Growth Accounting

GDP grows because economies deploy more labour and capital, and because they use them more efficiently. Growth accounting decomposes the rate of GDP growth into the share attributable to increases in labour input, capital input, and a residual called total factor productivity (TFP). It answers a deceptively simple question: is growth coming from working harder and investing more, or from discovering better ways to do things?

The basic decomposition

Suppose an economy’s output in a year is Y, produced with labour L and capital K. The simplest production function is Cobb-Douglas: Y = A × K^α × L^(1−α), where α is capital’s share of income and A is a productivity level.

Taking the growth rate of both sides:

ΔY/Y = ΔA/A + α × (ΔK/K) + (1−α) × (ΔL/L)

This identity says that output growth comes from three sources. The first term, ΔA/A, is total factor productivity growth (TFP growth). It captures improvements in technology, management, organisation, or any other factor that raises output given fixed K and L. The second term attributes growth to capital accumulation; the third to labour growth. The weights are labour’s and capital’s income shares.

This is the core of growth accounting. The quantities K, L, Y are measured from national accounts and labour statistics. The shares α and 1−α are estimated from income distribution data. What remains unexplained—TFP—is the famous Solow residual.

Measuring the inputs

Labour input is not just the number of workers; it is hours worked, adjusted for composition. An economy with 100 million workers working 30 hours a week has lower labour input than one with 100 million working 40 hours. Additionally, an hour of work by a university graduate is more productive than an hour by a high-school dropout, so economists adjust hours by education and experience.

This adjustment is crucial. Some countries achieve high growth by adding less-educated workers (rural-to-urban migration in China). Others achieve growth by raising the skill level of a stable workforce (Germany’s apprenticeship system). Growth accounting disentangles these.

Capital input similarly needs care. A dollar of investment in a semiconductor fab is not the same as a dollar in a shopping mall. Economists weight investment by expected returns and durability. Intangible capital—software, patents, brands, organisational knowledge—was long ignored but is increasingly included in modern accounts.

What is total factor productivity?

The Solow residual, TFP, is a catchall for everything not accounted for by measured factor increases. In practice, this includes:

• Genuine technological progress (better designs, faster computers).
• Improvements in management and organisation (just-in-time inventory, supply-chain optimisation).
• Measurement error in labour and capital (hidden skills, unmeasured capital stock).
• Returns to scale and market structure (monopoly rents or efficiencies from concentration).
• Changes in regulation and institutions (trade liberalisation, labour-market reform).

The Solow residual is not “true” technology alone. It is the measure of our ignorance. If we could perfectly measure all inputs and account for their quality improvements, TFP would shrink. Conversely, if we miss important inputs (say, human health or environmental quality), TFP overstates the true productivity gain.

Historical patterns in TFP growth

In the United States, TFP growth was around 2.5–3% per year from 1948 to the early 1970s, driving robust growth in living standards despite modest capital accumulation and stable labour force participation. From 1973 to the mid-1990s, TFP growth slowed to under 1% per year—the “productivity paradox” that puzzled economists as computers proliferated but output per worker stalled.

The 1995–2005 period saw TFP reaccelerate to 1.5–2% as information technology matured. Since 2010, TFP growth has been sluggish again in many developed economies, despite rapid progress in AI and biotech. This mismatch is one of the central puzzles in modern macroeconomics: either measurement is missing the gains from new technologies, or the productivity impact is truly small relative to hype.

Across countries, TFP growth varies dramatically. Post-World War II Japan and Germany had TFP growth above 3%, reflecting rapid catch-up with frontier technology. Sub-Saharan Africa and Latin America have often had negative or near-zero TFP growth, suggesting factor accumulation without efficiency improvement—or measurement challenges in informal economies.

The limits and critiques

Growth accounting is mechanical. It describes the data but does not explain causation. Suppose a government builds a highway but fails to repair it, and it crumbles within a decade. Capital was accumulated, then lost to depreciation, but TFP might fall if the broken highway reduces trade efficiency. The accounting would be correct but not satisfying to policymakers.

Another limit: labour and capital are not independent. Workers on a modern assembly line are far more productive than workers with hand tools, not because they are intrinsically more skilled but because capital deepening (more machinery per worker) makes them more productive. Growth accounting attributes this to capital, but the underlying causation is tangled.

A deeper critique: the Cobb-Douglas function assumes constant returns to scale and competitive markets. In reality, firms often have monopoly power, investment in R&D generates spillovers that firms cannot fully capture, and increasing returns to scale mean the total product exceeds the sum of marginal products. In such settings, the factor shares that weight the decomposition are not the “true” contributions.

Finally, growth accounting ignores distribution. Two economies might have the same TFP growth but distribute it very differently across workers. An economy in which TFP gains accrue entirely to capital owners is very different from one in which workers share them through wage growth. The aggregate number hides these political and social dimensions.

Growth accounting in policy

Despite its limits, growth accounting is a workhorse in policy analysis. If a country’s productivity growth has collapsed, officials know to investigate: Is capital investment too low? Is the workforce shrinking or aging? Or is TFP growth the culprit—suggesting problems with innovation, management, or institutions?

During the COVID-19 pandemic, growth accounting helped distinguish a supply shock (fewer workers and capital used) from a demand shock (less efficient use of inputs). Many economies initially saw negative TFP growth as factories operated at low capacity. As capacity recovered, TFP rebounded.

Long-run growth projections often start with growth-accounting assumptions: labour force growth (from endogenous-fertility-growth models or demographic projections), capital accumulation (from golden-rule-of-accumulation models), and an assumption about TFP (often conservative—1–2% per year for developed economies). The sum gives a baseline forecast of GDP growth.

See also