Skip to main content

What is Okun's law?

Okun's law is an empirical relationship between unemployment and economic growth discovered by economist Arthur Okun in 1962. It states that for every 1% the economy grows above its potential (trend) growth rate, unemployment falls by roughly 0.5 percentage points. Conversely, for every 1% the economy contracts below potential, unemployment rises by roughly 0.5 percentage points.

Okun's law is not a physics law—it is a statistical pattern observed repeatedly in historical data. It answers a simple question: when the economy grows, how much does unemployment fall? The answer: roughly half a percentage point per percentage point of growth. This relationship has held broadly across decades and countries, though the exact coefficient varies.

Okun's law matters because it connects two of the most important economic variables: the growth rate of output (GDP) and the unemployment rate. Policymakers, economists, and investors use it to forecast how recessions will affect employment, and to estimate how much growth is needed to meaningfully reduce unemployment. In recessions, Okun's law helps predict the unemployment spike. In recoveries, it helps estimate how many jobs will be created.

Quick definition: Okun's law quantifies that 1% excess growth (growth above potential) reduces unemployment by roughly 0.5 percentage points; the relationship between output growth and unemployment changes.

Key takeaways​

  • Okun's law states that for each 1% of growth above potential, unemployment falls by roughly 0.5 percentage points (the coefficient is approximately 0.5).
  • The relationship reflects the fact that firms do not cut or hire workers proportionally to output changes; there is a lag and deadweight loss.
  • Okun's law allows economists to estimate unemployment effects of recessions before jobs data arrives (output data comes faster than employment data).
  • The coefficient has varied over time—in the 1960s it was 0.3, in the 1990s it rose to 0.5, in the 2010s it fell to 0.3 again, suggesting the relationship is not stable.
  • When growth equals potential growth, unemployment stays flat (zero net job creation or loss). When growth exceeds potential, unemployment falls. When growth is below potential, unemployment rises.
  • The relationship breaks down in severe recessions and unusual years (like 2020), when output collapses faster than unemployment rises.

The basic arithmetic of Okun's law​

The simplest form of Okun's law can be written as:

Change in unemployment = -0.5 × (Growth rate - Potential growth rate)

Or equivalently:

Unemployment rate change = -0.5 × Output gap

Here is what this means. Imagine potential growth (the trend) is 2% and actual growth is 3%. The output gap is 1% (actual exceeds potential). Okun's law predicts unemployment will fall by 0.5 percentage points (- 0.5 × 1 = -0.5).

If actual growth is 2% (matching potential), the output gap is zero and unemployment stays flat.

If actual growth is 0% (a recession) and potential is 2%, the output gap is -2%. Unemployment rises by 1 percentage point (-0.5 × -2 = +1).

If actual growth is -2% (a severe contraction) and potential is 2%, the output gap is -4%. Unemployment rises by 2 percentage points (-0.5 × -4 = +2).

This arithmetic makes Okun's law easy to use as a forecasting tool. Once you estimate potential growth and project actual growth, you can forecast unemployment changes without waiting for employment data.

Why the relationship exists: production functions and labor demand​

Okun's law emerges from how firms operate. When output rises, firms need more workers. But the relationship is not one-to-one. Here's why:

Hoarding and lag. When firms see output rising, they do not immediately hire proportionally. They first exhaust slack: workers who are underutilized during normal times but become fully productive when demand rises. A manager working at 70% capacity during slow growth moves to 95% capacity during boom without hiring. A worker on a half-time schedule moves to full-time without a new hire. This hoarding means initial output growth does not require proportional hiring.

Expectations and uncertainty. Firms are unsure if demand will persist. An unexpected surge in orders might be temporary. Firms hesitate to hire permanently until they are confident demand has truly increased. This lag means employment responds slowly to output changes.

Productivity improvements. During recessions, firms cut staff aggressively even if demand only falls slightly. In recovery, they are reluctant to rehire. Once demand is clearly rising, they rehire. The path is not smooth; there is deadweight loss (unemployment during recovery before demand is fully confident).

