Why Decompose GDP Per Capita?
Real GDP per capita is a useful summary of average economic output relative to population. But a change in this measure does not, by itself, tell us which part of the economy changed. Average output per employed worker might rise, the employed share of the population might increase, or both might happen. Separating these components helps explain how changes in production per worker and labor utilization relate to GDP per capita.
The decomposition is an accounting identity: it follows from how the quantities are defined, not from a claim about what causes economic growth. For comparisons over time, use real GDP so that price changes alone are not mistaken for changes in output. Match GDP, employment, and population to the same period and geographic area.
The Accounting Identity
The key is to multiply by a ratio equal to one, \(E/E\). This inserts employment into the GDP-per-capita ratio without changing its value:
Thus, GDP per capita equals output per employed worker multiplied by the employed share of the population. The first component describes average output associated with each employed person; the second describes how many employed people there are relative to the population. The product has units of output per person: output per worker multiplied by workers per person.
This identity uses the total population as the denominator of the labor-utilization component. It is not the employment-to-labor-force ratio, which has the labor force as its denominator, nor is it necessarily an employment rate for a particular age group. If a problem specifies a different population group, use that group consistently in both the per-capita measure and the employment share.
Confirm that real GDP, employment, and population refer to the same place and period. Use compatible units.
Divide real GDP by the number of employed people, \(Y/E\).
Divide employment by population, \(E/N\).
Multiply the two components and check that the result equals \(Y/N\).
Worked Example: Separating the Components in Two Years
A hypothetical economy reports real GDP of $600 billion, a population of 30 million, and 15 million employed people in Year 1. In Year 2, real GDP is $693 billion, population is 33 million, and employment is 16.5 million. Calculate GDP per capita, output per employed worker, and the employed share in each year. Then compare their growth.
Method. Calculate each component from the matched level data, then compare Year 2 with Year 1. The GDP figures are real, so the comparison is not driven just by a change in the price level.
Year 1. Convert the GDP-to-worker ratio into dollars per worker:
The employed share is \(15/30=0.5\), or 50%. Therefore, GDP per capita is \(\$40{,}000\times0.5=\$20{,}000\). Directly, \(\$600\text{ billion}/30\text{ million}=\$20{,}000\), which checks the identity.
Year 2. Output per employed worker is \(\$693\text{ billion}/16.5\text{ million}=\$42{,}000\). The employed share is \(16.5/33=0.5\), again 50%. GDP per capita is \(\$42{,}000\times0.5=\$21{,}000\). Direct division gives \(\$693\text{ billion}/33\text{ million}=\$21{,}000\).
GDP per capita rose by \((21{,}000/20{,}000)-1=0.05\), or 5%. Output per worker also rose by \((42{,}000/40{,}000)-1=5\%\), while the employed share did not change. In this example, the entire proportional increase in GDP per capita corresponds to the increase in output per worker.
Common mistake avoided: Dividing Year 2 GDP by Year 1 population would mix periods and produce a ratio that is not GDP per capita for either year. Every ratio here uses quantities from the same year.
Decomposing Growth
The level identity also gives an exact relationship between growth rates. Let \(g_c\) be the proportional growth rate of GDP per capita between two periods, \(g_p\) the growth rate of output per employed worker, and \(g_u\) the growth rate of the employed share. Each growth rate is calculated as the later value divided by the earlier value, minus one. Since the level of GDP per capita is the product of the other two measures, their growth factors multiply:
Expanding the right-hand side gives the exact growth decomposition:
The final term, \(g_pg_u\), is the interaction between the two changes. Simply adding the two component growth rates leaves out this term. When both changes are small, the interaction may be small too, but the exact formula includes it. If the employment share falls while output per worker rises, the interaction term is negative.
There is also an exact additive decomposition using logarithms. Taking the natural logarithm of the level identity and subtracting the earlier period from the later period yields:
This works because the logarithm of a product is the sum of the logarithms. Log changes are often used to express growth contributions additively. For a finite change, a log change is not exactly the same number as the ordinary percentage change; the multiplicative growth-factor formula above gives the exact relationship among ordinary rates.
Worked Example: Combining Two Positive Growth Rates
In a hypothetical economy, output per employed worker rises from $50,000 to $54,000, while the employed share of the population rises from 60% to 62.5%. Find the percentage change in GDP per capita and show why adding the component growth rates alone is not exact.
Method. Calculate the proportional change in each component, then multiply their growth factors. As a level check, GDP per capita initially is \(\$50{,}000\times0.60=\$30{,}000\). Its later level is \(\$54{,}000\times0.625=\$33{,}750\).
Output-per-worker growth is \((54{,}000/50{,}000)-1=0.08\), or 8%. The employed share rises from 0.60 to 0.625, so its growth is \((0.625/0.60)-1=0.0416667\), or about 4.1667%.
Multiplying the growth factors gives \(1.08\times1.0416667=1.125\). Therefore GDP per capita rises by 12.5%. This agrees with its level change: \((33{,}750/30{,}000)-1=12.5\%\).
