🔍 Read the full analysis: Why 80% Of AI Jobs Are Now In Services on ThorstenMeyerAI.com
Get movie nights 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 9 analysis from ThorstenMeyerAI.com says about 79.6% of U.S. workers were in services in 2025, while service industries accounted for 86% of nonfarm payroll jobs. It argues the shift began well before manufacturing employment peaked in 1979 and uses Baumol’s cost disease to explain why labor-intensive services can grow more costly as productivity rises elsewhere. The piece points to AI customer-support research as evidence of potential productivity gains, but does not establish that AI will reduce service-sector employment.
About four in five U.S. workers were employed in services in 2025, according to International Labour Organization estimates cited by ThorstenMeyerAI.com in an analysis published November 9. The article argues that this long-running shift began well before the decline in manufacturing jobs and examines whether artificial intelligence can raise productivity in service work that still depends heavily on human labor.
The article reports that service-providing industries made up 86% of U.S. nonfarm payroll jobs in 2025, based on the author’s calculation using Bureau of Labor Statistics data, a topic explored further in a look at the sectors that shape jobs. That payroll measure excludes farm workers and self-employed people. The ILO estimate, which counts all workers, puts the service-sector share at 79.6% in 2025. These figures use different definitions and should not be treated as the same measure.
The BLS payroll series cited in the analysis shows services represented 62% of nonfarm payroll jobs in 1939, 67% in 1967 and 72% in 1979. The article notes that U.S. manufacturing employment peaked at nearly 20 million in June 1979. Its central point is that services were already the majority of nonfarm payroll jobs decades before that peak, so the shift cannot be explained only as a consequence of factories later shrinking.
To explain why services can become a larger part of the economy, the article turns to economist William Baumol’s cost disease. Some work, such as factory production, can produce more output per worker as technology improves. Other work—such as live music, teaching or personal medical care—can be harder to speed up without changing what the service provides. If wages rise across the economy, labor-intensive services may become more expensive even when their output per worker changes little.
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
How AI Could Change Service Productivity
The issue matters because services cover a wide range of activities people rely on, including health care, education, transport, retail and repairs. If these jobs are difficult to automate or accelerate, productivity gains in manufacturing and other sectors do not automatically translate into lower costs for service users. The analysis uses price changes as an illustration: it reports that, from 2000 to 2025, the quality-adjusted U.S. price of televisions fell 98%, while hospital services rose 274%, college tuition 188% and car repair 143%. These figures are presented as price comparisons over that period, not as proof that productivity alone caused the differences.
AI could alter that pattern if it helps workers complete more tasks in the same amount of time. The article cites a study of 5,172 customer-support agents in which an AI assistant increased issues resolved per hour by 15% on average. That is a productivity result for the studied setting; it does not by itself show that companies will cut staffing, reduce prices or reproduce the same gain in hospitals, schools and other services.
The wider consequence is still open. Higher productivity could allow existing workers to serve more people, change the mix of tasks they perform or reduce demand for some roles. It could also lead to new services and jobs. The analysis frames whether AI will shrink service employment or redirect workers into remaining tasks as an unresolved question, rather than a settled forecast.
Popular right now
Bose QuietComfort Earbuds (2nd Gen) – True Wireless Bluetooth Earbuds with Active Noise Cancellation
Vitamix 5200 Blender with 64 oz. Container, Black
DREAME A3 AWD 1000 LiDAR Robot Lawn Mower
As an affiliate, we earn on qualifying purchases.
Services Grew Before the Factory Peak
The article’s historical comparison challenges a familiar account in which services rose mainly after U.S. factories began shedding jobs. In the BLS payroll data it cites, services already accounted for 62% of nonfarm jobs in 1939. The share then rose through the decades, reaching 72% in 1979 and 86% in 2025. Manufacturing employment reached its reported high point in June 1979, but service employment had been expanding long before then.
Baumol and economist William Bowen examined the economics of performing arts in their 1966 book, Performing Arts: The Economic Dilemma. Their example was a 45-minute Schubert string quartet: four musicians performing for that length of time still require roughly three person-hours of labor. A year later, Baumol developed the idea in his paper “Macroeconomics of Unbalanced Growth: The Anatomy of Urban Crisis.” He distinguished activities where technology can repeatedly raise output per worker from work in which human effort is central to the service itself.
The distinction is not absolute. The source notes that retail changed through supermarkets and self-service, while transport technology made some forms of travel more productive. Baumol’s theory instead describes a difference in how readily productivity can grow. The article also cautions against explaining rising service costs only through inefficiency: it cites increases in haircuts and car repairs, alongside a comparison of U.S. health spending at 18.0% of the economy with Germany’s 12.4%. Those national spending figures show different totals, but do not by themselves identify the causes.
What the AI Evidence Cannot Yet Show
The customer-support study cited in the article reports a 15% average increase in issues resolved per hour among 5,172 agents using an AI assistant. The source material does not provide the study’s full methods, date, participant details or range of results, so those details cannot be assessed here. The reported gain also concerns one type of work; it is not evidence that AI has produced the same effect across the service economy.
It remains unclear whether productivity improvements will translate into lower prices, higher service quality, fewer jobs, new roles or some combination. The cited employment shares describe the size of the service sector, but do not forecast how AI will change it. The source also gives price and national health-spending comparisons without establishing that Baumol’s theory alone explains them.
The Next Test for Service-Sector AI
The next evidence to watch is whether productivity gains from AI extend beyond customer support and appear in other service industries. Future studies and company data could show whether workers handle more cases, whether service quality changes, and whether employers adjust staffing or prices. The source does not identify a specific next report or policy milestone.
For now, the article’s figures establish a long-term trend rather than a forecast: services made up a growing share of U.S. employment across the BLS series, and about four in five workers were in services under the ILO’s 2025 estimate. Whether AI substantially changes that pattern will depend on how broadly the technology can improve output without reducing the human attention that many services require.
Key Questions
Does 80% mean exactly 80% of U.S. jobs are in services?
No. The article cites an ILO estimate of 79.6% of all U.S. workers in services in 2025, rounded to about four in five. Separately, its BLS calculation says services represented 86% of nonfarm payroll jobs, a measure that excludes farm workers and self-employed people.
Did services grow only after manufacturing jobs declined?
No. The BLS payroll figures cited in the article put services at 62% of nonfarm jobs in 1939, decades before manufacturing employment peaked in 1979. The service share continued to rise before and after that peak.
What is Baumol’s cost disease?
It is an economic explanation for why labor-intensive services can become more costly when wages rise across the economy but productivity in those services is harder to increase. A live performance is a common example: cutting the musicians’ time may change the product rather than simply make it more efficient.
Does the customer-support study prove AI will eliminate service jobs?
No. The source cites a study of 5,172 support agents where an AI assistant raised issues resolved per hour by 15% on average. That finding concerns productivity in the studied setting; it does not establish what will happen to employment, prices or staffing across services.
Source: ThorstenMeyerAI.com
