Baumol Effect

Some services get more expensive because other jobs get more productive.

Also calledBaumol’s cost disease.
A cost pressure, not a verdict on how hard people work.

A haircut still takes someone’s time, even as machines turn out more goods each hour. As pay rises across the economy, a barbershop faces pressure to raise wages too. The same haircut can then cost more, without taking a minute longer.

Two jobs. One wage market.

Imagine a mug factory and a live string quartet. Better machinery lets a factory worker make more mugs in an hour. But four musicians still need an hour to perform the same one-hour concert.

Change how much the factory produces per worker-hour

Hourly pay rises in both jobs
Assumption: pay keeps pace with factory productivity.

$40 / hour
Starting pay: $20 / hour

The factory

One worker, one hour, better machinery.

20 mugs / hour

Starting output: 10 mugs per worker-hour

Workers needed: 1
Labour bill for one hour: $40

Labour cost per mug

$2 was $2

Unchanged

The quartet

Four musicians, the same one-hour concert.

1 performance / hour

Starting output: also 1 performance per hour

Musicians needed: 4
Labour bill for one hour: $160

Labour cost per performance

$160 was $80

Doubled

Pay doubles, but so does the factory’s output. Its labour cost stays at $2 per mug. The quartet’s output stays the same, so its labour cost doubles to $160 per performance.

Illustrative dollars, not historical data. This model counts wages only, with equal hourly pay in both jobs and unchanged concert length and audience size. It excludes materials, rent, profits, and how tickets are priced. There is no assumed number of years between settings.

The connection is competition for people. Higher output makes higher factory pay possible. As better-paying opportunities spread, services face pressure to raise wages to attract and keep staff. A violinist need not be ready to run a factory machine tomorrow for the wider job market to matter.

The Baumol effect is this upward cost pressure on work whose output per hour grows slowly, while wages rise with the more productive parts of the economy.

A rising cost is a clue, not a diagnosis.

People can work just as well.

Playing the piece twice as fast changes the concert. Giving each child half as much attention changes the care. Sometimes the time spent is part of what people are paying for.

Other causes still matter.

If a concert hall raises ticket prices because its rent jumps, that increase is not explained by the Baumol effect. Market power, waste, changing quality, and other costs can also raise a bill.

Technology can change the result. A bigger audience spreads the same performers’ labour cost over more listeners. A recording can reach many more still. That reaches more listeners per hour of performers’ work, but a recording is a different product from a seat at a live concert.

Look for time that is hard to compress.

The pattern can appear inside a service, even when other parts of that service improve quickly.

Childcare: A carer can use better scheduling software, yet still needs to spend time watching and helping a small group of children.
Education: Materials can be copied cheaply. A teacher’s careful conversation with a struggling student still takes time.
Healthcare: Some treatments and tests become far more efficient. Bedside attention and hands-on care can be harder to speed up.

Why the idea began with performers

Live performers could not turn out concerts the way factories turned out more goods: a string quartet still needed four players for the length of the piece. Economists William Baumol and William Bowen studied the finances of performing arts organizations and connected that constraint to rising pay in the wider economy. Their work explained how a well-run company could face growing financial pressure without its performers doing a worse job.

They developed the argument in the 1960s, including their 1966 book Performing Arts: The Economic Dilemma. The name “Baumol’s cost disease” came to describe the broader cost pressure on services with slow productivity growth.

Origins: Princeton’s account of Baumol, Bowen, and the quartet Original work: Baumol & Bowen, Performing Arts (1966)

The arithmetic, assumptions, and affordability

The basic calculation is labour cost divided by output. At the doubled setting:

Factory: $40 ÷ 20 mugs = $2 per mug
Quartet: (4 × $40) ÷ 1 performance = $160 per performance

More generally, labour cost per unit rises when hourly pay grows faster than output per worker-hour. Productivity need not be completely flat. The model uses identical wages and full wage growth in both jobs to isolate the relationship; actual wages can move at different speeds, and productivity gains do not automatically become pay rises.

Relative cost is the central point. The service becomes more costly compared with goods whose production improves faster. That can happen even if the service’s dollar price stays flat while goods get cheaper.

A rising relative price does not mean society must become unable to afford the service. More productive industries also create more goods and income. Who receives that income, and how services are funded, still determines who can afford access. “Cost disease” describes a pressure; it does not prescribe a policy.

Further reading: Baumol, The Cost Disease (Yale University Press, 2012)

When machines save time, human time can cost more.

Before asking why a service has failed to get cheaper, ask how much faster it can become while still giving people what they value.