Abstract
Using recent developments in econometric techniques, this article re-examines the export-led growth (ELG) hypothesis for Korea over 1963–2001. The Granger-causality tests was based on two testing approaches: vector error correction modelling (VECM) approach outlined in Toda and Philips; and the augmented levels VAR modelling with integrated and cointegrated processes (of arbitrary orders) separately introduced by Toda and Yamamoto (Citation1995) and Dolado and Lutkepohl (Citation1996). Empirical evidence from causality tests based on the two alternative approaches indicates that the causal link between real exports and real GDP growth is bi-directional. Additional determinants of growth are also found to be significant.