Fixed and variable costs. Firms have fixed costs (rent, management salaries, fixed equipment) that do not scale with output. A small output increase might be absorbed without hiring because fixed costs are already covered. Only when output is clearly, persistently higher do firms justify hiring.

These factors combine to create the 0.5 coefficient. Each percentage point of excess growth translates to half a percentage point of unemployment reduction, not one-to-one.

How Okun's law has evolved over time​

When Okun discovered the relationship in 1962, the coefficient was approximately 0.3: each 1% of excess growth reduced unemployment by 0.3 percentage points. In the 1960s, firms were more willing to hire when demand rose and more reluctant to fire when it fell, producing a tighter relationship.

By the 1990s, the coefficient had risen to about 0.5 or higher. Why? Globalization and outsourcing made labor more flexible. Firms could more easily cut or expand workforces in response to demand changes. Workers also became more mobile, moving between jobs more frequently. The relationship tightened.

By the 2010s, the coefficient had fallen back to around 0.3 to 0.4. Why? Gig work and contractor arrangements decoupled employment from official statistics. A firm might respond to demand increases by hiring independent contractors rather than employees, which does not show up in traditional employment data. Automation also means that output growth can occur without proportional employment growth. A factory producing 10% more output with the same workers shows output growth but no employment gain. Labor force participation also declined in the 2010s, meaning the pool of available workers was shrinking, so firms could respond to demand with hours rather than hires.

These shifts suggest Okun's law is not immutable. The coefficient depends on labor market institutions, technology, and global economic structure. The modern coefficient is probably closer to 0.3 to 0.4, implying each 1% of excess growth reduces unemployment by 0.3 to 0.4 percentage points rather than 0.5.

Using Okun's law to forecast unemployment in recessions​

Okun's law is most useful in recessions when you want to forecast unemployment before employment data arrives. Real GDP is released about 30 days after the quarter ends; employment data is released monthly. So if a recession begins, GDP numbers showing the contraction arrive faster than unemployment numbers showing the spike. (See Federal Reserve economic data for real-time GDP and unemployment data.)

Using Okun's law, economists can estimate the unemployment impact:

Example: 2008 recession. Real GDP fell at a 4.3% annualized rate in Q4 2008 (a -4.3% quarter). Potential growth was roughly 2.5%. The output gap was -6.8%. Using Okun with coefficient 0.5, the predicted unemployment change is -0.5 × -6.8 = +3.4 percentage points. Unemployment was 5.7% in November 2008 and rose to 9.3% by May 2009, an increase of 3.6 percentage points. Okun's law predicted 3.4, actual was 3.6—very close.

Example: 2020 pandemic shutdown. Real GDP fell at an 31.4% annualized rate in Q2 2020 (the worst quarter in U.S. history). Potential growth was 2%. The output gap was -33.4%. Using Okun with coefficient 0.5, predicted unemployment change is +16.7 percentage points. But actual unemployment only rose by about 10 percentage points (from 3.5% to over 14%). Okun's law overestimated. Why? The shock was so sudden and explicitly temporary that many firms furloughed workers rather than laying them off permanently, expecting rehiring once lockdowns ended. Unemployment benefits were also unusually generous, allowing workers to survive without frantic job search. The composition of unemployment changed, making the traditional Okun coefficient less accurate.

This shows Okun's law's strength (captures typical recessions) and weakness (breaks down in unusual crises).

Using Okun's law to estimate growth needed to reduce unemployment​

In recovery, Okun's law tells you how much growth is needed to bring unemployment down. For example, if unemployment is 5.5% and the estimated natural rate is 4%, there is 1.5 percentage points of slack. Using Okun with coefficient 0.5, you need 3% of excess growth (1.5 ÷ 0.5 = 3) to close the gap.

If potential growth is 2%, then actual growth needs to be 5% (2% potential + 3% excess) for one year to bring unemployment from 5.5% to 4%. If potential growth is 1.5%, actual growth needs to be 4.5%.

