Recent decisions
Shorter windows emphasize more recent expectations underlying production and employment plans.
Data & tools
The NGDP Gap
Nominal spending across planning and commitment horizons. Compare actual U.S. nominal GDP with forecasts made over the preceding quarters.
Loading the chart…
TightModerately tightApproximately neutralModerately looseLooseNBER recessions
1971Q1–2026Q2 · Hover, tap, or select a quarter to inspect its value. Black line: zero gap.
Download all windows (CSV) Data snapshot: October 12, 2026 · Actual GDP through 2026Q2 · Last checked: October 13, 2026
Reading the Horizons
Projects underway
Intermediate windows retain expectations that may remain relevant as investments progress toward completion or abandonment.
Longer commitments
Longer windows retain older expectations that may be embedded in financing obligations and other commitments.
These are interpretations to investigate. The data measure spending relative to earlier forecast paths; they do not directly observe the duration of projects or contracts, or identify the causal effect of monetary policy.
Methodology & sources
How the measure is constructed
One method, seventeen windows
For a window of W quarters, the benchmark averages W survey vintages, from the current quarter through t − (W − 1), with equal weights. Every setting uses the same sources, GDP vintage, and forecast-extension rules. Only the window length changes.
NGDP gap (%) = 100 × (actual nominal GDP ÷ average forecast − 1)
A positive gap means spending exceeds the benchmark; a negative gap means it falls short. The benchmark moves forward each quarter. A negative gap in the later 2010s does not preserve a fixed pre-crisis spending path.
Forecasts, extrapolation, and historical coverage
The underlying data are median forecasts from the Philadelphia Fed’s Survey of Professional Forecasters (SPF). Forecast growth paths are rebased onto a common revised GDP vintage. Missing internal forecast levels are interpolated in logs.
Beyond the reported quarterly horizon, forecasts use long-run real GDP growth and CPI inflation, adjusted by the terminal short-run difference between GDP-deflator and CPI inflation. Where those long-run forecasts are unavailable, the last available quarterly growth rate is extended. Longer windows consequently use more extrapolated forecasts.
The SPF reported GNP before 1992; its forecast growth is used as a proxy for GDP growth in this reconstruction. Only complete windows are shown, so the starting date changes with the slider. Recession shading uses the NBER-based USREC indicator.
Classification thresholds
The fixed categories below follow William J. Luther’s reported thresholds. They are descriptive bands, not confidence intervals.
| Category | NGDP gap |
|---|---|
| Tight | Below −1.25% |
| Moderately tight | −1.25% to below −0.75% |
| Approximately neutral | −0.75% through +0.75% |
| Moderately loose | Above +0.75% through +1.25% |
| Loose | Above +1.25% |
Sources and relationship to published measures
The approach draws on Beckworth (2020), Luther (2024), and Martinez et al. (2026). This explorer is based on Martinez, Schibuola, and Beckworth (2026), extended to compare alternative planning and commitment horizons using forecast windows of four to twenty quarters. It is not an exact numerical replication of the published series; data vintages and implementation can affect the values.
The Luther link opens his technical description, which cites the 2024 working paper, “Neutral Nominal Spending and the Nominal Spending Gap.”
- Philadelphia Fed: SPF median forecasts
- BEA nominal GDP via FRED
- NBER recession dates via FRED
- Calculation script and reproduction instructions
The automatic refresh checks public sources each weekday for new releases and revisions. All windows and the CSV download update together after validation. Historical gaps can change when source data are revised; if a refresh fails, the last successful version remains available.