Import intensity dynamics in Chinese economic sectors: time series clustering of input-output data (1981–2018)

Abstract

This paper analyzes the evolution of import intensity in Chinese domestic production across 18 economic sectors (1981–2018). Using Dynamic Time Warping (DTW) time-series clustering on annual input-output tables via the dtwclust R package, we identify four distinct temporal patterns of import dependence. As a pilot study using estimated import matrices derived via the proportionality assumption, these findings provide preliminary insights into China’s sectoral import intensity dynamics