Testing the Income Convergence in Korea

The purpose of this paper is to examine the hypothesis of income convergence among 15 Korean regions during 1985~2011, using the recently developed pairwise approach and CIPS panel unit root tests. This study also examines the club convergence hypothesis, and produces the results based on the mean different coefficient. The results based on pairwise tests show that the output differentials for most of the groups are not in favor of convergence based on the cotrending condition as well as cointegration condition. In addition, the mean difference (MD) strengthens a general conclusion of no convergence for all of the sub-groups of regions. Moreover, the results of CIPS test show that regional output gaps do not converge towards the national average of per-capita output, indicating shocks to output gap could be expected to have a permanent effect. Our empirical results imply that past policies by the Korean government have not been able to adjust the structural productivity differences facing with the persisting regional gaps. Thus, a policy aimed at alleviating the impact of the shocks on the output gaps would be absolutely necessary.