Growth accounting decomposes real output growth into contributions from measured production inputs and a residual total factor productivity component.
Growth accounting is a framework that decomposes real output growth into contributions from measured production inputs and a residual total factor productivity component. It helps answer whether an economy or industry grew mainly by using more labor and capital or by producing more than measured input growth can explain.
The framework is an accounting decomposition, not proof that any one factor caused growth. Its conclusions depend on the production model, data, prices, input boundaries, and assumptions used to convert input growth into output contributions.
For a simplified value-added model with capital and labor:
where:
Y is real output;K is capital input;L is labor input;s_K and s_L are capital and labor cost shares; andA is total factor productivity.Rearranging the formula makes TFP the residual:
The simple equation assumes a consistent production boundary and commonly relies on constant returns to scale and competitive factor-pricing assumptions so observed shares can stand in for output elasticities. Applied systems may use more flexible index-number methods and changing shares.
Assume real output grows by 5.0% during a year. Capital services grow by 4.0% and receive a 40% cost share, while labor input grows by 2.0% and receives a 60% share.
The approximate capital contribution is:
The approximate labor contribution is:
Residual TFP growth is:
The decomposition attributes 1.6 percentage points to measured capital-input growth, 1.2 points to measured labor-input growth, and 2.2 points to the residual. These are approximate growth contributions, not shares of revenue or proof of causal effects.
Growth accounting can also explain growth in output per labor hour. A simplified decomposition is:
This version separates:
Exact agency formulas and classifications differ. Some models incorporate labor composition inside labor input; others present it separately when decomposing labor productivity.
| Framework | Output measure | Inputs commonly included | Best use |
|---|---|---|---|
| Value-added growth accounting | Real value added | Capital and labor | Broad sectors and economy-wide analysis |
| Gross-output or sectoral-output model | Output before deducting all purchased intermediate inputs | Capital, labor, energy, materials, and services | Detailed industry production analysis |
A gross-output KLEMS model accounts explicitly for purchased intermediate inputs. A value-added model nets those inputs from output and focuses on capital and labor. Their TFP residuals answer related but different questions and should not be compared without reviewing definitions.
Economic analysis: The framework distinguishes growth through factor accumulation from residual productivity growth. That difference matters for assessing whether a growth path depends on continued expansion of labor and capital.
Industry research: Analysts can compare whether sectors grew through investment, labor input, intermediate inputs, or a residual efficiency component.
Forecasting: Long-run assumptions about labor, capital formation, and productivity can be made explicit rather than embedded in one output-growth rate. Forecasts remain uncertain and should use scenarios.
Business interpretation: The logic can organize analysis of automation, capacity, and workforce changes. Public company data usually do not support an official TFP decomposition, so firm-level estimates require caution and transparent proxies.
The TFP residual may include technological change, management, organizational improvements, resource reallocation, scale effects, capacity utilization, omitted inputs, and measurement error. Calling it a technology contribution overstates what the calculation alone establishes.
Cost-share weighting also relies on economic assumptions. Market power, markups, adjustment costs, regulation, taxes, and incomplete measurement can weaken the link between observed factor payments and true output elasticities.
Timing presents another problem. A large investment may add measured capital services before workers fully learn the system, or organizational improvements may occur before recorded output responds. Annual residuals can therefore be noisy even when longer-run trends are meaningful.
Growth accounting is an educational analytical framework. It does not provide a personalized investment recommendation or a definitive causal policy conclusion.