How this calculator works
Okun's law links the unemployment rate to the gap between actual and potential output. In the gap version, unemployment above its natural rate signals spare capacity, and each percentage point of excess unemployment corresponds to roughly two percentage points of lost output when the coefficient is 2. In the growth version, GDP growing faster than its trend reduces unemployment, with a coefficient of about 0.5 in the United States: a one-point rise in growth above trend cuts unemployment by half a point. The coefficient is empirical, not a law of nature; estimates vary by country, period and how potential output is measured, which is why the tool lets you supply your own.
Formula
Gap version: output gap = -beta x (unemployment rate - natural rate), where all rates are fractions. Growth version: change in unemployment = -beta x (actual GDP growth - trend growth). Rearranged, beta = -output gap / unemployment gap or beta = -change in unemployment / growth gap.
Worked example
Unemployment at 5% with a natural rate of 4% and a coefficient of 2 gives an unemployment gap of 1 point and an output gap of -2%, meaning output is 2% below potential. In the growth version, GDP growing at 2% against a 2.5% trend with a coefficient of 0.5 raises unemployment by 0.25 percentage points.
Advertisement
Frequently asked questions
Why is the coefficient different between the two versions?
The versions measure different things. The gap version relates a level of unemployment to a level of lost output, while the growth version relates a change in growth to a change in unemployment. Their coefficients are not interchangeable.
Is the Okun coefficient constant?
No. Estimates for the United States have ranged roughly from 1.5 to 3 for the gap version and from 0.3 to 0.7 for the growth version, and they shift with labour market structure, demographics and how potential output is estimated.
Does low unemployment necessarily mean the economy is overheating?
Not by itself. Okun's law describes an average historical relationship, not a causal rule. Participation changes, hours worked and productivity can all move output without moving unemployment one for one.