数字鸿沟对产业结构升级的空间效应研究

Research on the Spatial Effect of Digital Divide on Industrial Structure Upgrade

摘要:

在数字资源禀赋差异的基础上，从数字劳动力创新、数字资本创新、数字人力资本、数字技术创新和数字创新环境等维度，建立相对完整的数字鸿沟评价体系，选取2010至2021年中国31个省份空间面板数据，基于地理距离特征、经济距离特征和数字新基建距离特征的条件下，进行静态和动态空间面板模型分析。结果表明：我国数字鸿沟和产业结构升级具有空间集聚特征和空间联系。地理距离特征和数字新基建距离特征条件下，数字鸿沟加剧了产业空间割据，数字禀赋对产业结构升级具有空间效应，数字鸿沟对域外产业结构升级带来了抑制作用。经济距离因素是数字禀赋促进域内产业结构升级的主要因素，地理距离因素是抑制域外产业结构升级的主要因素。地理距离特征条件下的区域开放程度和数字新基建距离特征条件下的城镇化加剧产业割据。静态空间面板模型产生过高的估计偏误，动态空间面板模型可以修正静态模型。

Abstract:

On the basis of differences in digital resource endowments, a relatively complete digital divide evaluation system is established from the dimensions of digital capital investment, labor investment, technological innovation investment, human capital investment, and innovation environment, Using 31 provinces' spatial panel data from 2010 to 2021 in China. Static and dynamic spatial panel model analysis is performed under the condition of digital new infrastructure distance characteristics. It is found that: the digital divide and the upgrading of industrial structure have the characteristics of spatial agglomeration and spatial connection in Chin. Under the conditions of geographical distance characteristics and digital new infrastructure distance characteristics, the digital divide exacerbates the industrial spatial division, the digital endowment has a spatial effect on the upgrading of the industrial structure, and the digital divide has an inhibitory effect on the upgrading of the industrial structure outside the region. The economic distance factor is the main factor for digital endowment to promote the upgrading of the industrial structure within the region, and the geographical distance factor is the main factor for inhibiting the upgrading of the industrial structure outside the region. The degree of regional openness under the characteristics of geographical distance and the urbanization under the characteristics of distance between digital new infrastructure intensify industrial fragmentation. The static spatial panel model produces excessive estimation bias, and the dynamic spatial panel model can correct the static model.