🔍 Read the full analysis: Why 80% Of AI Work Is Service-Oriented on ThorstenMeyerAI.com
Get the little things that make your day 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, 2026, article argues that the U.S. shift toward service work began well before manufacturing employment peaked, and uses Baumol’s cost disease to explain why some services get more expensive. It also cites a study in which an AI assistant increased customer-support issues resolved per hour by 15% on average, while leaving open whether AI will reduce service-sector employment or redirect workers to other tasks.
A November 9, 2026, episode of ThorstenMeyerAI.com’s Work After Work series argues that the long-term rise of U.S. service employment reflects more than the decline of factory jobs: some services are hard to make more productive because they depend on human time. The article connects that pattern, known as Baumol’s cost disease, to price increases in services and to a customer-support study in which an AI assistant raised issues resolved per hour by 15% on average.
The episode draws on Bureau of Labor Statistics payroll data to show that service-providing industries accounted for 62% of U.S. nonfarm payroll jobs in 1939, rising to 72% in 1979 and 86% in 2025. The field, factory, and office offer another lens on the changing nature of work. Those figures exclude farm workers and self-employed people. A separate International Labour Organization estimate, which counts workers more broadly, puts services at 79.6% of U.S. employment in 2025 — roughly four in five workers.
The article says services were gaining share before U.S. manufacturing employment reached its peak in June 1979, when manufacturing employed almost 20 million people, and continued gaining share afterward. The history of work provides a broader view of these shifts. It uses the example of a live performance: four musicians playing a 45-minute Schubert quartet still require three hours of combined labor. By contrast, technology can reduce the labor needed to make manufactured goods.
In a 1967 paper, economist William Baumol described this difference between work where labor is mainly an input and work where labor is central to the product. When productivity grows in one part of the economy, wages can rise across sectors as workers seek comparable pay. Where output per worker is difficult to increase without changing the service, the article argues, labor costs can rise faster than productivity. AI’s impact on work also raises questions about how workers may be affected. It cites prices since 2000: quality-adjusted television prices fell 98%, while hospital services rose 274%, college tuition 188% and car repair 143%. The episode attributes the figures to official U.S. price measures but does not provide the underlying series in the supplied material.
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 Service Costs Follow Productivity
The framework matters because it links two visible changes: the growing share of jobs in services and the rising cost of some labor-intensive services. If manufacturing can produce more with fewer hours while teaching, health care or live performance still requires substantial direct human time, employment can shift toward services even without a sudden collapse in factory output. The episode’s historical figures show that the shift was underway decades before the 1979 manufacturing peak.
It also cautions against treating every service price increase as evidence of waste or profiteering. The article notes that U.S. haircut and car-repair prices more than doubled between 2000 and 2025, even though those services do not depend on student loans or health insurers. At the same time, it points to differences in national health spending: the United States spent 18.0% of its economy on health, compared with 12.4% in Germany. Baumol’s theory may help explain pressure shared across labor-intensive sectors, but it does not by itself explain why particular countries or services cost more.
AI could affect this pattern if it raises productivity in tasks that have relied on workers’ time. The customer-support result suggests that software can help workers handle more issues per hour. It does not show that AI can replace the personal attention involved in every service, nor does it establish what happens to total employment or wages.
Products Worth Considering
Arlo Pro Wireless Security Camera, 2K HDR, 6th Generation, 4-Camera System
Shark AI Ultra Robot Vacuum with Matrix Clean Navigation, Self-Empty Base, HEPA Filtration, Pet Hair Friendly, Wi‑Fi, Dark Grey
Bose Lifestyle Ultra Soundbar and Ultra Subwoofer Bundle
As an affiliate, we earn on qualifying purchases.
From Orchestras to Service Jobs
Economists William Baumol and William Bowen studied the finances of performing arts organizations in their 1966 book, Performing Arts: The Economic Dilemma. Their example contrasted labor-intensive live performance with manufacturing, where machinery and production methods can reduce the labor required for each unit. Baumol developed the wider argument in his 1967 paper, “Macroeconomics of Unbalanced Growth: The Anatomy of Urban Crisis.”
The episode quotes Baumol’s observation that a 45-minute Schubert quartet still takes three combined hours of musician labor. It contrasts that with an assembly-line example from earlier in the series: labor per car chassis fell sharply after Ford introduced a moving assembly line in 1913. The comparison is illustrative, not a claim that every factory task improves at the same rate or that all service work is resistant to technology.
The employment statistics use different measures. The BLS figures cited in the article cover payroll jobs outside farming and exclude self-employment; the ILO estimate counts employment across a broader group of workers. Their percentages should not be treated as directly interchangeable. Both, however, indicate that service work represents a large share of U.S. employment.
““When someone purchases an air conditioner he neither knows nor cares how much labor went into it.””
— William Baumol, quoted from his 1967 paper
AI’s Employment Effects Remain Open
The cited customer-support study involved 5,172 agents and found that an AI assistant raised the number of issues resolved per hour by 15% on average, according to the episode. The supplied material does not identify the study’s authors, publication date, workplaces, study design or duration, so those details cannot be assessed here. The result is a productivity measure, not evidence that employers cut jobs or that the same effect applies across service industries.
It is also unclear whether AI will reduce the number of workers needed, allow companies to serve more customers, change the kinds of tasks employees perform, or produce a mix of those outcomes. The episode frames that as an open question. Its price comparisons likewise do not establish that Baumol’s cost disease alone caused the increases; the article itself points to country-level differences in health spending and other possible influences.
Tracking Productivity and Jobs
The episode identifies the next issue for its series as whether AI will draw workers out of service jobs or increase productivity while shifting people into other work. Answering that will require evidence beyond a rise in cases handled per hour: readers would need to see how staffing, wages, service quality and total output change over time.
For the current claims, useful next checks include the original customer-support study and the underlying BLS, ILO and price-index data. Those sources could clarify how the measures were constructed and whether the observed patterns extend across sectors. The article does not announce a specific deadline or additional study, so the timing of further evidence remains unknown.
Key Questions
What is Baumol’s cost disease?
It is an economic explanation for why costs can rise in activities where productivity is difficult to increase. If wages rise across the economy while a service still needs roughly the same amount of direct labor, its cost per service may rise.
Does the article say 80% of U.S. workers are in services?
It cites an International Labour Organization estimate of 79.6% of U.S. employment in services in 2025. Its separate BLS figure is 86% of nonfarm payroll jobs, a measure that excludes farm workers and self-employed people.
What did the customer-support AI study find?
The episode says a study of 5,172 customer-support agents found that an AI assistant raised issues resolved per hour by 15% on average. The supplied article does not give enough study details to establish how broadly that result applies.
Does that result prove AI will eliminate service jobs?
No. A higher number of issues resolved per hour does not by itself show whether employers will reduce staffing, serve more customers, change workers’ tasks or combine those responses. The episode presents the employment effect as unresolved.
Source: ThorstenMeyerAI.com
