🔍 Read the full analysis: The 80% Service Pivot In AI on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
A November 9, 2026, article examines why services account for about four in five U.S. workers and how AI could affect productivity in those jobs. It cites a study of 5,172 customer-support agents that found an AI assistant increased issues resolved per hour by 15% on average; whether those gains reduce service-sector employment remains uncertain.
A November 9 report from ThorstenMeyerAI.com says services accounted for 79.6% of U.S. employment in 2025, and asks whether artificial intelligence can raise productivity in work where human attention is central. The article links that question to Baumol’s cost disease and cites a study of 5,172 customer-support agents in which an AI assistant increased issues resolved per hour by 15% on average. The result does not establish that AI will eliminate service jobs or bring down prices.
The report distinguishes the share of all workers from the share of payroll jobs. It gives an International Labour Organization estimate that services represented 79.6% of U.S. employment in 2025. Separately, its calculation from Bureau of Labor Statistics payroll data puts service-providing industries at 86% of nonfarm payroll jobs that year. The payroll measure excludes farm workers and self-employed people, so the two percentages describe different populations.
The article says service work was already a majority of nonfarm payroll employment well before U.S. manufacturing employment peaked in 1979. Its BLS-based series puts the service share at 62% in 1939, 67% in 1967 and 72% in 1979. The figures counter an account in which services grew only after factories declined: the shift was underway for decades before the manufacturing peak.
For evidence of AI’s possible effect, the report cites research involving 5,172 customer-support agents, a topic also explored in research on learned memory in customer service. An AI assistant was associated with an average 15% increase in issues resolved per hour. That is a measured result in one work setting, not a forecast for every service occupation. The report frames the broader employment effect as an open question: workers may be displaced, or productivity gains may change the amount and type of service work demanded.
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 issue matters because services make up most U.S. employment, but many service tasks depend on time, attention and human interaction. If AI helps workers handle more tasks per hour, it could ease some cost pressures or let organizations serve more people. Whether either outcome follows depends on how employers use the technology and whether customers accept AI-assisted service.
The report uses Baumol’s cost disease to explain why productivity gains are uneven. In manufacturing, machines and process changes can raise output per worker substantially. A live performance still takes musicians’ time, and attentive teaching or medical care cannot always be expanded without changing the service. If wages rise across the economy while output per hour rises more slowly in these fields, their costs can climb relative to goods produced more efficiently.
The cited support-agent result suggests AI can raise measured throughput in at least one service task. It does not show that the work’s quality stayed constant, that all employees benefited equally, or that the same gains will appear in health care, education or other fields. Those questions will shape whether AI changes costs, working conditions and access to services.
Popular right now
Apple AirPods Pro (1st Generation) with MagSafe Charging Case
Apple AirPods Pro 3 Wireless Earbuds with Active Noise Cancellation
Bose QuietComfort Earbuds (2nd Gen) – True Wireless Bluetooth Earbuds with Active Noise Cancellation
As an affiliate, we earn on qualifying purchases.
From Quartet to Service Jobs
The economic framework behind the report comes from William Baumol and William Bowen’s 1966 book on the performing arts and Baumol’s 1967 paper on unbalanced growth. Their illustrative example was a 45-minute Schubert string quartet: four musicians still require three hours of combined labor to perform it. Unlike factory production, the performance cannot be sped up without changing what the audience receives.
Baumol described a spectrum, not a strict division. Some service work has adopted productivity-enhancing tools, while some goods production still requires substantial labor. The distinction is whether labor is mainly an input that can be reduced while producing the same output, or whether human effort is part of the product. The report applies that framework to contemporary service jobs and AI.
It also compares selected U.S. prices from 2000 to 2025: the report says the quality-adjusted price of televisions fell 98%, while hospital services rose 274%, college tuition 188% and car repair 143%. These are specific price comparisons, not proof that productivity differences alone caused the changes. The article notes that health-system structures matter too, citing health spending at 18.0% of the U.S. economy versus 12.4% in Germany.
““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.””
— Thorsten Meyer, citing William Baumol and William Bowen’s account of the Schubert quartet
The Limits of the Productivity Finding
The cited study’s average 15% result does not establish how AI affects service quality, employee pay, job numbers or customer satisfaction. The source material does not provide the study’s full methods, the range of results across agents, or details that would show how broadly the finding applies beyond customer support.
It is also unclear whether productivity gains will lower prices, increase service availability, raise profits or lead employers to reduce staffing. Those outcomes can vary by industry and by how organizations deploy AI. The reported national employment shares describe the size of the service economy, but they do not forecast how many jobs AI may change.
The price comparisons in the report also do not isolate the causes of cost changes. Health spending, tuition and repair prices reflect more than labor productivity, and the figures alone cannot establish how much Baumol’s theory explains. The report’s contrast between U.S. and German health spending points to possible institutional differences, but does not quantify their effects.
Evidence Needed Beyond Support Work
The next test is whether measured gains in customer support can be replicated across occupations and organizations, with information about service quality as well as output per hour. Results from health care, education, retail and other fields would show whether AI can raise productivity where personal interaction is a central part of the service.
Readers should also watch for evidence about how employers distribute the gains: whether they expand services, change staffing, alter workloads or pass savings on to customers. The report identifies the employment consequences as an unresolved question and does not announce a policy decision or a timetable for wider adoption.
For now, the confirmed figures establish the scale of U.S. service employment and offer one measured example of AI-assisted productivity. They do not yet show that AI will reverse the long-running growth of services or solve the rising costs associated with labor-intensive work.
Key Questions
What does “the 80% service pivot” refer to?
It refers to the report’s ILO estimate that services made up 79.6% of U.S. employment in 2025. A separate BLS-based calculation puts services at 86% of nonfarm payroll jobs; that measure excludes farm workers and the self-employed.
What is Baumol’s cost disease?
It is an economic explanation for why costs can rise in fields where output per worker is hard to increase, even as wages rise alongside those in more productive industries. The report uses a string quartet as an example: the performance still takes four musicians three hours.
What did the customer-support AI study find?
In a study of 5,172 customer-support agents, an AI assistant increased issues resolved per hour by 15% on average, according to the report. That finding does not by itself establish effects on job numbers, service quality or other occupations.
Does the report show that AI will reduce service jobs or prices?
No. It presents the employment and price effects as unresolved. The cited productivity result is specific to customer support, and the report does not show how gains will affect staffing or what customers pay.
Source: ThorstenMeyerAI.com
