Guide

Output gap explained

Harbor Manufacturing's 2024 capacity plan assumed 2.1% GDP growth meant a “normal” year. Payroll added 340 temps across two plants while machine utilization sat at 74% and backlogs shrank. When Q3 revenue missed, the ops team discovered the error: they tracked growth, not slack. The economy was expanding, but still 1.8 percentage points below Congressional Budget Office potential — a negative output gap that should have signaled hiring restraint, not overtime expansion.

The output gap is the difference between actual GDP and potential GDP — the level of output an economy can sustain without accelerating inflation. Positive gaps mean overheating; negative gaps mean idle capacity and labor slack. Central banks, fiscal analysts, and corporate planners use gap estimates in Taylor rule benchmarks, Phillips curve inflation models, and business cycle phase calls. After Harbor rebuilt its capacity sleeve around gap-aware inputs, overtime cost spikes fell 28% and temp churn dropped. This guide covers what the output gap is, how potential GDP is estimated, positive vs negative slack, policy and market links, the Harbor Manufacturing refactor, a technique decision table versus growth-rate-only models, pitfalls, and a checklist.

What the output gap is

Formally, the output gap is:

Gap = (Y − Y*) / Y*

Where Y is actual real GDP and Y* is potential real GDP at full employment and normal capacity utilization. The gap is usually expressed as a percent of potential. A +2% gap means the economy is running 2% above sustainable capacity; a −3% gap means 3% of potential output is idle.

The concept sits at the center of macroeconomic stabilization policy. When the gap is positive, demand exceeds supply capacity, wage and price pressures build, and central banks typically tighten. When the gap is negative, spare capacity absorbs shocks without much inflation, and policy can stay accommodative. Unlike raw GDP growth rates, the gap is level-based: you can have positive growth during a negative gap (recovery) or slowing growth during a positive gap (cooling but still overheated).

The output gap is the GDP-side mirror of labor slack measures like unemployment minus NAIRU. The two should move together over long horizons but diverge when productivity shocks, participation changes, or sectoral mismatches break the simple link — see our Okun's law guide for when GDP and jobs decouple.

Measuring potential GDP

Potential GDP is not directly observable. Every gap estimate is only as good as its Y* assumption, and Y* is revised for years after the fact. Main approaches:

CBO and official estimates

The U.S. Congressional Budget Office publishes quarterly potential GDP built from a production-function model: trend labor force participation, NAIRU-based employment, capital stock growth, and total factor productivity. CBO gaps are the standard benchmark for U.S. fiscal scoring and many Fed staff charts. They are smooth, transparent, and slow to capture sudden shocks like pandemic reopening or AI productivity surges.

Statistical filters

Hodrick-Prescott (HP) and band-pass filters extract a trend from actual GDP time series. Filters are fast and model-free but sensitive to endpoint revisions: the last few quarters of “potential” can swing wildly when new data arrive. Real-time gap trackers on trading desks often blend filter output with survey-based capacity utilization.

Survey and market proxies

Capacity utilization from the Fed's industrial production report — how close plants run to physical limits.
PMI/new orders minus inventories — demand pressure relative to supply pipelines.
Unemployment gap — (u − u*) as a labor-side slack proxy when GDP revisions lag.
Inflation surprises — persistent core PCE above target with stable unemployment often implies a positive gap even before GDP data confirm.

No single method wins. Production teams and investors should treat gap readings as a range from multiple estimators, not a precise scalar.

Positive gaps, negative gaps, and inflation

Positive output gap (overheating)

When Y exceeds Y*, firms run overtime, capacity strains, and workers gain bargaining power. Marginal costs rise; firms pass through to prices. The Phillips curve links positive slack to accelerating inflation, though the slope has flattened in recent decades. Equity cyclicals often outperform early in positive-gap expansions; bonds suffer as rate-hike risk rises.

Negative output gap (slack)

Below-potential output means idle machines, underemployed workers, and weak pricing power. Disinflation or deflation risks rise if the gap persists. Automatic stabilizers widen fiscal deficits as tax receipts fall and transfer payments rise. Central banks face room to cut unless inflation expectations de-anchor from a supply shock.

Closing the gap vs crossing potential

Markets frequently confuse gap closing with overheating. GDP can accelerate from −2% gap toward zero (still slack) while headlines scream “strong growth.” Conversely, growth can slow from +1.5% to +0.5% gap while the economy remains above potential and inflationary. Direction of the gap matters as much as the GDP growth print.

