Page 119 - Contributed Paper Session (CPS) - Volume 1

CPS1196 Song X. et al.

Clustering Chinese cities' economic growth paths
with dynamic time warping
Song Xue, Hong Liu
Zhongnan University of Economics and Law, Wuhan

Abstract
In this research, we use the time-series cluster analysis to study Chinese cities'
economic growth patterns. In particular, dynamic time warping algorithm is
used to measure the time-series distance between two growth paths. 35
Chinese cities that are economically most important are included in the
research. The cluster analysis categorizes the 35 cities into five groups, each of
which exhibits distinct economic growth patterns. The five groups are: a.
service centers, b. deindustrializing cities, c. balanced-industrializing cities, d.
traditional industrial centers, and e. emerging industrial centers. This research
shows the potential of applying unsupervised machine learning techniques in
development economics, and can be extended in many ways.

Keywords
Dynamic time warping; time series clustering; urban growth; economic
development

1. Introduction
China's economic growth in recent decades has been accompanied by
unprecedented scale of urbanization (Démurger, 2001). In the last three
decades, over half a billion Chinese rural residents have moved to cities. The
urban population in China has increased from 19.4% in 1980 to 57.9% in 2017.
Given a base population growth from 1 billion to 1.38 billion during the same
period, this urbanization surge means over half a billion rural residents have
moved to cities in China, more than the population of United States and Japan
combined. After year 2006, 60% of China's urbanization occurs in large cities
that have more than 4 million residents.
While China's economic boom and the accompanying urbanization have
been the themes of voluminous research (Chan and Wan, 2017), the
heterogeneity within the economic development paths of Chinese cities is
often overlooked in extant literature. Despite their similarities in economic
achievements, cities in China are diverse in natural, social, and political factors
that would influence their growth paths (Wu, 2016). Some of the cities are
coastal harbors that have been the center of trade and commerce for
centuries, while some others are hundreds of miles inland, and have become
industrialized at the result of central planning (Alder 2016). In terms of the

108 | I S I W S C 2 0 1 9