What is the Baumol cost disease in services?

The short answer

William Baumol’s 1965 insight started with a simple observation: a Mozart string quartet requires the same four musicians and the same 40 minutes today as it did in 1787. Musicians can’t get more productive at performing the music as written. Yet musician wages have risen by orders of magnitude over 200 years, in line with general wage growth in the economy.

The implication is profound. In any sector where productivity can’t rise (or rises much more slowly than the rest of the economy), unit labor costs must rise faster than average. The relative price of those goods and services drifts higher. The share of total spending devoted to them grows over time, mathematically.

The angle that distinguishes Baumol’s framework is that this isn’t a market failure or policy mistake. It’s a logical consequence of uneven productivity growth across sectors. Healthcare, education and many public services have low productivity growth and rising spending shares not because they’re “broken” but because they’re succeeding at retaining workers in a wider economy where productivity is rising elsewhere.

→ New to public economics? Financial education hub

What the data shows

The most cited evidence comes from OECD Health Statistics, BEA detailed personal consumption data, and various Baumol-disease replications.

The empirical context (OECD, BEA, Baumol-Towse, 1960-2024):

• US healthcare spending: from ~5% of GDP in 1960 to ~18% in 2024 (CMS/OECD)
• OECD average general government spending: from ~28% of GDP in 1960 to ~42% in 2023 (OECD)
• US college tuition: rose 1,400% in nominal terms 1980-2020 vs ~250% general CPI (BLS)
• Performing arts ticket prices: rose ~3% per year above general inflation since 1960 (Baumol-Towse follow-up studies)
• US healthcare productivity growth: estimated near zero by Skinner-Staiger; arts productivity by definition flat in many disciplines
• Manufacturing productivity in same period: ~2.5% per year average

The exception worth noting: technology has eroded Baumol disease in some traditional service sectors. Telehealth, MOOCs and recorded music distribution have introduced productivity gains in fields where Baumol predicted none. The disease is not destiny — but it remains the dominant force in face-to-face care services.

→ Dataset: US Government Spending GDP

Why it happens — the macro mechanism

Baumol’s cost disease operates through the interaction of labor markets across sectors with different productivity growth rates.

Channel 1 — Wage equalization across sectors. In a competitive labor market, workers move between sectors in response to wage differentials. If wages in manufacturing rise as productivity grows, workers in healthcare and education must receive comparable wage increases or they will leave. But healthcare and education productivity grew much more slowly. So unit labor costs in these sectors rise persistently faster than in the rest of the economy.

Channel 2 — Inelastic demand for affected services. If demand for these services were highly elastic, rising prices would reduce quantity consumed and hold spending shares constant. But demand for healthcare, education and many public services is highly inelastic — people don’t reduce hospital visits when prices rise, especially in systems with insurance pooling or government funding. Rising prices feed directly into rising spending shares. The angle that distinguishes Baumol from a generic relative-price story: it’s the combination of slow productivity AND inelastic demand that produces the rising share. See our FAQ on automation and wages.

A third channel runs through public budgets.

Channel 3 — Public-sector dominance amplifies the effect. Many Baumol-affected services (healthcare, education, public administration) are heavily public-sector. Government spending is less price-sensitive and less responsive to cost pressures than private spending. This means Baumol dynamics translate more directly into rising public spending shares than into private substitution. See our FAQ on public debt tipping points.

Synthesis by regime. In the postwar period 1950-1970, rapid productivity growth in manufacturing, agriculture and construction outpaced services significantly — Baumol disease accelerated, with healthcare and education spending shares rising sharply. From 1980 to 2010, productivity growth slowed across the board (the TFP slowdown), narrowing the gap between manufacturing and services productivity rates and slowing Baumol pressure modestly — but the structural dynamic continued. Since 2010, generative AI and digital tools have begun penetrating Baumol-affected sectors (telemedicine, online education, automated administrative services), which could either erode the disease or push it into more concentrated face-to-face premium tiers. The future trajectory depends on which.

Baumol’s cost disease isn’t a disease at all. It’s the price of keeping musicians in orchestras when they could otherwise be programming software.

→ Working framework: Macro-financial regimes

What it means for different economic actors

Long-term investors. Industries subject to Baumol disease (healthcare, certain education segments) have shown remarkable revenue resilience over decades because their nominal spending shares grow even when volumes don’t. This makes them defensive in real terms but exposed to political pressure when public budgets squeeze.

Policymakers. Baumol dynamics imply that rising healthcare and education spending shares are not necessarily evidence of waste or mismanagement. The challenge is distinguishing structural Baumol effects (which should be expected) from genuine inefficiency (which can be addressed).

Workers. Workers in low-productivity service sectors benefit from Baumol pressure because their wages rise even when their productivity doesn’t. This is one structural reason why service-sector employment has been so resilient in advanced economies — Baumol creates wage pressure that automation can’t easily eliminate.

A common error is to treat rising healthcare costs as exclusively reflecting inefficiency or political failure. Some portion is structural Baumol, some portion is genuine inefficiency, and the two are difficult to disentangle empirically.

Practical observation

Go deeper

Related questions

• What is total factor productivity and why has it slowed?
• How does automation affect wages and employment?
• When does public debt reach tipping points?
• How does the rise of intangible capital change valuations?

Frequently asked questions

Why does Baumol’s argument explain why services get more expensive even without inefficiency?

The core insight is that wages in any given sector are determined by economy-wide labor market conditions, not by sector-specific productivity. A nurse’s wages must keep pace with a software engineer’s wages or nurses will retrain as software engineers. But hospitals can’t make a nurse’s hour of patient care more productive at the same rate that software firms make a programmer’s hour more productive. So hospital costs rise faster than software costs, and the relative price of healthcare drifts higher year after year. This isn’t waste — it’s the mathematical consequence of uneven productivity growth across an integrated labor market.

Has technology eliminated Baumol disease in some sectors?

Partially. Recorded music and broadcasting eroded the original Baumol example — orchestral concerts are still subject to the disease, but recorded music distribution has rapidly rising productivity. MOOCs and online courses have introduced productivity gains in some education segments. Telemedicine and AI diagnostics may begin doing the same in healthcare. But the face-to-face premium services (in-person medical care, elite tutoring, live performance) remain Baumol-affected. The disease has shifted to a smaller but still-growing footprint.

Should governments restrain Baumol-driven public spending growth?

This is a normative question that economic analysis can illuminate but not resolve. Restraining nominal spending growth in healthcare and education effectively means accepting falling real wages for workers in those sectors relative to other sectors — workers will leave, quality may fall, supply will shrink. The alternative — accepting persistently rising shares of public spending on these services — has fiscal sustainability implications. Different societies have made different tradeoffs. The Baumol framework clarifies the tradeoff without prescribing a resolution.

Last updated — 12 July 2026