Research On External Spillover Effect Of China’s Infrastructure Investment

Infrastructure is a basis for a country’s economy and determines the level and degree of economic development. Therefor, it has attracted the widespread attention of researchers all over the world. However, researchers have not obtained consistent conclusion on the issue about spillover effect of infrastructure, which is significant in the involving field. In this paper, we disscuss the spillover effect and the extent of spillover of infrastructure. Additionally, the threshold effect is considered. By analyzing these issues, we also propose the corresponding policy recommendations.This paper defines the infrastructure as a transmission system with spillover effects, which exerts effects on economic development by the network externality. Meanwhile, we select total factor productivity(TFP) method to quantify the external spillover effects of infrastructure and propose the hypotheses to be tested.The 1993-2012 annual provincial panel data is selected to estimate TFP by Solow residual method. We construct the model and derive the influence of every infrastructure variable on the TFP by the dynamic least squares(DOLS) regression. DOLS empirical results show that China’s provincial TFP contribution to the economy was significant in the last two decades.The spillover effects of the three core infrastructure(transport, information, energy) on TFP growth are statistically significant at 1% level. Energy still plays an important role as an traditional facility, and at the same time the effects of transport and information enhance. For other variables, government spending, import and export trade, industry structure play positive roles in promoting the economic, while foreign direct investment and human capital show the negative effects.In the fourth section, we define the threshold regression variable of infrastructure by the panel conversion model(PTR), and analyze the degree of mutation. PTR empirical analysis shows that threshold effect of the per capita information expenditure is significant. When the per capita information expenditure exceeds the threshold value, the external effects of transport, energy and import and export trade enhance, while the results from human and foreign direct investment are consistent with those from DOLS regression, which express negative effects. The threshold effect of industrial structure is not significant and government spending is slightly weakened. At present, the information investment of the most provinces in China is not enough and has not achieve the threshold value of the per capita information investment. As another threshold variable, energy expenditure also has an impact on the externality of infrastructure. Excessive investment will block the external growth of other variables. The investment of energy infrastructure needs to be controlled in a reasonable range, otherwise the elastic contribution from information, transport and industry bodies will be reduced significantly. Additionally, the fluctuation of human capital and foreign direct investment is little.Based on the theoretical analysis and empirical research, in the final part of this artical we propose the policy recommendations for the infrastructure investment. First of all, the overall infrastructure investment should strengthen continually, but part of the input should be slightly slowed down. Second, the information infrastructure investment should be expanded all over the country. The externality of the infrastructure can be utilized when the information infrastructure reaches a certain size. Third, the energy investment should be controlled within a seasonable range.