Unemployment, Output Gaps, and Okun’s Law

Why Link Output Gaps and Unemployment?

When an economy produces below its sustainable capacity, firms may need fewer workers or fewer hours of work than when output is near capacity. Unemployment often rises relative to its usual level as a result. The relationship is useful for interpreting business-cycle data: it provides a disciplined way to relate an output shortfall to cyclical unemployment, rather than treating either measure in isolation.

That relationship is summarized by Okun’s law. It is an empirical regularity, not an accounting identity or a fixed technological law. Its coefficient must be interpreted in context: estimates can differ across economies and periods, and a relationship estimated in one setting should not automatically be transferred to another.

Definition: Let $Y$ be actual real output and $Y^*$ be potential output, an estimate of the output the economy could sustain under normal conditions. The output gap, expressed as a percentage, is $G=100(Y-Y^*)/Y^*$. A negative gap means actual output is below potential; a positive gap means it is above potential. Let $u$ be the actual unemployment rate and $u^*$ a reference or trend unemployment rate. Cyclical unemployment is $u-u^*$, measured in percentage points.

The benchmark $u^*$ is not the same thing as zero unemployment. It is a reference rate for separating the cyclical component from other influences on unemployment. Both potential output and the reference unemployment rate are estimated rather than directly observed, so calculated gaps and cyclical unemployment inherit uncertainty from those estimates.

The Gap Form of Okun’s Law

In the gap form used here, the unemployment rate is expressed in percentage terms, and the output gap $G$ is expressed as a percentage. The coefficient $\beta$ states how many percentage points cyclical unemployment changes for a one-percentage-point change in the output gap.

Okun’s law, gap form: For a specified economy and period, a simple approximation is $u-u^*=-\beta G$, where $\beta>0$ is an estimated coefficient. The minus sign represents the usual inverse relationship: a negative output gap is associated with unemployment above its reference rate, while a positive output gap is associated with unemployment below it. The coefficient is not universal.

For example, if $\beta=0.40$, a one-percentage-point increase in the output gap is associated with a 0.40-percentage-point decrease in cyclical unemployment, according to this approximation. It does not mean unemployment changes by 40 percent. The units matter: the output gap is in percent, and the unemployment difference is in percentage points.

Before applying the equation, check that the question supplies a gap-form coefficient and that the rates and gaps use compatible units. In particular, a gap of $-3\%$ is entered as $-3$, not as $-0.03$, when $\beta$ is stated in percentage points of unemployment per one percentage point of output gap.

Why the Relationship Has a Negative Sign

A simplified labor-input argument explains the expected direction of the relationship. Suppose, for this illustration only, that the labor force is fixed, output per employed worker is constant, and potential output corresponds to employment at the reference unemployment rate. Write the unemployment rates as fractions between zero and one. With a fixed labor force, employment is proportional to one minus the unemployment rate. Under these assumptions, output relative to potential is proportional to employment relative to potential employment:

\frac{Y}{Y^*}=\frac{1-u}{1-u^*}

Subtract one from both sides. The left side becomes the output gap as a fraction of potential output. On the right, the numerator after subtracting one is $1-u-(1-u^*)=u^*-u$. Thus:

\frac{Y-Y^*}{Y^*} = \frac{Y}{Y^*}-1 = \frac{u^*-u}{1-u^*} = -\frac{u-u^*}{1-u^*}

This toy model gives an inverse relationship: if unemployment exceeds its reference rate, output is below potential. It is not a derivation of a universal empirical coefficient. In actual economies, hours per worker, productivity, labor-force participation, and employers’ adjustment decisions can also change. Those influences help explain why the measured coefficient differs from the simple model and varies across settings.

Applying Okun’s Law

Worked Example: From a Negative Output Gap to Unemployment

A hypothetical economy has potential output of $640 billion and actual output of $620.8 billion. A study for this economy estimates $\beta=0.35$ percentage points of unemployment per one-percentage-point output gap. The reference unemployment rate is 4.8%. Use the gap form of Okun’s law to estimate the output gap, cyclical unemployment, and actual unemployment rate.

Method. First calculate the output gap using potential output as the denominator. Then apply the estimated coefficient and add the cyclical component to the reference rate.

Step 1: Calculate the output gap. The difference between actual and potential output is $620.8-640=-19.2$ billion dollars. Therefore:

G=100\frac{620.8-640}{640} =100\frac{-19.2}{640} =-3.00\%

Step 2: Calculate cyclical unemployment. Use the stated gap-form coefficient and retain the negative sign of the gap:

u-u^*=-\beta G =-(0.35)(-3.00) =1.05\text{ percentage points}

Step 3: Find the actual unemployment rate. Add the cyclical component to the reference rate: $u=4.8\%+1.05$ percentage points $=5.85\%$.

Check. Output is below potential, so the model predicts unemployment above its reference rate. The estimate of 5.85% has that direction. The common mistake of dropping the minus sign would instead imply unemployment below 4.8%, contradicting the negative output gap.

Worked Example: Inferring the Output Gap From Unemployment

In a hypothetical economy, the current unemployment rate is 6.2%, the reference rate is 5.0%, and the estimated gap-form coefficient is $\beta=0.30$. Potential output is $900 billion. Assuming the coefficient applies to this period, estimate the output gap and actual output.

Method. Calculate cyclical unemployment first, then rearrange Okun’s law to solve for the output gap. Finally, use the definition of the gap to recover actual output.

