than its share of output in 2023. Output shares describe production, not how many people work in each sector; a separate measure is needed to explain that difference.
Products Worth Considering
LG UltraGear 34G600A-B 34-Inch Curved WQHD Gaming Monitor, 160Hz
Bose QuietComfort Ultra Wireless Over-Ear Headphones with Spatial Audio and Noise Cancellation, Black
LG UltraGear 32-inch Curved QHD Gaming Monitor (2560×1440, 180Hz, 1ms, 1000R, HDR10)
As an affiliate, we earn on qualifying purchases.
From Depression Accounts to GDP
The 1934 report measured national income, rather than today’s headline GDP. The article describes later changes in accounting: the United States developed gross national product estimates in the 1940s as it planned wartime production; the first international System of National Accounts appeared in 1953; and the United States adopted GDP as its headline measure in 1991, in line with international standards.
GDP is generally defined as the market value of final goods and services produced within a country during a given period. The Bureau of Economic Analysis presents it as the sum of consumption, investment, government spending and net exports. The report gives 2025 output figures of about $30.86 trillion for the United States, €4.53 trillion for Germany and 140.2 trillion yuan for China, attributing them to the respective statistical agencies. These totals describe economic output; they do not, on their own, indicate how income is distributed or how people experience economic conditions.
The article’s central point is not that GDP is useless, but that it answers a narrower question than “How well off are people?” Kuznets himself spelled out that limitation alongside the estimates that helped establish national accounting.
“The estimates “cover primarily only efforts whose results appear on the market place of our economy”.”
— Simon Kuznets, quoted in the 1934 report
What the Estimates Leave Open
The article’s examples do not establish a single monetary value for everything GDP excludes. The household-work figure is a valuation based on treating unpaid tasks like wage-paid work; results can depend on the method used. The search-engine figure comes from a 2017 experiment and records what participants said they would need to give up access, not a market price or a measure of GDP itself.
The cited $172 billion annual AI estimate is described as coming from a Stanford working paper. The source material does not provide the paper’s methods, publication status or uncertainty range, so the estimate should not be treated as a settled measure of realized economic benefit. It also does not specify a direct comparison between that consumer-value figure and total AI-industry revenue. More broadly, the article does not quantify how much unpaid work or AI-related value national accounts miss today, or how any proposed adjustment would affect official GDP figures.
Further Measures Beyond Output
The article presents no new official accounting standard or policy change. Its immediate implication is that readers should distinguish GDP from measures designed to describe household labor, income distribution, employment or consumer benefit. Those indicators can add information, but they answer different questions and rely on their own definitions and data.
The report points toward productivity as the next topic in its series, using the gap between workers’ share and output share across sectors as a starting point. For the AI valuation estimate, a fuller assessment would require examination of the working paper’s methodology, comparison period and treatment of consumer value. Until those details are established, the estimate is best read as one effort to measure benefits that market revenue alone may not capture.
Key Questions
What did the 1934 national-income report find?
It estimated that U.S. income produced fell from $83 billion in 1929 to $39.4 billion in 1932, a 53% decline. Income paid to individuals fell 40% over that period.
Why did Simon Kuznets warn against using national income as a welfare measure?
The estimates mainly covered activity recorded through markets. They did not fully count unpaid household work, show how income was distributed or measure the effort involved in earning it.
Does unpaid household work count in GDP?
Most unpaid household services, such as cooking and caregiving for one’s own family, are not recorded as market production in standard GDP accounts. The article cites a German estimate valuing this work at €987 billion in 2013.
What does the $172 billion AI figure represent?
It is an annual estimate of value to U.S. consumers from AI chatbots, attributed in the article to a Stanford working paper. It is not presented as chatbot revenue, and the source material does not provide enough methodological detail to assess the estimate fully.
Does GDP show how many people work in each sector?
No. GDP measures the value of output, while employment statistics count workers or jobs. The article notes that farming accounted for a larger share of China’s workers than of its GDP in 2023.
Source: ThorstenMeyerAI.com
