关键词: 智能化, 全要素生产率, 企业创新, 倾向得分匹配, 双重差分法

Abstract:   The report of the 19th National Congress of the Communist Party of China first proposed that China′s economy has shifted from a stage of rapid growth to high-quality development. High-quality development should be based on the improvement of production factors, productivity and total factor efficiency. With the emergence of ‘industry 4.0’, the world is facing a huge opportunity for networking, digitization and intelligence, which is also put forward by ‘China manufacturing 2025’. The development of intelligent relevant technologies and industries is gradually becoming the core factor of the global economy. In 2018, the scale of China′s robot market was US8.74 billion, with a year-on-year increase of 39.2%, accounting for 30% of the total amount of global robots. And the global output of industrial robots was 390 thousand, with an increase of 14.6%. The output of industrial robots in China was 148 thousand, accounting for 38%. As the world′s second largest economy, intelligence has played an important role in China. Therefore, it is urgent to promote the development of enterprises and the transformation of industrial structure through intelligence.Based on the relevant intelligence policies of China, using the panel data of listed companies from 2009 to 2018, this paper studies the effect of intelligence on total factor productivity and innovation of enterprise, and analyzes the heterogeneity of factor-intensity, ownership structure and R&D-density through the method of propensity score matching and difference-in-difference. The results are as follows. First, intelligence can improve the total factor productivity of enterprises in the current period. However, the impact of intelligence on innovation is insignificant. Second, intelligence has a lag and dynamic impact on the total factor productivity and innovation. The relationship between intelligence and total factor productivity shows a "U" trend. The effect of intelligence on the innovation functions from the second year, and appears an inverted "U" trend. Third, the effect of intelligence on total factor productivity in labor- and resource- intensity enterprises is greater than that of capital-intensity enterprises, while the effect of innovation is the opposite. Fourth, state-owned enterprises have a bigger positive impact of intelligence on total factor productivity than that of non-state-owned enterprises. In addition, the effect of innovation is not significant. Fifth, enterprises with lower R&D-density have stronger and more positive effect of intelligence on TFP than that of higher R&D-density enterprises. However, intelligence has an insignificant effect on the innovation of all types of enterprises in the current period. The possible innovations and contributions of this paper are as follows: First, we find the intelligent pilot policies, proposed by the Ministry of Industry and Information Technology, the National Development and Reform Commission and the Ministry of Science and Technology. This paper also uses the control group through the propensity score matching method and analyzes the impact of intelligence on enterprise performance through the difference in difference method. Third, the lag term of dummy variables is also introduced into the model to study the lag effect of intelligence on enterprise performance. Finally, this paper studies the different effects of intelligence on enterprise performance considering the factor intensity, ownership structure and R&D density.

Key words: intelligent technology, total factor productivity, enterprise innovation, propensity score matching, difference in difference (DID) method