Tutorials › VEE Economics › Workshop: Growth Rates and Growth Accounting

VEE Economics · Growth, Productivity, and Living Standards · Tutorial 922 of 998

Workshop: Growth Rates and Growth Accounting

Learn to calculate real growth consistently across time and to decompose output and GDP-per-capita growth into interpretable sources.

Intermediate 11 min read

What You'll Learn

  • Convert nominal GDP into real GDP using a price index and calculate its growth rate.
  • Distinguish a one-period growth rate from a cumulative change and a compound annual growth rate.
  • Apply the Cobb-Douglas growth-accounting relation to output, capital, labor, and productivity.
  • Calculate a total factor productivity contribution as a residual when required inputs are given.
  • Decompose GDP-per-capita growth into productivity, capital deepening, and labor input per person.
  • Recognize when percentage-point approximations differ from exact growth-rate calculations.

Why Growth Calculations Need Care

Economic growth questions often require more than subtracting one number from another. A reported rise in nominal GDP may reflect higher prices as well as more real production. A change over several years depends on compounding, not on multiplying one year's growth rate by the number of years. And aggregate output growth can have several sources, whose contributions must be measured and interpreted consistently.

This workshop develops a practical sequence for those calculations: measure real GDP growth, handle compound rates, and then apply growth accounting. The methods are aligned with the SOA VEE Economics (Validation by Educational Experience) syllabus. Growth accounting organizes measured changes; by itself, it does not establish that a particular factor caused the observed growth.

Growth Rates and Real GDP

Definition: For a positive quantity \(X\), its one-period percentage growth rate from period \(t-1\) to period \(t\) is \(g_X=(X_t-X_{t-1})/X_{t-1}\). The percentage growth rate is \(100g_X\). A real quantity is adjusted for changes in prices, so its growth is intended to measure changes in volume rather than the effect of inflation.

If nominal GDP is \(Y^N\) and a price index is \(P\), with the index expressed relative to a base of 100, real GDP in base-period prices is \(Y^R=Y^N/(P/100)\). The index must be used consistently for both periods. If the price index is instead expressed as a ratio, such as 1.20, the corresponding calculation is \(Y^R=Y^N/1.20\).

Nominal growth and inflation do not combine by simple subtraction exactly. If nominal GDP growth is \(g_N\) and the price index growth rate is \(\pi\), then the exact real growth rate is \(g_R=(1+g_N)/(1+\pi)-1\). Subtracting inflation from nominal growth is a useful approximation when the rates are small, but it is not an exact identity.

Worked Example: Separating Real Growth From Price Growth

A hypothetical economy has nominal GDP of $840 billion in Year 1 and $990 billion in Year 2. Its GDP price index, with a base of 100, is 120 in Year 1 and 132 in Year 2. Calculate real GDP in each year and its growth rate. Also compare nominal growth with inflation.

Method. Deflate nominal GDP in each year using the corresponding price index, then calculate the percentage change in real GDP.

Year 1 real GDP. The price-index multiplier is \(120/100=1.20\), so \(840/1.20=700\) billion dollars.

Year 2 real GDP. The multiplier is \(132/100=1.32\), so \(990/1.32=750\) billion dollars.

Real growth. The increase is \(750-700=50\) billion, and the base is Year 1 real GDP: \(50/700=0.0714286\), or approximately 7.14%.

Comparison. Nominal GDP growth is \((990-840)/840=150/840=0.1785714\), or about 17.86%. Inflation, measured by the index, is \((132-120)/120=12/120=0.10\), or 10%. The exact conversion gives \(1.1785714/1.10-1=0.0714286\), or 7.14% real growth. The result agrees with the calculation from real GDP levels. The subtraction approximation, 17.86%-10%=7.86%, is not exact.

Check and trap. The real GDP levels rise from 700 to 750, so real growth must be positive and less than nominal growth in this case. Applying the Year 1 deflator to both nominal GDP figures would fail to remove the change in prices and is the common mistake this method avoids.

Compounding Across Multiple Years

A constant annual growth rate applies to the current level each year. If a quantity grows at rate \(g\) for \(n\) years, its ending value is its starting value multiplied by \((1+g)^n\). The cumulative percentage change is \((1+g)^n-1\), not \(ng\), except as a rough approximation for small rates and short intervals.

Definition: The compound annual growth rate (CAGR) is the constant annual rate that would take an initial value \(X_0\) to a final value \(X_n\) over \(n\) periods. It is \((X_n/X_0)^{1/n}-1\), assuming both values are positive and the periods are equally spaced.