The interaction is \(0.08\times0.0416667=0.0033333\), or about 0.3333 percentage points. Adding the two component rates and this interaction gives \(8\%+4.1667\%+0.3333\%=12.5\%\).
Common mistake avoided: Adding 8% and 4.1667% gives only 12.1667%. That omits the positive interaction term, so it is not the exact change in GDP per capita.
Worked Example: Higher Output per Worker but Unchanged GDP per Capita
A hypothetical economy has real GDP of $1,200 billion, population of 40 million, and employment of 24 million in its first period. In a later period, real GDP is $1,320 billion, population is 44 million, and employment remains 24 million. Determine the change in GDP per capita and decompose it into changes in output per worker and the employed share.
Method. Calculate both components and GDP per capita in each period. This makes it possible to distinguish a change in productivity from a change in labor utilization.
Initially, output per worker is \(\$1{,}200\text{ billion}/24\text{ million}=\$50{,}000\). The employed share is \(24/40=0.60\), or 60%, so GDP per capita is \(\$50{,}000\times0.60=\$30{,}000\). Direct division gives \(\$1{,}200\text{ billion}/40\text{ million}=\$30{,}000\).
In the later period, output per worker is \(\$1{,}320\text{ billion}/24\text{ million}=\$55{,}000\). The employed share is \(24/44=6/11\), approximately 0.54545, or 54.545%. GDP per capita is \(\$55{,}000\times6/11=\$30{,}000\). Direct division confirms the result: \(\$1{,}320\text{ billion}/44\text{ million}=\$30{,}000\).
Output per worker grew by \((55{,}000/50{,}000)-1=10\%\). The employed share changed by \(((6/11)/0.60)-1=-1/11\), approximately \(-9.0909\%\). Applying the exact formula, \(1.10\times(10/11)=1\), so GDP per capita had zero growth. In the expanded formula, \(10\%-9.0909\%-0.9091\%=0\%\), where \(-0.9091\) percentage points is the interaction term.
Common mistake avoided: Concluding that GDP per capita increased by 10% because output per worker increased by 10% ignores the fall in the employed share. In this example, the two components together leave GDP per capita unchanged.
Interpretation and Common Traps
This decomposition clarifies what GDP per capita does and does not show. It does not identify a cause for a change in output per worker or in the employed share. It organizes the accounting: one component is output per employed person, and the other is employed people relative to the population. A rise in one component can be partly or fully offset by a fall in the other.
- Using nominal GDP for a real growth comparison. A rise in nominal GDP can reflect higher prices rather than more real output. Use real GDP when interpreting changes in real GDP per capita.
- Using the labor force as the denominator by mistake. The identity here uses \(E/N\), employed people divided by the specified population. Substituting employment divided by the labor force changes the measure and breaks this particular decomposition.
- Mixing periods, areas, or population definitions. The employment count must correspond to the GDP and population used. If the population definition changes, the resulting employed shares are not directly comparable without addressing that change.
- Adding percentage changes without the interaction. For ordinary growth rates, use \((1+g_p)(1+g_u)-1\), or include \(g_pg_u\) in the expanded formula.
- Confusing the identity with a causal explanation. The equation accounts for the level and its changes, but does not establish why either component changed.
The distinction matters when interpreting economic performance. Two economies with the same GDP per capita can combine different output per worker and employed shares. Likewise, GDP per capita can remain stable even when its components move substantially in opposite directions. Keeping the components visible prevents a summary statistic from being mistaken for a full account of what changed.
Check Your Understanding
Work through each problem by identifying the two components before calculating or interpreting their changes.
- A hypothetical economy has real GDP of $900 billion, population of 30 million, and employment of 18 million. Calculate output per employed worker, the employed share, and GDP per capita. (Answer: $50,000 per worker; 60%; $30,000 per person.)
- GDP per capita is $24,000 and output per employed worker is $40,000. If population is 50 million, find the employed share and the number employed. (Answer: employed share = 60%; employment = 30 million.)
- Output per worker grows by 3%, while the employed share falls by 2% of its initial value. Find the exact growth rate in GDP per capita. (Answer: \(1.03\times0.98-1=0.0094\), or 0.94% growth.)
- Output per worker rises from $64,000 to $68,000, while the employed share falls from 62.5% to 60%. Calculate initial and final GDP per capita, and verify the exact growth decomposition. (Answer: $40,000 initially and $40,800 finally; GDP per capita grows 2%, from 6.25% productivity growth and -4% utilization growth with a -0.25 percentage-point interaction.)
- A hypothetical economy's real GDP rises from $2.4 trillion to $2.52 trillion. Its population rises from 60 million to 63 million, and employment rises from 36 million to 37.8 million. Calculate output per worker, the employed share, and GDP per capita in both periods. What changed? (Answer: output per worker is $66,666.67 in both periods; the employed share is 60% in both; GDP per capita is $40,000 in both. Neither component nor GDP per capita changed.)