Who Gets Paid? The Labor Share Across Industries

When a sector produces a dollar of GDP, who gets it?

The Bureau of Economic Analysis tracks three components of every industry’s value added. Compensation of employees captures wages, salaries, and benefits — the portion that flows directly to workers. Gross operating surplus captures profits, rents, interest income, and depreciation — the portion that flows to capital owners. And taxes on production and imports captures sales taxes, property taxes, and excise taxes — the portion that flows to government before any income tax is assessed.

In the aggregate economy, the split is roughly 53% compensation, 40% surplus, and 7% taxes. But those averages conceal enormous variation across industries. In the sectors where most Americans work — healthcare, education, professional services — the labor share exceeds two-thirds. In the sectors that produce the most GDP — real estate, information, mining — the capital share dominates. The composition of who gets paid depends entirely on which industry is doing the producing.

The chart reads like a census of economic power. At the top, management companies send 88.7 cents of every dollar to workers — holding companies whose primary asset is the people who run them. Education (80.1%) and healthcare (80.2%) sit right behind, industries where the “product” is essentially human expertise delivered in person. Professional services (68.2%) and administrative services (69.9%) round out a cluster where at least two-thirds of value added flows to paychecks.

At the bottom, the picture inverts entirely. Real estate sends just 5.5 cents of every dollar to workers. The remaining 94.5 cents flows to capital owners (84.6%) and taxes (9.9%). This is not exploitation — it is the nature of the industry. Most of real estate’s $4.1 trillion in value added comes from imputed rent on owner-occupied housing, an economic concept where homeowners are counted as “paying rent to themselves.” There are no workers to pay because no service is being rendered.

Mining (22.9% labor) and agriculture (29.7% labor) follow the same capital-heavy logic — these are industries dominated by expensive equipment, land, and natural resources, where machines and mineral rights generate value that far exceeds the labor bill. Information, at 37.2%, sits in this camp as well: the sector that includes Google, Meta, and Netflix produces $1.6 trillion in GDP with relatively few employees. The technology platforms that define the modern economy are enormously productive per worker — which means most of their value flows to shareholders, not to paychecks.

The table separates cleanly into three tiers. Labor-intensive industries — those where compensation exceeds 60% of value added — include the seven sectors at the top: management, healthcare, education, admin services, other services, professional services, and construction. These are the economy’s paycheck factories, industries where the primary input is human time and expertise.

Mixed industries sit in the middle, with compensation between 40% and 60%. Manufacturing (47.2%) lands here — a sector that once would have been firmly in the labor camp but has shifted toward capital as automation replaced assembly-line workers. Finance (50.0%) sits exactly at the dividing line, a sector of well-paid professionals that nonetheless generates nearly as much surplus from invested capital as it pays in compensation. Retail (45.8%) has been drifting steadily toward the capital side as self-checkout kiosks and e-commerce algorithms replace cashiers and floor associates.

Capital-intensive industries — those where surplus exceeds compensation — include information (37.2%), agriculture (29.7%), mining (22.9%), and real estate (5.5%). These are industries where physical assets, intellectual property, land, and natural resources do most of the work.

The labor share is not static. Over the past 27 years, several major industries have seen dramatic shifts in how their value added is distributed. The pattern, almost without exception, runs in one direction: toward capital and away from workers.

Manufacturing’s labor share fell from 57.1% in 1997 to 47.2% in 2024 — a decline of nearly ten percentage points. This is the story of automation rendered in accounting data. American factories produce far more output than they did in 1997, but they do it with fewer workers commanding a smaller share of the value created. The workers who remain are more productive, but the surplus generated by robots, software, and advanced machinery flows to capital owners.

Retail’s decline is even more dramatic: from 58.5% to 45.8%. In 1997, retail was a solidly labor-intensive industry — stores full of cashiers, stock clerks, and floor associates generating paychecks from every transaction. By 2024, e-commerce, self-checkout, and warehouse automation had fundamentally changed who gets paid when Americans buy things. Amazon’s warehouse robots and Walmart’s automated fulfillment centers generate enormous value added with a fraction of the labor that traditional retail required. The retail dollar increasingly flows to the owners of the logistics platforms, not to the workers inside them.

Information dropped from 42.9% in 1997 to 37.2% in 2024, but the journey was anything but smooth. During the dot-com boom, the sector’s labor share surged to 52.2% in 2000 as startups hired furiously and paid handsomely. Then the bubble burst, and the survivors learned to generate revenue with leaner workforces. By 2007, the share had fallen to 36.8%. It has hovered in the mid-30s ever since, occasionally ticking up when hiring booms hit Big Tech, then falling back as the platforms scaled revenue faster than headcount. The information sector today generates $1.6 trillion in GDP — and most of that value flows to shareholders and intellectual property owners, not to employees.

Healthcare stands as the exception. Its compensation share has remained remarkably stable in the range of 78–82% across the entire 27-year period. Healthcare is an industry that has stubbornly resisted automation. You cannot replace a nurse with software. You cannot outsource a surgery to a robot — not yet. Healthcare’s labor share stayed high because the industry’s product is, fundamentally, human expertise applied to human bodies. As long as that remains true, healthcare will continue to send the vast majority of its income to workers.

Real estate deserves special attention because it is the single largest private industry by GDP — $4.1 trillion, more than manufacturing — and yet it sends just 5.5% of its value added to workers. How can the economy’s biggest sector be its least labor-intensive?

The answer lies in how the BEA measures housing. When a homeowner lives in a house they own, the BEA counts the imputed rent — the amount the homeowner would have paid if they rented the same house on the open market — as real estate value added. This is economically sound: the house is providing a service (shelter) that has real economic value. But there are no workers involved. No one earns a wage from a homeowner occupying their own property. The entire imputed value flows to “capital” — meaning the homeowner’s equity in the house.

Imputed rent accounts for the majority of the real estate sector’s value added. Strip it out, and the sector’s labor share would be much higher — property managers, real estate agents, and maintenance workers do exist. But with imputed rent included, real estate’s 5.5% compensation share is an accurate reflection of an industry whose output is overwhelmingly generated by assets, not people.

This is where the data becomes more than an accounting exercise. If the American economy is shifting from goods-producing sectors to services, the composition of those services determines whether workers share in the growth.

The good news: several of the fastest-growing service sectors are labor-intensive. Professional services — the consultants, engineers, and software developers — rose from 5.8% of GDP to 8.0%, and 68% of that value flows to workers. Healthcare rose from 6.5% to 7.5%, with 80% going to workers. These are industries where growth means more paychecks.

The troubling news: the two largest sectors in the entire economy — real estate (13.8% of GDP) and finance (7.6%) — have labor shares of just 5.5% and 50.0% respectively. Information (5.4% of GDP) sends just 37.2% to workers. As these capital-intensive sectors grow as a share of GDP, the economy’s aggregate labor share may fall for purely structural reasons, even if no individual industry changes its split at all.

This is the composition effect: the economy is not just shifting from goods to services — it is shifting from labor-intensive activities to capital-intensive ones. Manufacturing was 57% labor in 1997 and has fallen to 47%. Retail was 58.5% labor and has fallen to 45.8%. The industries replacing them at the top of the GDP rankings are dominated by assets, algorithms, and intellectual property. The aggregate labor share is falling not because employers became greedier, but because the economy’s center of gravity moved toward industries that inherently pay workers less of what they produce.

In the next episode, we bring the full series together: the final scoreboard of winners and losers across all ten dimensions of America’s industrial transformation — size, growth, labor intensity, and the structural forces that will shape the economy in the decades ahead.