Transport Infrastructure, Spatial Clusters and Regional Economic Growth in China

Abstract

China's transport infrastructure distribution and its economic activities have largely the same pattern of spatial clusters. This paper aims to determine whether causal linkages exist between transport infrastructure investment and economic growth in China at national and regional levels. We examine causality in a panel cointegration and a Granger causality framework using time series data throughout the 1978–2008 period. The empirical findings show that in the long run, at the national level, there is unidirectional Granger causality from economic growth to transport infrastructure; at the regional level, there exists bidirectional causality in the affluent eastern region while the low-income central and western regions exhibit unidirectional Granger causality from economic growth to transport infrastructure. These results imply that an improvement in transport infrastructure alone is not sufficient for stimulating economic growth in the underdeveloped areas of China. To better realize the economic benefits brought by transport infrastructure, the Chinese government should pay serious attention to the development of complementary factors in the central and western provinces.

Notes

In the discussion of Chinese economic development, we choose GDP per capita as an indicator of development. We are aware that GDP per capita as a development indicator has serious limitations because it does not count income inequality, is not (always) correlated with well-being and is not strongly correlated with social indicators (including gender equality, access to education and health). However, alternative measures such as the human development index are not available on a longitudinal as well as regional basis. Therefore, we work here with GDP per capita as an imperfect but still meaningful development indicator to explain Chinese economic phenomena in this paper.

The data are collected from the newspaper Economy of the 21st Century (translation from the Chinese).

All the figures in Section 2.3 are calculated by the author based on the data collected from SSB of China and Ministry of Transportation (1984–2009).

Highway: average daily volume is 25 000–100 000 cars. Class roads include four classes according to their average daily volume. Class 1: average daily volume is 15 000–55 000 cars; class 2: average daily volume is 3000–7500 medium-duty trucks; class 3: average daily volume is 1000–4000 medium-duty trucks; class 4: average daily volume is 200–1500 medium-duty trucks; substandard: average daily volume is less than 200 medium-duty trucks.

This information is collected from ‘Energy and Policy in China’, published by the official website of The Central People's Government of the People's Republic of China. Available at: http://www.gov.cn/zwgk/2007-12/26/content_844159.htm.

More details for the econometric methods, including panel unit root test, panel integration test and Granger causality test, can be found in Appendix 1.

In this paper, we choose investment in transport infrastructure as the indicator of transport development, which may neglect the benefits from the transport stock brought by transport investment. However, the data of transport capital investment induced are not available. Thus, we choose the data of transport investment instead.

LLC and IPS represent the panel unit root tests of Levine et al. (2002) and Im et al. (2003) respectively. ADF-Fisher and PP-Fisher represent the Maddala and Wu (1999) ADF-Fisher and PP-Fisher panel unit root tests, respectively. The LLC, IPS, ADF-Fisher and PP-Fisher examine the null hypothesis of non-stationarity.

More details of the results are available upon request from the authors.