中国经济已经进入到了发展的新阶段，创新驱动和提质升级成为了发展的主旋律。作为我国经济发展的重要支柱，制造业是我国经济转型升级的重中之重，制造业创新驱动发展是制造业高质量发展的核心路径。

当下，我国数字经济发展日新月异，数字经济将对制造业的创新升级发挥重大作用。经过多年的建设，我国的数字经济发展已经有了良好的基础。近年来我国高度重视数字经济和制造业的融合发展，先后发布了《中国制造2025》、《关于深化“互联网+先进制造业”发展工业互联网的指导意见》等重要规划。5G、大数据、云计算、人工智能、区块链等技术，将促进制造业和数字经济深度融合，驱动我国制造业的创新升级。

数字经济是当下迅速发展的新经济形态，制造业创新是我国制造业转型升级的核心路径，数字经济和制造业创新既是我国经济创新发展的关键因素，也有着融合发展的坚实基础，研究数字经济对制造业创新的促进作用可以进一步加强对数字经济和创新的认识，也对推动数字经济和制造业创新发展具有重要意义。本文分几个步骤对数字经济促进制造业创新发展这个主题进行了研究。首先，本文通过回顾已有文献，概括了数字经济和制造业创新能力的概念。并整合了过去的研究相关成果，提出了可能贡献的研究点。其次，本文通过机理分析将数字经济促进制造业创新能力概括为五种机制：数据计算机制、主体连接机制、市场扩大机制、提升金融配置效率、重塑劳动力市场机制。第三，通过相对差异指数分析了制造业创新能力的省级差异，通过NRI方法构建了2010年-2019年各省的数字经济指数，并通过耦合协调分析初步分析了数字经济和制造业创新能力的关系。第四，进行实证研究，构建了面板模型，考察了数字经济对制造业创新总量、人均制造业创新量、新产品销售收入的关系，并进行了分地区和分城市群的区域异质性分析。同时使用解释变量滞后变量、替换被解释变量和空间面板模型进行稳健性分析。通过门槛回归考察了数字经济和制造业创新的非线性关系。第五，创新效率能够体现出创新投入和产出的转化能力，本文通过三阶段DEA模型测算了2010-2019年各省的制造业创新效率，并通过Tobit模型研究了数字经济对制造业创新效率的影响，并分地区进行了异质性分析，通过滞后解释变量回归进行了稳健性检验。

通过前文研究，有七点发现。一是发现我国的数字经济发展整体呈现不断发展的状况，大部分省份的数字经济指数都有了不断提升。数字经济发展程度比较高的大部分集中在东部地区和省份，安徽等中部省份提升幅度比较大。二是我国的制造业创新能力在区域间存在一定差异，东部制造业创新能力比较强，西部省份制造业创新能力相对较弱，安徽等中部省份制造业创新能力发展较快已进入全国前列，制造业创新能力整体在提升，但是东北省份制造业创新能力有减弱趋势，省级制造业创新能力差异在2015年后开始减少。三是我国的数字经济和制造业创新能力耦合协调程度整体并不高，同时体现出了较大的地区差异，有着较大的提升空间。我国的制造业创新能力和数字经济的耦合协调度在不断提升，一部分发达省份已经进入高度耦合协调的区间。四是发现数字经济显著促进了制造业创新总产出，数字经济能够促进制造业创新人均产出，通过空间面板数据模型、替换被解释变量为研发销售占比、采用核心解释变量滞后项均结论稳健，数字经济随着时间推移促进制造业创新总量作用能够更加突出。在区域异质性方面，东部和西部的系数显著为正，说明数字经济显著提升了东部和西部制造业的创新总产出，其中在东部的系数更为显著。通过空间自相关分析发现数字经济和制造业创新具有一定的空间集聚的效应，高值集聚地区主要集中在东南部，低值集聚地区主要集中在西部。通过门槛回归发现人均GDP能够促进数字经济对制造业创新能力的影响。五是数字经济能够促进制造业新产品销售销售收入，代表数字经济能够促进制造业创新市场转换能力，通过解释变量滞后项分析结论稳健。六是发现我国的制造业创新效率整体不高，在创新效率的空间分布上，我国的制造业创新效率高的地区主要集中在东部地区，创新效率相对低的地区主要集中在西南和西北地区，安徽等中部省份在创新效率的提升上有良好表现。我国的制造业创新效率有了一定提高，通过环境因素对松弛变量的分析，可以发现产业结构和科研机构数量代表的科研氛围能够促进科研人员对于制造业创新能力的作用。七是发现数字经济能够促进制造业效率的提升，东部的效应更为显著，而在中部和西部的效应不显著。通过构建核心解释变量滞后项回归效果仍然显著。

