Growth Accounting: Output and Input Contributions

Why Decompose Output Growth?

When real output rises, the increase may reflect more labor input, more capital input, or greater output from given quantities of those inputs. Growth accounting organizes these sources into a common calculation. It does not say that any one source is the sole cause of growth; rather, it gives a model-based accounting of how each source contributes.

The basic method uses a production function and weights the growth of each input by its output elasticity. This lets an analyst compare input growth rates with output growth in percentage-point terms. The method is aligned with the SOA VEE Economics (Validation by Educational Experience) syllabus.

The Production Model and Its Assumptions

Let \(Y\) denote real output, \(K\) the quantity of capital services used in production, \(L\) the quantity of labor services, and \(A\) productivity. For this tutorial, use the Cobb–Douglas production function:

Y=A K^{\alpha}L^{1-\alpha}, \qquad 0<\alpha<1.

The parameter \(\alpha\) is the output elasticity with respect to capital: holding the other terms fixed, a 1% increase in \(K\) raises \(Y\) by approximately \(\alpha\) percent. The labor elasticity is \(1-\alpha\). The exponents sum to one, so the model has constant returns to scale: multiplying both inputs by the same factor while holding \(A\) fixed multiplies output by that factor.

The simple decomposition relies on this production form, stable elasticities over the period being studied, and comparable measures of real output and inputs. In applications, the capital and labor weights may be approximated by their income shares when the relevant competitive-market and constant-returns assumptions are reasonable. An income share is not automatically an output elasticity in every setting; for example, departures from the model assumptions may weaken that interpretation.

Definition: Growth accounting decomposes output growth into weighted contributions from productivity growth and input growth. In the Cobb–Douglas model, the weights are the output elasticities \(\alpha\) for capital and \(1-\alpha\) for labor. A contribution is a component of the calculated growth rate, usually reported in percentage points; it is not itself the input’s growth rate.

Deriving the Growth Decomposition

Take natural logarithms of the production function. This converts the product into a sum and makes the role of each exponent explicit:

\ln Y=\ln A+\alpha\ln K+(1-\alpha)\ln L.

For continuous growth rates—or, equivalently, instantaneous percentage growth—differentiate with respect to time. Since the growth rate of a variable \(Z\) is \(d\ln Z/dt\), the result is:

g_Y=g_A+\alpha g_K+(1-\alpha)g_L.

For annual data, this equation is also commonly used as an approximation when changes are not large. Here \(g_Y\), \(g_A\), \(g_K\), and \(g_L\) are expressed in the same units, typically percentages per year. The contributions are \(g_A\) from productivity, \(\alpha g_K\) from capital, and \((1-\alpha)g_L\) from labor. Add the contributions to obtain approximate output growth.

For finite changes, it is useful to distinguish this approximation from an exact relationship. Define \(g_Z=(Z_1-Z_0)/Z_0\), the ordinary percentage change from period 0 to period 1. The production function implies:

1+g_Y=(1+g_A)(1+g_K)^\alpha(1+g_L)^{1-\alpha}.

This exact expression follows by taking the ratio of the production function in period 1 to the function in period 0. Taking logarithms gives an exact additive expression for log changes:

\ln(1+g_Y)=\ln(1+g_A)+\alpha\ln(1+g_K)+(1-\alpha)\ln(1+g_L).

For small changes, \(\ln(1+g)\) is approximately \(g\), which explains why the weighted-growth equation is a good approximation. For larger changes or when exact finite growth is requested, use the multiplicative expression instead. Do not treat the approximation as an exact identity for ordinary percentage changes.

Identify the model and weights.
Write the production function and identify \(\alpha\) and \(1-\alpha\). Confirm that the rates refer to the same period and use the same growth convention.

Calculate each contribution.
Multiply capital growth by \(\alpha\), labor growth by \(1-\alpha\), and include productivity growth directly.

Add and interpret.
Sum the contributions for approximate output growth. Report both the total and the separate contributions, with units such as percentage points per year.

Worked Applications

Worked Example: Calculating Contributions and Exact Growth

A hypothetical economy is described by \(Y=AK^{0.3}L^{0.7}\). Over one year, productivity grows by 1%, capital input by 2%, and labor input by 3%. Calculate the approximate growth contribution of each source and output growth. Then calculate exact finite output growth from the model.

Method. Use the weighted growth equation for the approximate contributions, then use the finite-change production-function ratio for exact growth. This keeps percentage-point accounting separate from exact percentage changes.

The productivity contribution is 1.00 percentage point. The capital contribution is \(0.3(2.00\%)=0.60\) percentage point. The labor contribution is \(0.7(3.00\%)=2.10\) percentage points. Thus approximate output growth is:

1.00+0.60+2.10=3.70\%.

For exact growth, enter the finite growth rates as decimals in the multiplicative formula:

1+g_Y=1.01(1.02)^{0.3}(1.03)^{0.7}.

The input factors are \(1.02^{0.3}\approx1.005958\) and \(1.03^{0.7}\approx1.020907\). Therefore, \(1+g_Y\approx1.01(1.005958)(1.020907)\approx1.037260\), so exact output growth is approximately \(3.726\%\). The exact result is slightly higher than the 3.70% approximation. The check is that the exact multiplier exceeds one and is close to the approximate multiplier \(1.037\).

