Do We Reject Restrictions Identifying Fiscal Shocks? Identification Based on non-Gaussian Innovations

Identification of Structural VAR Models via Independent Component Analysis: A Performance Evaluation Study

“…Despite the plausibility of the identification strategy, it is important to test such restrictions, at least for two reasons: (i) as Caldara and Kamps (2017) show, the use of plausible range of estimated elasticities may lead to dynamical responses and fiscal shocks that significantly differ in size and persistence; (ii) the authors are not able to distinguish the contemporaneous relationship between government spending and tax revenues and, consequently, whether a tax shock has an immediate effect on spending. Our empirical exercise is similar to the recent study by Karamysheva and Skrobotov (2022), who exploit the non-Gaussianity of the underlying reduced-form residuals to estimate and identify the same VAR as BP, using the GMM estimator derived in Lanne and Luoto (2021). Differently from this study, however, our exercise tests whether the identifying restrictions proposed by BP are rejected through our inference procedure.…”

Identification of Structural VAR Models Via Independent Component Analysis: A Performance Evaluation Study

“…Tax ordered first dynamic responses and fiscal shocks that significantly differ in size and persistence; (ii) the authors are not able to identify the contemporaneous relationship between government spending and tax revenues and, consequently, to check whether a tax shock has an immediate effect on spending. Our empirical exercise is similar to the recent study by Karamysheva and Skrobotov (2022) , who exploit the non-Gaussianity of the underlying reduced-form residuals to estimate and identify the same VAR as BP, using the GMM estimator proposed by Lanne and Luoto (2021) . Differently from this study, however, our exercise tests whether the identifying restrictions proposed by BP are rejected through our inference procedure.…”