The formula follows by applying the same growth factor in each period: \(X_n=X_0(1+g)^n\). Dividing by \(X_0\) gives \(X_n/X_0=(1+g)^n\). Taking the \(n\)th root and subtracting one gives the CAGR formula. The CAGR summarizes the interval; it does not imply that the actual growth rate was constant in every year.

Worked Example: Compounding Income and Removing Inflation

In a hypothetical case, nominal income per person is $40,000 at the start of a four-year period. It grows by 5% each year, while the price level grows by 3% each year. Find the nominal income after four years, the cumulative nominal increase, and the annual real growth rate.

Method. Compound the nominal income for four years. For real growth, compare the nominal growth factor with the price-level growth factor.

Nominal income path. After Year 1, income is \(40{,}000(1.05)=42{,}000\). After Year 2, it is \(42{,}000(1.05)=44{,}100\). After Year 3, it is \(44{,}100(1.05)=46{,}305\). After Year 4, it is \(46{,}305(1.05)=48{,}620.25\). Equivalently, \(40{,}000(1.05)^4=40{,}000(1.21550625)=48{,}620.25\).

Cumulative nominal increase. The increase is \(48{,}620.25-40{,}000=8{,}620.25\) dollars. Relative to the starting income, that is \(8{,}620.25/40{,}000=0.21550625\), or 21.550625%.

Annual real growth. The exact real growth factor per year is \(1.05/1.03=1.0194174757\). Subtracting one gives a real annual growth rate of 0.0194174757, or approximately 1.94%.

Check and trap. The nominal income rises by more than 5% in total because the 5% applies to a successively larger amount. The annual real rate is less than the 5% nominal rate because prices also rise. Subtracting 3% from 5% gives a 2% approximation, not the exact 1.94% rate.

Growth Accounting

Growth accounting separates output growth into contributions from measured inputs and a productivity residual. A standard setup represents output as \(Y=A K^\alpha L^{1-\alpha}\), where \(Y\) is real output, \(K\) is capital input, \(L\) is labor input, \(A\) is total factor productivity (TFP), and \(0<\alpha<1\) is the capital output elasticity. The formula assumes this Cobb-Douglas production representation and a constant \(\alpha\) over the period being analyzed. Labor input should be defined consistently with the data, for example as hours worked or a quality-adjusted labor measure.

Taking logarithms gives \(\ln Y=\ln A+\alpha\ln K+(1-\alpha)\ln L\). The change in log output is therefore the change in log productivity plus the \(\alpha\)-weighted change in log capital plus the \((1-\alpha)\)-weighted change in log labor. Log changes add exactly. In applications, percentage growth rates are often used as approximations to log changes, especially when rates are modest.

$$ \Delta \ln Y=\Delta \ln A+\alpha\,\Delta \ln K+(1-\alpha)\,\Delta \ln L $$

In the percentage-point approximation, output growth is the sum of the TFP contribution, \(\alpha g_K\), and \((1-\alpha)g_L\). If \(g_Y\), \(g_K\), and \(g_L\) are observed, TFP growth can be calculated as a residual:

$$ g_A \approx g_Y-\alpha g_K-(1-\alpha)g_L $$

The weights matter. Under the stated model, a 1 percentage point increase in capital growth contributes \(\alpha\) percentage points to output growth, not a full percentage point. A residual TFP estimate collects the portion of measured output growth not accounted for by the weighted input growth. It can also reflect measurement error, omitted inputs, or a mismatch between the model and the data; it should not automatically be interpreted as a direct measure of a single technological change.

Worked Example: Calculating TFP and GDP-Per-Capita Contributions

A hypothetical economy has one-year real GDP growth of 3.5%, capital input growth of 6%, labor input growth of 2%, and population growth of 1%. Use a Cobb-Douglas growth-accounting model with \(\alpha=0.35\). Calculate the TFP residual and decompose GDP-per-capita growth into TFP, capital deepening, and labor input per person.

Method. First calculate the weighted capital and labor contributions to aggregate output growth. Then solve for TFP as the residual. For the per-capita decomposition, divide the production function by population.

Aggregate contributions. Capital contributes \(0.35(6\%)=2.10\) percentage points. Labor contributes \((1-0.35)(2\%)=0.65(2\%)=1.30\) percentage points.

