🔍 Read the full analysis: Understanding AI's 80% Service Focus 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 November 9 report examines why services account for about four in five U.S. workers and how Baumol’s cost disease helps explain why some service costs rise while manufactured goods get cheaper. It cites a study in which an AI assistant increased customer-support issues resolved per hour by 15% on average, but whether that productivity gain reduces service-sector employment remains unknown.
A November 9 report from ThorstenMeyerAI.com examines why about four in five U.S. workers are in services and whether artificial intelligence can lift productivity in work that has historically been difficult to automate. It cites a study of 5,172 customer-support agents in which an AI assistant raised issues resolved per hour by an average of 15%, while leaving open whether such gains will reduce jobs or change the work people do.
The report uses Baumol’s cost disease to explain why some services can become more expensive even as manufactured goods get cheaper. In a 1967 paper, economist William Baumol described a difference between work where labor is an input that technology can reduce and work where human effort is central to the product. A live string quartet, for example, still requires four musicians to perform together; productivity cannot be raised simply by having them play faster without changing the performance.
The report’s employment figures show that the service shift predates the decline of factory jobs. Based on its calculation from U.S. Bureau of Labor Statistics payroll data, services accounted for 62% of nonfarm payroll jobs in 1939, 72% in 1979 and 86% in 2025. The BLS series excludes farm workers and self-employed people. Using a different measure that counts all workers, International Labour Organization estimates put services at 79.6% of U.S. employment in 2025.
Price comparisons illustrate the uneven pattern. The report says the official, quality-adjusted U.S. price of a television fell 98% between 2000 and 2025, while hospital services rose 274%, college tuition 188% and car repair 143%. Those figures are attributed to the source report; the excerpts provided do not specify the underlying price series. The report also notes that U.S. health spending was 18.0% of the economy, compared with 12.4% in Germany, suggesting that Baumol’s theory alone does not account for differences in the total bill.
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 central question is whether AI can change the productivity limits that have helped push up costs in parts of the service economy. If an AI tool lets an employee resolve more support requests in the same hour, that could increase output without adding staff. Whether employers pass on any resulting savings to customers, use the time for more complex cases, or reduce hiring is not established by the cited productivity figure.
The issue reaches beyond customer support. Services include healthcare, education, transport, retail and government, among other activities. In some of these fields, personal attention is part of the service rather than an avoidable input. That can limit how far productivity gains go without changing quality or access. The report presents AI as a technology aimed at service work, but it does not show that the same results apply across these varied occupations.
For readers, the distinction matters because productivity growth and lower prices are not automatic synonyms. The report’s figures contrast large price reductions for televisions with increases in several services. They also show why attributing rising costs to one cause, such as waste or market power, may miss broader differences in how work is produced. The comparison with Germany indicates that policy and national systems can shape spending alongside the economic pressures Baumol described.
Popular right now
Vitamix Touch & Go Advance 32-Ounce Bar Blender, White
Dyson Airsmooth Styling Brush
Dyson V8 Cyclone Cordless Vacuum
As an affiliate, we earn on qualifying purchases.
From Quartets to Cost Disease
Baumol and economist William Bowen examined the finances of performing arts in their 1966 book, Performing Arts: The Economic Dilemma. The example that later became well known is a 45-minute Schubert quartet: four musicians contribute a total of about three hours of labor. The time required for the live performance has not fallen in the way factory labor per manufactured item can fall when production methods improve.
Baumol developed the broader argument in his 1967 American Economic Review paper, “Macroeconomics of Unbalanced Growth: The Anatomy of Urban Crisis.” The report says wages in less productivity-driven activities still tend to rise alongside wages elsewhere, or workers may leave for better-paid jobs. When output per worker grows more slowly but labor costs rise, the cost of producing those services can increase. Baumol described the distinction as a matter of degree, not a strict division: technology has also changed some service work, including retail and travel.
The report places this theory against a longer employment trend. U.S. manufacturing employment reached almost 20 million in June 1979, according to the source, but services had already accounted for a majority of nonfarm payroll jobs decades earlier. The figures use different employment definitions, so the BLS payroll share and the ILO all-worker estimate should not be treated as directly interchangeable.
““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 and William Bowen, quoted from their 1966 book as cited by economist Alex Tabarrok
AI’s Employment Effects Remain Open
The source reports a 15% average increase in issues resolved per hour in a study of 5,172 customer-support agents, but the supplied material does not identify the study’s date, research team, employer, methods or duration. It also does not explain how the average varied among workers or what counted as a resolved issue. The result should not be read as evidence that every AI system or workplace will achieve the same gain.
The employment consequences are also unsettled. The report asks whether AI will shrink service-sector work or move workers toward tasks that remain. It provides no figures on layoffs, hiring, wages, customer satisfaction or the quality of AI-assisted support. Nor does it establish whether productivity improvements will lower prices. The service-sector shares and price changes are historical measures, not forecasts of what AI will do next.
Some figures also rely on different definitions and sources. The 86% figure refers to nonfarm payroll jobs in the BLS series, while the 79.6% figure uses an ILO estimate covering all workers. The supplied material does not include full citations for the price comparisons, so their precise measurement methods cannot be checked from the excerpt alone.
Evidence Needed on Jobs and Costs
The next evidence to watch is whether AI productivity gains appear outside the customer-support study and persist over time. Further research could clarify how results differ by task, experience and workplace, and whether higher issue-resolution rates affect service quality as well as output.
Employment and price data will be needed to assess the wider effects. Changes in staffing, wages, hours, customer wait times and service costs would help show whether employers use AI to reduce labor, expand capacity or redirect workers. Until that evidence is available, the report frames AI’s effect on the service economy as an open question rather than a settled forecast.
Key Questions
What does the report mean by AI’s “80% service focus”?
It refers to the estimate that 79.6% of U.S. workers were in services in 2025, according to the International Labour Organization figures cited by the report. The report’s separate BLS-based calculation says services made up 86% of nonfarm payroll jobs that year; the measures cover different worker groups.
What is Baumol’s cost disease?
It is an economic explanation for why costs can rise in activities where productivity is hard to increase. If wages rise across the economy while some work still requires much the same labor, producing those services can become more expensive relative to goods whose production has grown more efficient.
What did the customer-support study find?
The report says an AI assistant increased issues resolved per hour by an average of 15% among 5,172 customer-support agents. The supplied material does not provide enough study details to determine how broadly the result applies.
Does the study show that AI will eliminate service jobs?
No. The reported productivity result does not establish what happened to employment, wages or hiring. The source says it remains unclear whether AI will reduce service-sector jobs or shift workers toward other tasks.
Why did service prices rise while television prices fell?
The report attributes part of the difference to how easily technology can reduce labor per unit of output. Television production can become more efficient, while some services rely on human time and attention. The report also notes that institutional and national differences affect costs, particularly in healthcare.
Source: ThorstenMeyerAI.com
