The nexus of transport infrastructure and economic output in city-level China: a heterogeneous panel causality analysis

The nexus between transport infrastructure and economic output has been subject to heated debates. Existing studies investigating their nexus between transportation and economic growth tend to focus on macroscales and utilize panel data but assume homogeneity across individual units. This paper aims to handle the problem of geographic heterogeneity and panel heterogeneity embedded in Granger causality between transportation and economic growth in China. According to a panel data of 216 cities from 1999 to 2014, heterogeneous Granger non-causality model at the nation level and region-specific heterogeneous Granger non-causality models are established. Results show that the heterogeneity of their mutual causality across cities cannot be neglected. In the direction from transport infrastructure to economic output, this paper finds homogeneous causality in all geographic regions except central China and it proves that the pulling effect of transportation on economic growth is positive in China but relatively marginal in eastern and western regions. In the reverse direction, heterogeneous causality can be found in all geographic regions. However, western region finds negative impact of economic growth on transportation. Overall, bidirectional causality is observed in 53.24% of the cities (115 out of 216) in the panel. Unidirectional causality is observed in about 39.35% (85 out of 216) and no causality in 7.41% (16 out of 216). This paper concludes with some main findings and region-specific policy implications.