TFP residual. The input contributions sum to \(2.10+1.30=3.40\) percentage points. Thus \(g_A\approx3.50-3.40=0.10\%\). The three aggregate contributions add to \(0.10+2.10+1.30=3.50\%\), matching the given output growth.

Per-capita decomposition. Let \(N\) denote population. Dividing \(Y=A K^\alpha L^{1-\alpha}\) by \(N\), or taking growth rates, gives the approximate relation \(g_{Y/N}=g_A+\alpha(g_K-g_L)+(g_L-g_N)\). Here GDP-per-capita growth is \(3.50\%-1.00\%=2.50\%\). Capital deepening contributes \(0.35(6\%-2\%)=1.40\) percentage points. Labor input per person contributes \(2\%-1\%=1.00\) percentage point. Adding the contributions gives \(0.10+1.40+1.00=2.50\%\).

Check and trap. The per-capita contributions sum to the calculated 2.50% growth. Subtracting population growth from aggregate growth gives GDP-per-capita growth, but it does not finish the accounting decomposition. The separate capital-deepening and labor-per-person terms are needed to interpret its sources.

Using the Per-Capita Identity

The identity in the worked example follows from the production function and the definition of output per person. Since \(Y/N=A(K/L)^\alpha(L/N)\), output per person depends on TFP, capital per unit of labor input, and labor input per person. Taking growth rates yields the stated decomposition. This makes clear why GDP-per-capita growth is not necessarily the same as labor-productivity growth: labor input per person may also change.

Key result: Under the Cobb-Douglas assumptions above, the percentage-point approximation is \(g_{Y/N}\approx g_A+\alpha(g_K-g_L)+(g_L-g_N)\). The terms are TFP growth, capital deepening per unit of labor input, and labor input per person, respectively. Use the same definition of labor input throughout.

For example, if employment rises faster than population, labor input per person can increase and contribute to GDP-per-capita growth even if productivity measures are unchanged. Conversely, labor input per person can fall and offset some gains from TFP or capital deepening. The decomposition is an accounting framework, not a claim that each contribution is an independently identified causal effect.

Common Mistakes and Practical Checks

  • Using nominal GDP as real GDP. Deflate each year's nominal value with that year's price index before calculating real growth.
  • Adding annual growth rates over a long interval. Apply the growth factor repeatedly, or use the CAGR formula to summarize the interval.
  • Subtracting inflation for an exact real rate. Use \((1+g_N)/(1+\pi)-1\) for exact conversion; subtraction is an approximation.
  • Omitting the factor weights. In the Cobb-Douglas approximation, multiply capital growth by \(\alpha\) and labor growth by \(1-\alpha\).
  • Confusing aggregate growth with per-capita growth. Subtract population growth to obtain the approximate per-capita rate, then use the per-capita identity if contributions are requested.
  • Calling the residual pure technology. TFP is the unaccounted-for component under the selected model and measurements; it can include other influences and data limitations.

Before finalizing a calculation, check the base used in each percentage change, confirm that the price index has the correct scale, and verify that reported contribution terms sum to the stated total (allowing for rounding). Report growth contributions in percentage points so they are not confused with percentage changes in the contributions themselves.

Check Your Understanding

For each problem, calculate the requested rate or contribution and check that the result is consistent with the given data.

  1. A hypothetical economy's nominal GDP rises from $1,200 billion to $1,260 billion. Its GDP price index, base 100, rises from 150 to 157.5. Calculate real GDP in both years and real GDP growth. (Answer: $800 billion in each year; 0.00% real growth.)
  2. Hypothetical real income per person rises from $40,000 to $46,305 over three years. Calculate its compound annual growth rate. (Answer: 5.00% per year.)
  3. In a Cobb-Douglas model, \(\alpha=0.40\), capital input grows by 5%, labor input by 1%, and TFP by 1.5%. Find approximate real output growth and show the weighted contributions. (Answer: 4.10% growth; capital 2.00 percentage points, labor 0.60 percentage points, TFP 1.50 percentage points.)
  4. Real output grows by 4.2%, population by 1.2%, capital input by 4%, and labor input by 2%. With \(\alpha=0.30\), calculate the TFP residual and GDP-per-capita growth. (Answer: TFP growth 1.60%; GDP-per-capita growth 3.00%.)
  5. Nominal GDP grows by 8% while the price index grows by 3%. Calculate exact real GDP growth and compare it with the subtraction approximation. (Answer: 4.8544% exact; 5.00% by subtraction, an overstatement of about 0.1456 percentage points.)
Clinical Biostats