Okun's Law
By Sitraka Forler · Lecturer, Durham Business SchoolUpdated 13 September 2026 About this site
Empirical link between unemployment and the output gap.
Each 1 percentage-point rise in unemployment above its natural rate is associated with roughly a 2% fall in real GDP below potential. It is a rule of thumb, not an exact law, but useful for translating labour-market data into growth estimates.
Okun's law: the exchange rate between jobs and output
The blue line is the rule of thumb, output gap = −c × (u − u*). The faint dots are episodes that only loosely obey it.
u is 2.0pts above natural ⇒ output gap = −2.00 × 2.0 = −4.0% of potential GDP
The intuition
Okun's law is the empirical exchange rate between the labour market and output: when unemployment sits one percentage point above its natural rate, GDP tends to sit about two per cent below its potential. Formally, (Y − Y*)/Y* ≈ −c × (u − u*), with c around 2 for the United States. Arthur Okun spotted the pattern in 1962, and it has held loosely ever since. In the chart, the blue line is that rule; the faint dots are the sort of historical episodes it is fitted to, close to the line but never on it. It is a regularity, not a theorem.
Why roughly two for one? Because the unemployment rate understates slack. When demand falls, firms do not only shed workers: they cut hours, freeze hiring, and hoard staff they expect to need later, while discouraged workers stop being counted at all. Output therefore falls by more than the headcount alone suggests, which is why c is about 2 rather than 1. Drag the coefficient slider and the blue line tilts: a steep line means each point of unemployment signals a lot of lost output, typical of labour-hoarding economies like Japan; a flat one means the jobless rate does most of the adjusting.
The law earns its keep in both directions. Forecasters use it to translate a GDP forecast into an unemployment path: growth one point below trend for a year lifts the jobless rate by roughly half a point. Central banks run it the other way, backing an output gap out of the unemployment gap when potential GDP is unobservable. Just respect the scatter in the chart: real episodes miss the line by a point of GDP quite routinely, so treat any Okun number as a first approximation, not a measurement.
Formula / theory
(Y − Y*)/Y* ≈ −c · (u − u*) with c ≈ 2
From a jobless rate to a euro cost, and back
- Find the unemployment gap: u − u* = 6.5% − 4.5% = 2.0 percentage points above the natural rate.
- Apply Okun's law with c = 2: output gap = −2 × 2.0 = −4.0% of potential GDP.
- Convert to money: with potential GDP of €500bn, the shortfall is 0.04 × €500bn = €20bn of output lost per year.
- Sense-check on the chart: set the gap slider to 2.0 and c to 2.00, and the gold dot sits at exactly −4.0% on the blue line.
- Now invert it: if GDP grows 1% below trend for a year, unemployment rises by about 1 ÷ 2 = 0.5 percentage points.
- Stack it: two such years add roughly 2 × 0.5 = 1.0 point to the unemployment rate.
Okun's law is a currency converter between the labour market and GDP: multiply the unemployment gap by c to get the output cost, divide a growth shortfall by c to get the jobs cost, and allow a point or so of slippage either way.
Common pitfalls
- Treating c as a universal constant. It is roughly 2 on post-war US data, but it is larger where firms hoard labour and adjust hours instead of headcount (Japan, much of continental Europe) and smaller where unemployment swings freely, and it drifts across decades. Re-estimate it on recent, local data before you rely on it.
- Feeding structural unemployment through the formula. Okun links cyclical gaps only. If u rises because the natural rate u* itself rose (skills mismatch, demographics, policy), there is no corresponding output loss, and applying −c × (u − u*) with a stale u* estimate manufactures a phantom recession.
- False precision. Real episodes routinely sit a full point of GDP off the fitted line, exactly like the scatter in the chart, so an Okun-implied output gap quoted to one decimal place is decoration. Use it to size the story, then check it against direct output-gap estimates.
- Mixing the gap version with the difference version. The gap form needs estimates of u* and potential GDP; the growth-rate form, Δu ≈ −(g − g*) ÷ c, needs trend growth instead. Lifting a coefficient estimated for one into the formula of the other quietly changes what every number means.
Frequently asked questions
What does Okun's law state?
Okun's law states that for every percentage point unemployment rises above its natural rate, real GDP falls roughly c per cent below potential, with c about 2: (Y − Y*)/Y* ≈ −c × (u − u*). Arthur Okun documented the pattern for the US in 1962. It is an empirical regularity rather than a structural law, but it has proved remarkably durable as a first approximation.
What is the Okun coefficient?
The Okun coefficient is the c in output gap = −c × unemployment gap: the percentage of GDP lost per point of excess unemployment. It is roughly 2 on post-war US data but varies widely, higher in labour-hoarding economies such as Japan, lower where hiring and firing are flexible, and it shifts across decades, so practitioners re-estimate it rather than assume it.
Is Okun's law a real economic law?
No. It is an empirical regularity, a fitted line through noisy episodes, not a relationship derived from theory. The coefficient is unstable across countries and decades, and individual recessions can miss the line by a percentage point of GDP or more, because hours, participation and labour hoarding absorb part of every shock. Economists still use it constantly, but as a rule of thumb with error bars.
How is Okun's law used in practice?
Two main jobs. Forecasters translate growth forecasts into unemployment paths: a year of growth 1% below trend implies unemployment rising about 0.5 points when c = 2. Policy institutions run it in reverse, inferring the unobservable output gap from the unemployment gap, an input to Taylor-rule calculations and cyclically adjusted budget balances. In both uses it supplies speed and transparency, not precision.
Test yourself
Further reading
Where to go deeper. Free means a full, legal copy is online.
A modern cross-country re-estimation of the law's slope (free author PDF).