🔍 Read the full analysis: AI Market Shift: 80% Service-Driven on ThorstenMeyerAI.com
Get smart everyday buys 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 Nov. 9, 2026, report says services account for about four in five U.S. workers, a shift that began well before the manufacturing employment peak in 1979. It links rising costs in labor-intensive services to Baumol’s cost disease and cites a study in which an AI assistant increased customer-support issues resolved per hour by 15% on average. The report does not establish whether AI will reduce service-sector employment.
A report published November 9, 2026, says services accounted for 79.6% of U.S. employment in 2025, and argues that the shift toward service work was underway decades before manufacturing employment peaked in 1979. The report connects the pattern to economist William Baumol’s account of why labor-intensive services can become more expensive, while pointing to an AI customer-support study that found a 15% average increase in issues resolved per hour.
The employment figures use two measures. The International Labour Organization estimated that services made up 79.6% of all U.S. workers in 2025. Separately, the report’s calculation from Bureau of Labor Statistics payroll data found that service-providing industries accounted for 86% of nonfarm payroll jobs that year. The payroll series excludes farm workers and self-employed people, so the figures describe different groups and should not be treated as interchangeable.
The BLS payroll calculation traces the rise back to 1939, when services already represented 62% of nonfarm payroll jobs. The share reached 67% in 1967 and 72% in 1979, the year U.S. manufacturing employment reached its reported peak of nearly 20 million. The report uses this timeline to challenge the idea that services grew only after factory jobs began to decline.
To explain why some services cost more over time, the report revisits Baumol’s 1967 theory. In sectors where productivity can rise through machinery and scale, fewer labor hours may be needed for the same output. In work such as live performance or personal care, labor is more directly part of the service being purchased. The report says wages in both types of work tend to rise, putting pressure on costs in sectors where labor time is harder to reduce.
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
The report’s account matters because services employ a large share of U.S. workers and include areas central to household budgets, such as health care, education, repairs and personal care. If labor productivity in these services rises more slowly than in manufacturing, costs can increase even when providers are not simply choosing to charge more. That is the mechanism described by Baumol’s cost disease; it is not, by itself, proof that any particular price rise has a single cause.
The cited AI study offers a concrete, but limited, example of technology being applied to service work. Across 5,172 customer-support agents, an AI assistant was associated with 15% more issues resolved per hour on average. The report presents this as evidence that AI can improve output in at least one service setting, not as evidence that the same gains will occur across hospitals, schools, shops or other occupations.
For workers and consumers, the unresolved consequence is how productivity gains are distributed. Faster issue resolution could lower costs, allow staff to handle more requests or change the number and type of jobs needed. The cited result alone does not establish which outcome will follow, or whether service-sector employment will shrink.
Products Worth Considering
FLEXISPOT Electric Adjustable Workbench, 55 x 28 Inches, White
LG UltraGear 27-inch QHD OLED Gaming Monitor, 240Hz
Dell UltraSharp U2724DE 27-Inch WQHD IPS Black Monitor
As an affiliate, we earn on qualifying purchases.
A Service Shift Before 1979
The report places the current employment mix in a longer timeline. Its BLS series begins in 1939 and shows service industries held a majority of nonfarm payroll jobs at the outset. Their share then rose steadily through the 1967 publication of Baumol and William Bowen’s study of the performing arts and through the 1979 manufacturing employment peak. This sequence supports the report’s point that service growth did not begin as a simple consequence of factory decline.
Baumol and Bowen’s 1966 book, Performing Arts: The Economic Dilemma, examined the finances of theater, opera, music and dance. The report highlights their example of a 45-minute Schubert string quartet: four musicians perform for 45 minutes, totaling three hours of labor. Unlike factory production, the performance cannot easily be shortened without changing the product. Baumol developed a broader account of this productivity difference in his 1967 paper, “Macroeconomics of Unbalanced Growth: The Anatomy of Urban Crisis.”
The report also compares official U.S. price changes since 2000: it says quality-adjusted television prices fell 98%, while hospital services rose 274%, college tuition 188% and car repair 143%. These are figures reported in the source, not a full explanation of each price trend. The comparison between U.S. and German health spending—18.0% and 12.4% of their economies, respectively—also indicates that institutional differences may affect costs beyond the general productivity pressures Baumol described.
How Far AI Gains Will Reach
The source gives a result from one customer-support study but does not provide its full methodology, the time period over which the 15% average was measured, or details about how the participating agents and workplaces were selected. The figure therefore cannot establish how AI will affect other service occupations, nor does it show whether the productivity change reduced costs or altered employment.
The employment shares also depend on the measure used: the ILO figure covers all workers, while the BLS calculation covers nonfarm payroll jobs and excludes the self-employed. The report’s price comparisons do not isolate the causes of each increase or decrease. It remains unclear how much of the cost growth reflects labor productivity, institutional arrangements, demand, or other factors.
Evidence Needed on Jobs and Costs
The next evidence to watch is whether productivity gains from AI appear in more service workplaces and persist over time. Useful comparisons would track output, service quality, prices and staffing together, rather than treating a rise in issues resolved per hour as a complete account of an AI system’s effects. The source does not announce a new study or policy decision, so no specific follow-up date is established.
For the broader employment picture, later labor data can show whether the service share continues to rise and how it changes across industries. Until those measures and wider workplace studies are available, the report’s central conclusion is limited: services have long made up most U.S. employment, and AI has shown a productivity gain in one customer-support study, but the effects on service jobs and prices remain unsettled.
Key Questions
What does the report mean by an 80% service-driven workforce?
It refers to the International Labour Organization’s estimate that services accounted for 79.6% of U.S. workers in 2025. A separate BLS-based calculation puts services at 86% of nonfarm payroll jobs; that measure excludes farm workers and the self-employed.
Did the rise of services begin after U.S. factory employment peaked?
No. The report’s BLS payroll calculation says services made up 62% of nonfarm payroll jobs in 1939, rising to 72% by 1979, the year manufacturing employment peaked. The shift was already underway well before that peak.
What is Baumol’s cost disease?
It describes pressure on costs in work where labor time is difficult to reduce, such as live performance or personal care. If wages rise across the economy while productivity gains are harder to achieve in those services, their costs can rise relative to sectors where technology allows more output per labor hour.
Does the customer-support study show that AI will eliminate service jobs?
No. The report cites a study of 5,172 customer-support agents in which an AI assistant raised issues resolved per hour by 15% on average. That result does not establish what happened to employment or whether the same effect will occur in other jobs.
What remains unknown about AI’s effect on service work?
The source does not provide enough study detail to judge how broadly the measured productivity gain applies. It also does not establish whether AI will lower service prices, change staffing levels, improve service quality or shift workers into other roles.
Source: ThorstenMeyerAI.com
