›› 2018, Vol. 30 ›› Issue (5): 173-178.

Previous Articles     Next Articles

A Study on the Driving Factors of Domestic Export Value-added Changes Based on Data Mining

Xing Guanyuan, Feng Zongxian   

  1. School of Economics and Finance, Xi'an Jiaotong University, Xi'an 710061
  • Received:2017-02-13 Online:2018-05-28 Published:2018-05-29

Abstract:

Using the data from World Input-Output Database (WIOD), based on Koopman et al (2012), Wang et al (2013) with the 7th Framework Program of the European Union (EU-7), this paper calculates the rates of changes in domestic export value-added (Va) and total Export (Vs) towards different countries during the period of 1996 to 2011. Based on the EMD method, we decompose the annu-al growth rates of Vs and Va, and find that the trend RES of Va is the IMF3 component of Vs. We find the growth rate of export value-added increases from 1996, then reaches the peak in 2004 and declines after 2004, which shows a U-shaped feature. Besides, the var-ying tendency of IMF1 and IMF2 which are components of Va are completely consistent with components of Vs. Further, based on the Gravitational Model proposed by Evenett and Keller (2002), we find that the correlation coefficient between IMF1 and dI, which is highly correlated with GDP, is 0.652; the correlation coefficient between IMF2 and Vs trade cost changes is -0.447. This suggests that the driving factors of IMF1 and IMF2 from Va are GDP and trade cost respectively. All the results provide a technologic support and deci-sion-making basis for accurately identifying the driving factors of domestic export value-added and their effects in China.

Key words: domestic export value-added, Empirical Mode Decomposition technology, driving factors, World Input-Output Database