This is why slow-growth recoveries take years to bring unemployment down. After the 2008 recession, when unemployment peaked at 10%, the economy needed to grow at 3% to 4% above potential to get unemployment back to 4% in reasonable time. Growth was often 2% to 3%, so recovery took years. After the 2020 shutdown, unemployment spiked to 14.7% but growth surged to 5% to 6% (above potential), so unemployment fell back to 3.5% within two years.

Potential growth and the output gap​

Okun's law hinges on the concept of potential growth: the long-run growth rate the economy can sustain without accelerating inflation. Potential growth depends on labor force growth, capital investment, and productivity growth. It is typically 2% to 2.5% in the modern U.S.

The output gap is actual growth minus potential growth. A positive gap (actual exceeds potential) means unemployment is below the natural rate and the economy is overheating. A negative gap means the economy is below capacity, unemployment is above the natural rate, and there is slack.

Okun's law says the unemployment rate change is proportional to the output gap. But this requires knowing potential growth, which is unobservable and must be estimated. If economists overestimate potential growth, they will underestimate the output gap and underestimate unemployment changes. If they underestimate potential growth, they will overestimate the output gap.

This adds uncertainty. In real time, economists do not know the true output gap, so Okun's law forecasts have error bands. A 3% recession might produce 1.5 to 2 percentage point unemployment rise, not exactly 1.7, depending on actual potential growth.

Real-world Okun's law examples​

Post-2008 recovery (2009-2015). After the 2008 recession, unemployment peaked at 10% in October 2009 and the natural rate was estimated at 5%. Growth was 2% to 2.5% while potential was 2%, so the output gap was small (0% to 0.5%). Okun predicted unemployment would fall by 0% to 0.25 percentage points per year. In practice, unemployment fell by 0.5 to 1 percentage point per year. Okun underestimated, possibly because potential growth fell due to slower productivity growth (people revised down potential from 2.5% to 2%). The slower potential growth meant actual growth exceeded potential more than it appeared, making the output gap larger.

2011-2019 expansion. From 2011 to 2019, growth averaged 2.3% while potential was estimated at 2%. The output gap was tiny, implying unemployment should have fallen slowly. Yet unemployment fell from 9% to 3.5%, a decline of 5.5 percentage points. Okun's law significantly underestimated. Retrospectively, economists revised down potential growth estimates further, suggesting much of this period was excess growth that the models had not captured in real time.

2020-2021 recovery. Growth exploded after the shutdown, hitting 6.7% in 2021 while potential was 1.5% to 2%. The output gap was roughly 5%. Okun predicted unemployment would fall by 2.5 percentage points. Unemployment fell faster than this, dropping by roughly 3.5 percentage points in a year. The faster decline likely reflects the artificial nature of the unemployment spike (furloughs rather than permanent layoffs, generous benefits) and pent-up demand.

These examples show Okun's law works in typical circumstances but underperforms in unusual scenarios.

Factors that shift the Okun's law coefficient​

The coefficient in Okun's law has varied from 0.2 to 0.7 depending on the time period and country. Several factors shift it:

Labor market rigidity. Rigid labor markets (hard to fire workers, strong unions) produce steeper Okun coefficients. Firms cannot cut staff freely, so small output falls do not reduce employment much, but large contractions do. Flexible labor markets produce flatter coefficients; firms adjust quickly.

Technology and automation. High automation means output can increase without proportional employment increases. This flattens the Okun coefficient (output growth produces less employment growth).

Globalization and outsourcing. Firms can respond to demand changes by outsourcing or onshoring rather than hiring permanently. This complicates the Okun relationship.

Demographic change. Aging populations and declining labor force growth flatten the Okun coefficient. With fewer people in the labor force, firms cannot grow employment much even if output grows.

