The Output Gap and Potential Growth
The output gap measures how far actual GDP sits from what the economy could sustainably produce without overheating — a number nobody observes directly, but that quietly drives central bank policy decisions.
Prerequisites: GDP and the National Accounts, The Business Cycle
A economy growing at 3% a year sounds healthy. Whether it actually is healthy depends on a number you can't look up on any data release: how fast the economy could grow without straining its own capacity. If potential growth is only 2%, that 3% is running hot, building up inflationary pressure. If potential growth is 4%, that same 3% is actually underperforming, leaving workers and factories idle. The gap between what's happening and what's sustainable is the output gap, and central banks lean on it heavily even though it can never be measured directly — only estimated.
The output gap is actual GDP minus potential GDP, expressed as a percentage of potential GDP. A positive gap means the economy is running above its sustainable capacity, historically a sign of building inflationary pressure; a negative gap means there's slack — unused capacity — and typically disinflationary or recessionary pressure. Because potential GDP is never directly observed, the gap is always an estimate, and can be revised significantly after the fact.
What "potential" actually means
Potential GDP is not an economy's maximum physically possible output — it's the level of output consistent with stable inflation, given the economy's available labor, capital, and productivity, all fully but not over-utilized. It grows over time as the labor force expands, capital is invested, and productivity improves, but that growth rate itself shifts slowly and is estimated using models that combine data on labor-force trends, capital stock, and historical productivity growth — there's no single official gauge, and different institutions (a central bank, the CBO, the IMF) can each produce a different estimate for the same economy in the same quarter.
In words: take how much bigger (or smaller) the economy actually is than its estimated sustainable size, and express that difference as a percentage of the sustainable size.
Worked example
An economy's actual GDP for the year is $21.5 trillion, and the central bank's economists estimate potential GDP at $21.0 trillion. Output gap = ($21.5tn − $21.0tn) / $21.0tn ≈ +2.4%. A positive gap of this size signals the economy is running above sustainable capacity — consistent with tight labor markets and upward pressure on wages and prices — and historically supports an argument for tighter monetary policy. Two years later, after a slowdown, actual GDP is $21.3 trillion while potential GDP, still growing, is now estimated at $21.6 trillion: output gap = ($21.3tn − $21.6tn) / $21.6tn ≈ −1.4%. The negative gap signals slack in the economy — underused capacity, typically consistent with easing inflationary pressure — and would tend to support an argument for looser policy.
What this means in practice
Central banks use estimated output gaps as an input to interest-rate decisions precisely because the gap tracks where inflationary pressure is likely heading before it fully shows up in the inflation data itself. But because potential GDP is unobservable and estimated with a lag, output-gap estimates are revised — sometimes substantially — as more data comes in, and policymakers know in real time they are working from a noisy, uncertain signal, not a hard number.
The output gap is not a fact reported alongside GDP — it's a model-dependent estimate built on an unobservable "potential" that different institutions calculate differently and that gets revised, sometimes sharply, well after the fact. A policy decision defended by "the output gap was positive" can look wrong in hindsight simply because the potential-GDP estimate underlying it was later revised, not because the decision itself was unreasonable given what was known at the time.
Further reading
- CBO, 'How CBO Estimates Potential GDP'
- Blanchard, Macroeconomics (ch. 9, 'The Phillips Curve')