Can we rely on real time figures for cyclically adjusted budget balances?

This paper analyses the reliability of real time estimates of cyclically adjust budget balances (CABs). We find that real time CABs are not better at forecasting the ex post figures than simpler benchmarks. Further, we find that real time CABs have low power in detecting fiscal slippages, and in correctly identifying fiscal improvements. Around half of the real time errors in CABs can be attributed to revisions in the cyclical component of the budget balance, and around one half to revisions in the deficit to GDP ratio across vintages. That means it will be difficult to use them to reliably monitor the health of public finances. Lastly, we find that CABs are systematically less reliable under conditions of poor or deteriorating public finances, which means they are at their most unreliable precisely when they are needed most