Cyclical vs. structural shifts. Structural shifts (permanent job losses) produce different Okun patterns than cyclical shifts. The 2008 recession (structural shock) produced a steeper coefficient than typical cyclical slowdowns. (See Federal Reserve research on Okun's law for updated analysis.)

Common mistakes​

Mistake 1: Treating Okun's law as a precise prediction tool. Okun's law is a loose empirical relationship, not a deterministic law. It predicts the direction and rough magnitude of unemployment changes but not exact figures. Using it to forecast unemployment changes to the nearest 0.1 percentage point is false precision.

Mistake 2: Using a fixed coefficient without updating. The Okun coefficient changes over time. Using the 0.5 coefficient from the 1990s in the 2020s will produce inaccurate forecasts. Modern estimates are closer to 0.3 to 0.4.

Mistake 3: Ignoring that potential growth is unobservable. Okun's law requires knowing potential growth, which is estimated, not observed. Errors in potential growth estimates translate to errors in Okun predictions. In real time, economists disagree on potential growth, so Okun forecasts have wide error bands.

Mistake 4: Assuming the relationship holds in all recessions. Okun's law works in typical cyclical recessions but breaks down in supply shocks (oil crises), financial crises, and pandemic shocks. During the 2020 shutdown, unemployment spiked far beyond what Okun predicted because the shock was so unusual.

Mistake 5: Forgetting that the lag varies. Okun's law describes the average relationship between output and unemployment over a quarter or year. But within a quarter, some unemployment is reported before growth data arrives. The timing matters for forecasting.

FAQ​

How accurate is Okun's law?​

Okun's law is about 70-80% accurate in typical recessions, with standard error around 0.5 percentage points. In a typical recession producing a -2% output gap, Okun predicts unemployment will rise by 1 percentage point, but actual could be 0.5 to 1.5 points. In unusual crises, accuracy drops to 50%.

Why is the Okun coefficient sometimes negative or very small?​

In unusual periods (very shallow slowdowns, productivity shocks without demand shifts), the Okun coefficient can be near zero or negative. This happens when output and unemployment decouple—output falls but unemployment doesn't rise, or unemployment rises even as output grows. These periods suggest the underlying relationship has shifted.

Does Okun's law apply internationally?​

Similar relationships exist in other countries, but the coefficient varies. The U.K. coefficient is often around 0.4, lower than the U.S. Eurozone coefficients are varied by country (Germany ~0.3, Italy ~0.5). Differences reflect labor market institutions and economic structures.

Can Okun's law predict inflation?​

Okun's law connects output to unemployment, not directly to inflation. However, through the Phillips curve (unemployment to inflation), you can extend Okun to predict inflation from growth. Low unemployment (predicted by high growth via Okun) correlates with high inflation (via Phillips curve).

What is the relationship between Okun's law and the output gap?​

Okun's law is, at its core, a relationship between the output gap and the unemployment gap. The output gap is actual minus potential growth; the unemployment gap is actual unemployment minus the natural rate. Okun says these move together with coefficient around 0.5.

How often is Okun's law updated or revised?​

Economists continuously debate and revise Okun's law estimates. The Federal Reserve, Congressional Budget Office, and academic researchers publish updated estimates quarterly or annually. However, the broad pattern (1% output gap ≈ 0.5% unemployment change) remains surprisingly robust despite variations in the exact coefficient.

Summary​

Okun's law quantifies the relationship between economic growth and unemployment: for each percentage point the economy grows above its potential growth rate, unemployment falls by roughly 0.5 percentage points (the coefficient has ranged from 0.2 to 0.7 historically, currently around 0.3 to 0.4). The relationship exists because firms do not adjust employment proportionally to output changes; they hoard labor during booms and lay off more than necessary during busts. Okun's law is useful for forecasting unemployment changes from growth data, which arrives faster than employment data, though it breaks down in unusual crises like the 2020 pandemic shutdown. The coefficient has shifted over time due to changes in labor market institutions, technology, and globalization, suggesting the relationship is empirical rather than fundamental.

Next​

→ The original Phillips curve