Step 1: Calculate cyclical unemployment.

u-u^*=6.2\%-5.0\%=1.2\text{ percentage points}

Step 2: Solve for the output gap. Starting from $u-u^*=-\beta G$, divide by $-\beta$:

G=-\frac{u-u^*}{\beta} =-\frac{1.2}{0.30} =-4.0\%

Step 3: Convert the gap into output. Since $G=100(Y-Y^*)/Y^*$, a gap of $-4.0\%$ means actual output is $96.0\%$ of potential:

Y=0.96(900\text{ billion}) =864\text{ billion dollars}

Check. Unemployment is above its reference rate, so the inferred gap should be negative; $-4.0\%$ is consistent. A frequent error is to divide by the unemployment rate or multiply by $\beta$ when rearranging. The equation requires division of the unemployment difference by $-\beta$.

Worked Example: A Narrower Gap During Output Growth

A hypothetical economy’s potential output rises from $1,000 billion to $1,030 billion over a year. Its output gap moves from $-4.0\%$ to $-1.0\%$. The reference unemployment rate stays at 5.0%, and the estimated coefficient is $\beta=0.40$. Calculate actual output in each period, actual output growth, and the change in unemployment predicted by Okun’s law.

Method. Convert each gap into an output level using the potential output for that same period. Calculate output growth from the two actual output levels. Separately apply Okun’s law to each gap; a change in the gap is not itself the rate of output growth.

Step 1: Calculate actual output in the first period. A $-4.0\%$ gap means output is $96.0\%$ of potential:

Y_0=0.96(1{,}000)=960\text{ billion dollars}

Step 2: Calculate actual output in the second period. A $-1.0\%$ gap means output is $99.0\%$ of that period’s potential:

Y_1=0.99(1{,}030)=1{,}019.7\text{ billion dollars}

Step 3: Calculate actual output growth. The increase is $1{,}019.7-960=59.7$ billion dollars. Relative to initial actual output:

100\frac{1{,}019.7-960}{960} = 100\frac{59.7}{960} =6.21875\%

Rounded to two decimal places, actual output growth is 6.22%.

Step 4: Calculate predicted unemployment in each period. In the first period, cyclical unemployment is $-0.40(-4.0)=1.60$ percentage points, so $u_0=5.0\%+1.60=6.60\%$. In the second period, it is $-0.40(-1.0)=0.40$ percentage points, so $u_1=5.0\%+0.40=5.40\%$.

Step 5: Calculate the change in unemployment. The predicted change is $5.40\%-6.60\%=-1.20$ percentage points. Unemployment falls by 1.20 percentage points as the gap narrows.

Check. Actual output grew while the economy remained below potential in both periods. A shrinking negative gap indicates recovery relative to potential; it does not say output is still falling. The common mistake is to report the three-percentage-point change in the gap as output growth. Output growth must be calculated from the output levels, which also reflect growth in potential output.

Interpretation and Limits

Okun’s law is useful for organizing evidence, making approximate calculations, and comparing the direction and scale of cyclical changes. It does not, by itself, identify the cause of an output gap, prove that a change in unemployment caused a change in output, or establish what policy would eliminate the gap. The equation summarizes a relationship in observed data; it is not a complete model of the labor market or the business cycle.

The coefficient can change because economies differ in labor-market institutions, production structures, and the ways firms adjust employment, hours, and productivity. Even within one economy, the relationship may differ across periods. A temporary disruption, for example, may affect hours worked differently from a broad decline in demand. Measurement also matters: potential output and the reference unemployment rate are estimates, and revisions to those estimates can change the calculated gap or cyclical component.

For these reasons, use a coefficient only with the definition, units, economy, and period specified in the problem. If a question supplies a coefficient, treat it as the applicable estimate for that calculation rather than as a universal constant. If no coefficient is supplied, do not invent a precise numerical prediction.

Key takeaway: In the gap form, cyclical unemployment equals minus the estimated coefficient times the output gap. A negative gap therefore corresponds to unemployment above its reference rate in the model. Correct interpretation depends on units, sign, and the limits of the empirical estimate.

Check Your Understanding

For each problem, use the gap form of Okun’s law and show the sign and units in your calculation.

A hypothetical economy has actual output of $873 billion and potential output of $900 billion. Its coefficient is $\beta=0.32$, and its reference unemployment rate is 4.6%. Find the output gap, cyclical unemployment, and predicted actual unemployment rate. (Answer: gap $-3.0\%$; cyclical unemployment $+0.96$ percentage points; unemployment rate $5.56\%$)
Unemployment is 7.1%, the reference rate is 5.6%, and $\beta=0.375$. If potential output is $750 billion, find the implied output gap and actual output. (Answer: gap $-4.0\%$; actual output $720 billion)
A hypothetical economy’s output gap changes from $+1.5\%$ to $-2.5\%$. The reference rate is 5.2%, and $\beta=0.40$. Find predicted unemployment in each period and the change in the rate. (Answer: first period $4.60\%$; second period $6.20\%$; increase $1.60$ percentage points)
Potential output rises from $1,000 billion to $1,030 billion while the output gap moves from $-3.0\%$ to $-1.0\%$. Calculate actual output growth. With $\beta=0.30$ and a constant reference unemployment rate of 4.8%, also calculate the change in predicted unemployment. (Answer: actual output growth $5.12\%$; unemployment falls by $0.60$ percentage points)
In one period, the output gap is $-2.0\%$ and cyclical unemployment is $+0.90$ percentage points. Use these data to estimate $\beta$. If the same estimate applies when the output gap is $+1.2\%$, and the reference rate is 5.0%, find predicted unemployment. (Answer: $\beta=0.45$; predicted unemployment $4.46\%$)

What You'll Learn