Worked Example: Solving for Capital Growth

Suppose annual log-growth rates in a production model are \(g_Y=1.2\%\), \(g_A=1.0\%\), and \(g_L=-1.0\%\). The capital elasticity is \(\alpha=0.4\), so the labor elasticity is \(0.6\). Find the capital log-growth rate needed to match output growth.

Method. Substitute the known rates into the additive log-growth decomposition and solve for the missing rate. Because the rates are log-growth rates, the equation is additive in this form.

$$ 1.2=1.0+0.4g_K+0.6(-1.0). $$

The weighted labor term is \(0.6(-1.0)=-0.6\), so the equation becomes \(1.2=1.0+0.4g_K-0.6=0.4+0.4g_K\). Subtract 0.4 from both sides: \(0.8=0.4g_K\). Dividing by 0.4 gives \(g_K=2.0\%\).

Check by substitution: \(1.0+0.4(2.0)+0.6(-1.0)=1.0+0.8-0.6=1.2\%\), as required. A common error is to stop at the wrong constant after combining the productivity and labor terms; writing the intermediate equation makes the arithmetic check visible.

Worked Example: Separating Total Growth From Growth Per Worker

A hypothetical production model has \(\alpha=0.35\). Annual log-growth rates are \(g_A=0.8\%\), \(g_K=2.5\%\), and \(g_L=1.0\%\). Calculate the three contributions to output growth, then find the implied growth rate of output per worker.

Method. First calculate the weighted contributions to total output growth. Then subtract labor growth from output growth to obtain output-per-worker growth, since \(Y/L\) is output per worker.

The labor weight is \(1-\alpha=0.65\). The productivity contribution is 0.800 percentage point; the capital contribution is \(0.35(2.5\%)=0.875\) percentage point; and the labor contribution is \(0.65(1.0\%)=0.650\) percentage point. Total output log growth is therefore \(0.800+0.875+0.650=2.325\%\).

Output-per-worker log growth is \(g_Y-g_L=2.325\%-1.0\%=1.325\%\). As a check, divide the production function by \(L\) to get \(Y/L=A(K/L)^{0.35}\). Its log-growth equation is \(g_Y-g_L=g_A+0.35(g_K-g_L)\). Substitution gives \(0.8+0.35(2.5-1.0)=0.8+0.525=1.325\%\). The two calculations agree.

Interpreting Contributions and Avoiding Traps

The weights apply to input growth, not to output growth. If capital grows by 4% and its elasticity is 0.3, its approximate contribution is 1.2 percentage points—not 4 percentage points. Likewise, the labor contribution is its growth rate multiplied by the labor elasticity. A declining input has a negative contribution when its weight is positive; the other sources may still make total output rise.

A contribution is measured relative to the model and the selected period. It should not be described as a direct causal effect without further assumptions. For instance, a high capital contribution says that capital growth, weighted by the assumed elasticity, accounts for a large component of measured output growth in the decomposition. It does not establish why capital grew or whether the observed relationship would remain unchanged under a different policy.

A common exam trap is to report the growth rate of an input as its contribution, or to use the capital elasticity as the labor weight. Another is to mix ordinary finite percentage changes with log growth rates without noting the difference. Check that the two weights sum to one, that each rate covers the same interval, and that the signed contributions add to output growth under the stated convention.

Key takeaway: Under \(Y=AK^\alpha L^{1-\alpha}\), approximate output growth is productivity growth plus capital growth weighted by \(\alpha\) plus labor growth weighted by \(1-\alpha\). State the model, show the weighted terms, and distinguish approximate growth accounting from exact finite-change calculations.

Check Your Understanding

For each problem, use the stated growth convention and show the weighted contributions before checking the final answer.

In a Cobb–Douglas model, \(\alpha=0.25\), productivity grows at 0.6% per year, capital at 4.0%, and labor at 2.0%. Find approximate output growth. (Answer: 3.10% per year.)
Annual log-growth rates are \(g_Y=1.5\%\), \(g_A=0.5\%\), and \(g_L=-0.5\%\), with \(\alpha=0.4\). Find \(g_K\) and verify the weighted contributions. (Answer: capital growth 3.25%; contributions are 0.50, 1.30, and -0.30 percentage points, summing to 1.50%.)
Let \(\alpha=0.3\), with annual log-growth rates \(g_A=1.2\%\), \(g_K=0.0\%\), and \(g_L=2.0\%\). Find output growth and output-per-worker growth. (Answer: output growth 2.60%; output-per-worker growth 0.60%.)
For \(\alpha=0.4\), ordinary finite growth rates over a period are \(g_A=1\%\), \(g_K=3\%\), and \(g_L=-1\%\). Find approximate growth using the weighted equation and exact growth using the production-function ratio. (Answer: approximate growth 1.60%; exact growth \(1.01(1.03)^{0.4}(0.99)^{0.6}-1\), approximately 1.587%.)
Annual log-growth rates are \(g_Y=2.4\%\), \(g_A=0.4\%\), and \(g_L=1.0\%\), with capital elasticity \(\alpha=0.35\). Solve for capital growth and state its contribution. (Answer: capital growth approximately 3.857%; capital contribution 1.35 percentage points.)

What You'll Learn