Assessing risks to fiscal policy: Calibrating stochastic debt sustainability analysis models

Stochastic debt sustainability analysis (SDSA) is a key tool for assessing risks in modern fiscal policymaking. Yet, designing and calibrating these models poses significant challenges. Using a panel of up to 24 OECD countries with data spanning from 1980, this paper argues that pooling data across countries provides a more representative reflection of fiscal risks, effectively extends sample periods, and a trend-cycle approach better accounts for structural changes. The choice of calibration method is policy relevant as it substantially alters the required primary balance adjustments to stabilise the debt ratio implied by the models. Given that real-time macroeconomic risks are skewed to the downside during economic upswings, the paper proposes a state-contingent (SC-SDSA) approach conditioned on contemporaneous output gap estimates. This suggests that risks are greater during economic booms than conventional approaches and that larger structural primary balances are required than in normal times to ensure adequate fiscal buffers.