The OECD has published an Economics Department working paper assessing how stochastic debt sustainability analysis models should be calibrated for fiscal risk analysis. The paper argues that pooling data across countries and separating structural trends from cyclical shocks can produce more representative estimates of fiscal risk than short, country-specific historical samples, and that the calibration method can materially change the primary balance adjustment implied to stabilise debt. Using panels of up to 24 OECD countries with data extending back to 1980, the paper compares stylised European Commission and International Monetary Fund approaches with a pooled trend-cycle ARMA model. It finds that short country histories can overstate risk for countries hit by recent severe shocks and understate it for countries with unusually calm recent cycles. The paper also develops a state-contingent SDSA approach conditioned on contemporaneous output gap estimates. In that framework, downside growth risks are more pronounced during booms than conventional approaches suggest, and larger fiscal buffers are needed. In the benchmark case, a boom at the 90th percentile of the output gap would require a structural primary balance around 1.5 percentage points of GDP higher than in normal times to stabilise the debt ratio over five years.