数字乡村、网络溢出和农业绿色全要素生产率

DIGITAL VILLAGE, NETWORK SPILLOVER AND AGRICULTURAL GREEN TOTAL

关键词:  数字乡村  农业绿色全要素生产率  社会网络分析  网络溢出  空间杜宾模型

Key words:  digital village  agricultural green total factor productivity  social network analysis  network spillover  Spatial Durbin Model

摘要:

目的 探索数字乡村发展对农业绿色全要素生产率的影响及作用机制，为实现农业高质量发展提供新的经验依据。方法 文章基于2014—2020年中国31个省（市、自治区，不含港澳台）面板数据，分别利用信息熵指数法和数据包络分析法对数字乡村发展水平和农业绿色全要素生产率进行测度，运用修正的引力模型和社会网络分析法识别数字乡村发展水平的网络关联特征，在此基础上通过所识别的非对称空间网络权重，构建时空双固定的空间杜宾模型探讨数字乡村发展对农业绿色全要素生产率的网络溢出效应。结果 （1）数字乡村发展水平较高的省份主要分布在长三角、京津冀及珠三角地区，分省数字乡村发展水平在样本期内均呈不同程度的上升趋势，但省份间差异明显。（2）省际之间数字乡村发展水平的联系，已超越地缘意义上的相近，呈现为复杂、多线程的空间网络结构，网络溢出效应较为明显。（3）基于空间关联网络，数字乡村存在显著的空间自相关性，能显著提高当地农业绿色全要素生产率，但对相关联省份具有负向溢出效应。结论 各地应将数字乡村作为推动农业高质量发展的重要抓手，适度有序推进自身的数字乡村发展进程；与此同时，重视数字乡村的协调发展，减少因区域发展不平衡而产生的空间负外部性，根据各省域数字乡村发展的空间关联规律，制定差异化的数字乡村发展政策。

Abstract:

The paper aims to explore the impact and mechanisms of digital village on agricultural green total factor productivity, so as to provide new empirical evidence for achieving high-quality agricultural development. Based on the panel data of 31 provinces (cities or autonomous region） in China from 2014 to 2020, the digital village index and agricultural green total factor productivity were measured by using the information entropy method and data envelopment analysis method, respectively. The network association characteristics of digital village were identified by using the modified gravity model and social network analysis method. According to the identified asymmetric spatial network weights, the Spatial Durbin Model under dual fixed effect was constructed to explore the network spillover effect of digital village on agricultural green total factor productivity. The results were showed as follows. (1) Provinces with higher level of digital village development were mainly distributed in the Yangtze River Delta, Beijing-Tianjin-Hebei Region, and Pearl River Delta. The digital village development level by province showed an upward trend to varying degrees during the sample period. However, there were significant differences among provinces. (2) The linkage of digital village between provinces surpassed the similarity in the geographical sense, and presented a complex, multi-threaded spatial network structure with obvious network spillover effects. (3) Based on the spatial correlation network, digital village had significant spatial autocorrelation, which could indicate the improvement of the province’s agricultural green total factor productivity, but had a negative spillover effect on the associated provinces. Therefore, all provinces should take digital village as a key strategic direction to promote high-quality agricultural development and promote their own digital village development process in an appropriate and orderly manner. Meanwhile, it should pay attention to the coordinated development of digital village, reduce the negative spatial externalities caused by unbalanced regional development, and formulate differentiated policies complying with the spatial correlation principles of digital village development in each province.