Asymmetric Panel Causality Tests with an Application to the Impact of Fiscal Policy on Economic Performance in Scandinavia

Tests for conducting asymmetric Granger causality within a panel system are introduced in this paper. It is shown how the cumulative sums of negative and positive shocks can be constructed to investigate whether the potential causal effects of these shocks are asymmetrical or not within a panel system. These tests can be based on asymptotic or bootstrap distributions. The tests are applied to assess the impact of contractionary as well as expansionary fiscal policy on the economic performance of the three Scandinavian countries. The results show that allowing for asymmetry in the panel causality testing has important repercussions for the underlying causal inference.