In 1966, William Baumol and William Bowen published a study about the performing arts that accidentally explained one of the most persistent frustrations of contemporary economic life. Their observation was deceptively simple: a Mozart string quartet written in 1787 still requires four musicians and roughly the same number of minutes to perform today. No amount of technological progress can make it faster. Meanwhile, manufacturing output per worker has increased roughly fortyfold since Mozart's era.
This asymmetry—some sectors racing ahead in productivity while others remain stubbornly labor-intensive—creates what economists now call Baumol's cost disease. It is not really a disease at all, but rather an iron law of relative prices. When productivity soars in manufacturing and technology, wages rise across the entire economy. Workers in stagnant-productivity sectors must be paid competitive wages or they will leave for more productive industries. The result: the relative cost of labor-intensive services climbs inexorably, even when nothing about those services has changed.
The sectors that cannot be automated away
Healthcare provides the starkest illustration. A nurse performing a physical examination takes roughly the same time today as in 1970. A surgeon's hands still move at human speed. Diagnostic technology has improved dramatically, but the core acts of care—listening, examining, treating—remain stubbornly bound to human attention and time. As wages in the broader economy rise, hospitals must pay more to attract and retain staff, and those costs flow directly to patients and insurers.
Education follows the same pattern. A classroom of twenty students still requires a teacher's sustained presence. Online learning has created efficiencies at the margins, but the fundamental act of instruction—explaining, questioning, correcting—resists compression. Universities have responded by increasing class sizes, relying on adjunct faculty, and raising tuition. The cost disease explains why education spending as a share of GDP has risen steadily across wealthy nations even as educational technology has proliferated.
Personal services complete the trifecture. A haircut takes as long as it ever did. A dental cleaning, a therapy session, a legal consultation—all remain anchored to the irreducible reality of one professional serving one client at a time.
Why this matters for policy
Baumol's insight carries uncomfortable implications for political debates about costs. When politicians promise to make healthcare or education affordable through market competition alone, they are often fighting against a structural force that competition cannot overcome. A hospital cannot achieve the productivity gains of a semiconductor factory. A university cannot scale like a software company.
This does not mean nothing can be done. Administrative bloat, regulatory capture, and rent-seeking all layer additional costs atop the fundamental disease. But it does mean that some portion of rising costs in these sectors is not waste, corruption, or inefficiency—it is the mathematical consequence of paying twenty-first-century wages for tasks that cannot be performed any faster than they were performed in the twentieth.
Baumol himself, who died in 2017, was optimistic about the implications. He argued that rising productivity in manufacturing makes society wealthy enough to afford more expensive services. The disease is painful only if we refuse to acknowledge its existence and allocate resources accordingly.
Our take
Baumol's cost disease is the economic equivalent of gravity—invisible, omnipresent, and indifferent to ideology. It explains why your parents could afford a doctor's visit on a factory wage and you cannot, even though you earn more in real terms. It explains why tuition rises faster than inflation and why live music feels like a luxury. Understanding it will not make the bills smaller, but it might make the frustration more bearable. Some things simply cost what they cost because they require a human being's undivided time, and that time grows more expensive every year that the robots get faster.




