Testing for Granger Causality in Heterogeneous Panels with Cross-sectional Dependence: Causal Interrelationships between Export and Growth

This paper proposes a panel Granger causality test for heterogeneous panels with cross-sectional dependence. We define a panel VAR model with unobserved common factors and apply the PANIC procedure to obtain the de-factored data. We then estimate the lag augmented (LA)-VAR models for each cross-section and define the panel statistic based on the meta-analytic approach that combines the p-values of the individual Wald statistics. The Monte Carlo simulations indicate that the test shows good size and power; and appears suitable for the panels where cross-sections may have different unit root or co-integration properties. We finally re-investigate causal interrelationships between export and economic growth in OECD countries by comparing the results from our testing procedure with those from the existing methods. The key finding is that accounting for cross-sectional dependency with factor modelling approach plays a crucial role to determine the direction of causality for country-specific results. A fresh information is that export and economic growth do not cause each other in most of EU countries.