贸易网络与企业出口动态

Trade Network and Firm Export Dynamics

本文探讨国际贸易网络对企业出口动态的影响及其作用机制。首先以2000—2016年跨国双边贸易数据为基础构建国别贸易网络指标体系,衡量一国在全球贸易体系中的相对位置,并由此对接中国企业层面出口数据构建企业出口网络指标,再进一步在引力模型中研究企业贸易网络对出口动态的影响。研究发现:将企业贸易网络指标引入到引力模型后,模型对于企业出口动态的解释力更强。企业当期的一阶、二阶和高阶贸易网络指标提高一个单位,会使得企业—市场下期成功出口的概率分别提升4.86%、4.70%和8.93%,出口额分别增加68.55%、65.25%和128.86%,成功存活的概率分别提升5.97%、5.76%和10.90%。在基准回归中加入扩展地理距离项与控制不同层面固定效应后,结果依然稳健。异质性分析表明,中国企业的出口会根据不同层面的目的地网络信息进行调整,网络信息对企业出口动态的影响存在产品异质性。动态分析表明,相较于地理距离,企业贸易网络指标对企业出口动态有着更强的解释力。本文提出了将网络分析方法与计量经济学相结合的新思路,具有一定的学术价值。

This paper discusses the influence of the international trade network on firm export dynamics and its mechanism.We first use the 2000-2016 cross-border bilateral trade data to build the country-level network indicator to measure the country’s relative position in the global trading system,then connect the Chinese firm export data to construct the firm export network indicators,further study the influence of trade network on export dynamics in the standard gravity equation.The results show that:the model has a more substantial explanatory power for the firm export dynamics when the firm trade network indicator was brought into the standard gravity equation.That the first,second and high order trade network indicators of firm’s current period rose by one unit will increase the firm-market export probability in the next period by 4.86%,4.70%and 8.93%,its export volume by 68.55%,65.25%,and 128.86%,its survival rate by 5.97%,5.76%,and 10.90%accordingly.The results are still robust by adding the extended geographic distance term or controlling the fixed effects of different levels in the benchmark regression.The heterogeneity analysis shows that the export of Chinese firms will adjust to varying levels of destination network information.The influence of network information on the export dynamics of firms has product heterogeneity.Dynamic analysis shows that,compared with geographic distance,firm trade network indicators have a more substantial explanatory power for firm export dynamics.This paper puts forward a new idea of how to combine the network analysis method with empirical economics tests,which has particular academic value.