Fiscal policy and external adjustment: New evidence

•We estimate the effects of fiscal-policy shocks on the current account and the real exchange rate.•Shocks are identified through conditional heteroscedasticity.•The results are generally inconsistent with the predictions of standard economic models.•The exchange rate puzzle is worse than you think.

Relatively little empirical evidence exists about countries' external adjustment to changes in fiscal policy and, in particular, to changes in taxes. This paper addresses this question by measuring the effects of tax and government spending shocks on the current account and the real exchange rate in a sample of four industrialized countries. Our analysis is based on a structural vector autoregression in which the interaction of fiscal variables and macroeconomic aggregates is left unrestricted. Identification is instead achieved by exploiting the conditional heteroscedasticity of the structural disturbances. Three main findings emerge: (i) the data provide little support for the twin-deficit hypothesis, (ii) the estimated effects of unexpected tax cuts are generally inconsistent with the predictions of standard economic models, except for the US, and (iii) the puzzling real depreciation triggered by an expansionary public spending shock is substantially larger in magnitude than predicted by traditional identification approaches.