🔍 Read the full analysis: AI's New Normal: 80% In Services on ThorstenMeyerAI.com
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
TL;DR
A November 9 report says services account for about 79.6% of U.S. employment in 2025, while their share of nonfarm payroll jobs rose from 62% in 1939 to 86% in 2025. It links rising costs in labor-intensive services to Baumol’s cost disease and points to a study in which an AI assistant increased customer-support agents’ issues resolved per hour by 15% on average. How AI will affect service-sector employment and costs remains uncertain.
A report published November 9, 2026, says services made up about 79.6% of U.S. employment in 2025, while a separate payroll-jobs measure puts the share at 86%. The report argues that the shift toward services began well before U.S. manufacturing employment peaked in 1979, and examines whether AI can raise productivity in work where human attention is often part of the service itself.
The two percentages use different measures. The International Labour Organization estimate covers all workers and puts the U.S. services share at 79.6% in 2025. The report’s calculation from Bureau of Labor Statistics Current Employment Statistics puts services at 86% of nonfarm payroll jobs that year; that survey excludes farm workers and self-employed people. The payroll share was 62% in 1939, 67% in 1967 and 72% in 1979.
The report uses economist William Baumol’s cost-disease theory to explain why some services become more expensive even as other parts of the economy grow more productive. A factory can produce goods with fewer labor hours as technology improves. A live quartet, by contrast, still requires musicians to play the piece in real time. If wages rise across the economy, labor-intensive services must pay more too, even when they cannot cut the time needed to provide them as readily.
Prices cited in the report illustrate the difference: from 2000 to 2025, the official quality-adjusted price of a television fell 98%, while hospital services rose 274%, college tuition 188% and car repair 143%. The report also cites a study of 5,172 customer-support agents in which an AI assistant increased issues resolved per hour by an average of 15%. That result is a productivity measure in one study, not evidence that AI has reduced service-sector employment overall.
Infographic 02 · Baumol's cost disease
Why the workers moved to services
Baumol's mechanism in four steps, and what it looks like in US jobs.
- 1966Baumol and Bowen: the quartetA 45-minute Schubert quartet still takes about three hours of human work
- 1967Two kinds of workWhere labour is a tool, machines cut it again and again; where labour is the product, they can’t
- WagesPay rises everywhereTeachers, nurses and musicians must earn more too, or they leave
- ResultCosts rise, people moveIf we keep buying the same mix, more and more workers end up in the slow sector
Service industries, share of US nonfarm payroll jobs
US jobs, annual average 2025 (millions)
Payroll shares exclude farm workers and the self-employed (BLS establishment survey). Health spending: national health expenditure as a share of gross domestic product.
Sources: W. J. Baumol, “Macroeconomics of Unbalanced Growth”, American Economic Review 57(3), 1967, BLS Current Employment Statistics (own calculation of shares), CMS, National Health Expenditure, Table 1.
Powered by Thorsten Meyer AI
AI Meets Labor-Intensive Services
Services now account for about four in five U.S. workers by the all-worker estimate, so changes in service productivity could affect a large part of the workforce and household spending. The report’s examples span health care, education, repairs and customer support: areas where costs, access and the amount of human attention provided can matter directly to consumers.
The customer-support result offers evidence that AI can help workers complete more tasks in at least one setting. It does not show whether employers will use the gain to handle more requests, reduce staffing, improve service or combine those outcomes. That distinction matters: a rise in output per worker is not the same as a decline in the number of workers.
The price comparisons also caution against explaining every increase through Baumol’s theory alone. The report notes that U.S. health spending equaled 18.0% of the economy, compared with 12.4% in Germany. That contrast suggests national institutions and policy also shape the final bill, alongside labor costs and productivity.
Products Worth Considering
Bose QuietComfort Headphones (2nd Gen), Wireless Over-Ear Noise-Cancelling Headphones, Eucalyptus Green
Bose QuietComfort Wireless Noise-Cancelling Headphones, Twilight Blue
DJI Mic 3 (2 TX + 1 RX + Charging Case) Wireless Microphone
As an affiliate, we earn on qualifying purchases.
Services Grew Before Factory Decline
The report challenges the idea that services expanded only after manufacturing jobs disappeared. In the BLS payroll series, services already represented 62% of nonfarm payroll jobs in 1939. Their share reached 72% in 1979, the year U.S. manufacturing employment peaked at nearly 20 million, and continued climbing to 86% in 2025. The series excludes self-employed and farm workers, so it should not be treated as the share of every U.S. worker.
Baumol and economist William Bowen examined the economics of performing arts in their 1966 book, Performing Arts: The Economic Dilemma. In a 1967 paper, Baumol developed the broader argument: some activities gain output per labor hour through technology and scale, while others depend more directly on workers’ time. The report calls this difference a matter of degree, not a strict split; retail, for example, has changed through supermarkets and self-service.
Baumol and Bowen’s quartet example makes the point concrete. A 45-minute Schubert performance requires four musicians for that time, or three labor-hours in total. The report contrasts that constraint with factory production, where technological changes have reduced the labor required to make goods. The analogy helps explain persistent cost pressures, but it does not by itself account for every sector’s prices.
“Human ingenuity has devised ways to reduce the labor necessary to produce an automobile, but no one has yet succeeded in decreasing the human effort expended at a live performance of a 45-minute Schubert quartet much below a total of three man-hours.”
— William Baumol, quoted in the report from his 1967 paper
What AI Productivity Will Change
The report does not specify the customer-support study’s publication details, study period, participating organizations or the range of results among agents. The 15% average therefore cannot establish how well the same tool would work across other employers or service occupations.
It is also unclear whether AI productivity gains will reduce prices, increase service capacity, change job duties or lead to fewer positions. The employment shares describe the structure of the economy, not the effect of AI on jobs. The report frames displacement versus movement into other work as an open question and provides no forecast that resolves it.
The cited price figures show changes between 2000 and 2025, but the material does not provide full methodological detail for each measure. The cross-country health-spending comparison likewise does not isolate which policies or market features explain the difference between the United States and Germany.
Tracking AI’s Effects on Services
The report presents the customer-support study as an early indication of what AI might do in the service economy, not a final answer about jobs or prices. A clearer assessment will require evidence across more occupations, with details on study design and whether productivity gains persist in day-to-day operations.
Readers can also track whether higher output per worker leads employers to handle more requests, change staffing or alter service quality. Future employment and price data will help show whether AI is changing the long-running relationship between wages, productivity and service costs. The source does not name a specific next study, forecast or release date.
Key Questions
What does the 80% figure refer to?
The report cites an ILO estimate that services accounted for 79.6% of all U.S. workers in 2025. A separate BLS-based calculation says services made up 86% of nonfarm payroll jobs; that measure excludes farm workers and self-employed people.
Why can services get more expensive as other products get cheaper?
Baumol’s cost-disease theory says that some work remains tied to human time, such as a live performance or personal care. If wages rise across the economy but these services cannot increase output per labor hour as quickly, their costs can rise relative to goods that become cheaper to produce.
Does the 15% AI result mean customer-support jobs will disappear?
No. The report says an AI assistant raised issues resolved per hour by an average of 15% in a study of 5,172 agents. It does not report that the study found job losses or establish how employers will use the productivity gain.
Did services grow only after manufacturing jobs declined?
No. The BLS payroll series cited by the report shows services at 62% of nonfarm payroll jobs in 1939, decades before the 1979 manufacturing employment peak. Their share rose to 86% by 2025 on that same payroll measure.
Source: ThorstenMeyerAI.com