Policy links: Taylor rule, fiscal stance, and markets

John Taylor's 1993 reaction function adds 0.5 times the output gap to the recommended policy rate alongside the inflation gap. A +1% gap implies roughly 50 basis points of additional tightening relative to neutral, holding inflation at target. See our Taylor rule guide for coefficient variants and r-star sensitivity.

Fiscal policy also responds to gaps, often unintentionally: progressive taxes and unemployment insurance deepen deficits in negative gaps and shrink them in positive ones. Discretionary stimulus should target measured slack, but political calendars rarely align with CBO revisions.

For asset allocators, gap sign helps position cyclical vs defensive sectors, curve steepness trades, and credit spreads. A deeply negative gap with anchored inflation expectations often favors duration; a positive gap with rising core services inflation favors underweighting long bonds regardless of headline growth deceleration.

Harbor Manufacturing capacity sleeve refactor

Harbor's legacy plan keyed off year-over-year GDP growth and industry PMI alone. The model hired into expansions that were still below potential and cut temps late when growth turned positive but slack remained wide.

The refactor added three gap-aware layers:

CBO gap overlay — quarterly potential GDP from public CBO tables, lagged one release to match corporate planning cycles.
Plant-level utilization band — internal machine-hours versus nameplate capacity, mapped to a −2% to +2% local slack scale.
Wage pressure sentinel — regional employment cost index momentum; flags positive-gap hiring even when national GDP looks soft.

Hiring triggers now require both positive national gap momentum and local utilization above 82%, unless backlog coverage exceeds 14 weeks. Temp pools scale down when the national gap is below −1% regardless of PMI headlines. Overtime authorization needs plant manager sign-off when gap estimates disagree by more than 1 percentage point across CBO and utilization proxies.

Results after four quarters: overtime cost spikes down 28%, temp turnover down 19%, and stockout incidents unchanged — the team stopped over-hiring without sacrificing fill rates.

Technique decision table

Your situation	Prefer	Avoid
Setting macro rate-path or fiscal stance benchmarks	CBO or OECD gap + Taylor rule overlay	Headline GDP growth alone
Real-time trading around data releases	Blended filter + unemployment gap + PMI	Single HP filter endpoint
Corporate capacity and hiring plans	National gap + local utilization surveys	Industry revenue growth without utilization
Inflation nowcast when GDP is stale	Services PMI, wage growth, unemployment gap	Assuming negative gap because growth slowed one quarter
Long-horizon potential revision debates	Production-function decomposition (labor, capital, TFP)	Extrapolating pre-shock trend lines through 2020–2022

Common pitfalls

Confusing growth with slack. Positive GDP growth can coexist with a deep negative gap during early recovery.
Treating CBO gap as real-time truth. Potential GDP revisions can erase half a point of slack years later.
Ignoring supply shocks. Negative gaps during energy or pandemic supply disruptions may still see inflation if expectations de-anchor.
Single-estimator certainty. HP filters and CBO models diverge materially at turning points; use bands.
Labor-GDP decoupling. Productivity booms widen Okun gaps; staffing models need both sides.
Level vs change errors. Markets price gap surprises, not levels everyone already knows from consensus.
International comparability. Euro area gap estimates use different potential methodologies; cross-country rankings shift with definitional choices.
Overfitting corporate plans to one quarter. Harbor's mistake was reacting to a growth print without the level gap context.

Production checklist

Define potential GDP source (CBO, OECD, internal model) and revision policy.
Compute gap as percent of potential, not dollar difference alone.
Track gap level and change separately in dashboards.
Blend at least two estimators; flag when they disagree beyond a tolerance band.
Cross-check GDP gap with unemployment gap and capacity utilization.
Map gap sign to policy stance benchmarks (Taylor rule, fiscal impulse).
For corporate planning, overlay national gap with local utilization data.
Document supply-shock overrides when inflation and gap signs conflict.
Reconcile quarterly with Okun-style GDP–employment consistency checks.
Version potential GDP assumptions; replay major decisions when Y* revises.
Stress-test plans at ±1 pp gap estimation error.
Publish uncertainty bands, not point estimates, to downstream teams.

Key takeaways

The output gap measures how far actual GDP sits from sustainable potential — positive means overheating, negative means slack.
Growth rates and gap levels tell different stories; positive growth can still mean a negative gap during recovery.
Potential GDP is unobservable and revised often; use multiple estimators and treat gaps as ranges.
Central bank reaction functions, Phillips curves, and fiscal stabilizers all lean on gap concepts — Harbor cut overtime spikes 28% by planning on slack, not growth alone.
Cross-check GDP gaps with labor slack and utilization surveys before hiring, rates, or allocation calls.