Analysis of Homogeneous and Heterogeneous Causality between Economic Growth and Export in Panel Data using Dumitrescu-Hurlin Method

Export and economic growth are of those economic variables which have parallel behaviors with respect to one another. However there are different views regarding the causality between them. This paper attempts to test the causality between export and economic growth by referring to Dumitrescu and Hurlin (2012) test and data for 91 countries between 1980i-2012. To this end, Granger causality test has been applied in Homogenous Non Causality and Heterogeneous Non Causality methods with panel data. The result does not reject the causal relationship directed from export to economic growth in either methods. Although there is homogeneous causality directing from economic growth to export, but such causality is rejected in Dumitrescu and Hurlin (2012) test. In other words, Heterogeneous Non Causality method rejects the causality relation directing from economic growth to export, therefore such causal direction cannot be generalized to all countries. Accordingly, by separating the subsets of countries in accordance with income category into five groups, the impact of export on economic growth has been estimated by the system of generalized method of moments (GMM-SYS) estimator; this estimation shows that there is a significant positive effect in LI, LMI, UMI, HI and OECD countries and this effect increases in parallel with the increase in income group. Furthermore, the extent to which this effect bears is significantly different across the income groups.