Fiscal Multiplier Bench

Problem solved

Finance ministries, sovereign credit teams, and macro funds need defensible multiplier estimates by spending category for the country and the cycle phase that matter to the decision in front of them. Generic 1.0 or 0.5 multipliers do not survive an investment committee. The framework produces a country-specific bench with explicit cycle-phase and monetary-regime conditioning.

Inputs

• Quarterly real GDP, government consumption, public investment, transfers, and tax revenue (IMF GFS, country fiscal accounts)
• Output gap series (IMF WEO, OECD Economic Outlook, country central bank)
• Policy rate, inflation, and exchange rate regime classification
• Public debt to GDP and primary balance trajectory
• Discretionary fiscal-shock identification series (Romer-Romer, Blanchard-Perotti, narrative)
• Cross-country panel for prior-tightening (Auerbach-Gorodnichenko regime-switching estimates)

Outputs

• Multiplier point estimate by spending category: government consumption, public investment, transfers, tax cuts, with 90 percent CIs
• Cycle-phase conditioning: recession multiplier versus expansion multiplier
• Monetary regime conditioning: at the effective lower bound versus normal policy rate
• Cumulative multiplier at 1, 4, 8, 12 quarter horizons
• Crowding-out estimate via interest-rate and exchange-rate channels
• Comparison row against a five-country peer panel for the same category and phase

Method

1. Step 1. Build the country fiscal panel from 1990 or earliest available, quarterly. Reconcile IMF GFS to country accounts; resolve the cash-versus-accrual distinction.
2. Step 2. Identify discretionary shocks. Use Blanchard-Perotti (timing restrictions on tax and spending) as the base specification, with Romer-Romer narrative shocks as a robustness case where a country-specific narrative archive exists.
3. Step 3. Estimate a baseline local-projection model (Jorda) by spending category. Report cumulative multipliers at 1, 4, 8, and 12 quarter horizons with Newey-West standard errors.
4. Step 4. Layer in regime-switching (Auerbach-Gorodnichenko style). Use the output gap as the state variable. Report recession-state and expansion-state multipliers separately.
5. Step 5. Add the monetary regime layer. Stratify by whether the policy rate was at the effective lower bound during the shock window. Floating versus pegged exchange rate regime is a secondary stratification.
6. Step 6. Compute crowding-out. Estimate the response of the policy rate, the long rate, and the real effective exchange rate to the same shocks. Report channel decomposition.
7. Step 7. Benchmark. Pull the same multiplier from the cross-country prior panel for five peer countries. Report the country point estimate against the peer median and IQR.

Assumptions

• The output gap is observable with reasonable lag. The framework uses the IMF WEO output gap as the base, with the country central bank's series as a robustness case.
• Discretionary shocks are exogenous to current-quarter output conditional on the timing restrictions. This is the Blanchard-Perotti identifying assumption, and the framework reports the narrative-shock alternative as a robustness check.
• Multipliers are local approximations. Large fiscal expansions outside the historical sample are extrapolated with named caution.
• The peer panel is a rough discipline check, not a structural benchmark. Country-specific institutions matter more than the panel average.

Limitations

• Quarterly fiscal data quality is highly variable across emerging markets. Pakistan, Egypt, Argentina samples carry quarter-to-quarter revisions that widen confidence intervals.
• The narrative-shock identification is feasible only for countries with a documented archive of fiscal-policy decisions (US, UK, Japan, a handful of others). For most emerging markets, the timing-restriction identification carries the weight.
• Multipliers shift over time. The framework reports a rolling 15-year window in addition to the full sample to surface structural shifts.
• The framework does not produce a debt-sustainability multiplier directly. That is a separate calculation that uses these multipliers as an input.

Example application

Applied to Brazil's 2026 fiscal trajectory: the new fiscal framework, the markets' skepticism, and the political constraint on primary balance adjustment. The Multiplier Bench runs the seven steps on Brazilian data 1996 to 2025, conditioning on the cycle phase Brazil enters 2026 in (small positive output gap) and the monetary regime (high real rates, no ELB issue). Public investment carries a four-quarter multiplier of about 1.1 in the baseline; transfers come in around 0.6. The brief uses these as the anchor for its scenario tree on primary balance paths. See Brazil fiscal trajectory 2026.