基于以上分析，本文从六个方面提出了政策建议。一是加强政策支持，多方协力完善政策环境，通过产业政策、金融政策、财政政策、教育政策多方面的协同联动加强数字经济和制造业的融合发展；二是提高人力资本，政府通过加强培训和大学教育、职业技术教育、金融手段支持工业企业的数字化和智能化改造，同时通过加强宣传和引导推广、鼓励人力资源加入制造业企业；三是基于数字经济对制造业创新存在区域异质性作用，因此在不同区域要采用符合实际情况的方法，通过发达地区的带头作用形成一定的区域带动作用，促进不同区域间的交流；四是加强数字经济的基础设施建设，发挥数字经济对制造业创新的机制作用。五是鼓励数字经济企业和制造业企业的相互合作，开展协同创新；六是我国的制造业创新效率有一定提升空间，需要支持企业探索提升数字经济创新促进作用的管理机制，打破企业内部的数据藩篱，形成全流程的数字化。

本文可能的创新点是，一是数字经济高速发展以来，成为了学术界关注的重点领域，本文尝试把数字经济和制造业创新联系起来，分析了两者之间的关系，尝试做出一些边际贡献。二是通过空间自相关分析、变异系数、耦合协调、三阶段DEA、面板数据模型、Tobit模型等分析数字经济和制造业创新的发展情况和地区间的异质性，研究数字经济和制造业创新能力的关系，全面考察了数字经济和制造业创新能力的省际差异和影响关系；三是对制造业创新能力通过创新总产出、人均创新产出、新产品销售收入、创新效率等进行全面的衡量，考察了数字经济对制造业创新产出、创新市场转换能力、创新投入产出比等多个角度的制造业创新能力的影响，从而对制造业创新能力和数字经济的关系有了全面的考察，这一点在以往研究中较少出现；四是为了提高稳健性和解决内生性问题，通过纳入空间效应的面板数据模型、替换被解释变量为研发销售占比、解释变量的滞后项回归，从而从多个角度考察结论的稳健性和可靠性。

China's economic development has entered a new stage of development, and innovation driven, quality improvement and upgrading have become the main theme of development. As an important pillar of China's economic development, manufacturing industry is the top priority of China's economic transformation and upgrading. The development of digital economy is changing with each passing day, which has become highly valued by the party and the state and incorporated into the top-level design. Digital economy will play an important role in the innovation and upgrading of manufacturing industry. After years of construction, China's digital economy has a good foundation. In recent years, China issued the plan of "made in China 2025" in 2015, and issued the "guiding opinions on deepening the development of the Internet plus advanced manufacturing industry" in 2017. 5g, big data, cloud computing, artificial intelligence, blockchain and other technologies will promote the deep integration of manufacturing industry and digital economy, generate new economic forms and promote the in-depth development of China's manufacturing industry.

Digital economy is a rapidly developing new economic form, and manufacturing innovation is the core path of China's manufacturing transformation and upgrading. By returning to the existing literature, this paper summarizes the concepts of digital economy and manufacturing innovation ability. It also integrates the relevant results of past research, and puts forward the research points that may contribute. Then, through mechanism analysis, this paper summarizes the promotion of manufacturing innovation by digital economy into five mechanisms: data computer system, subject connection mechanism, market expansion mechanism, improving the efficiency of financial allocation, reshaping the labor market and so on. Next, the provincial differences of manufacturing innovation are analyzed through the relative difference index, the digital economy index of each province from 2010 to 2019 is constructed through the NRI method, and the relationship between digital economy and manufacturing innovation is preliminarily analyzed through the coupling and coordination analysis. Then, the empirical research is carried out, using the panel model, taking the digital economy index as the core explanatory variable, and using the total amount of manufacturing innovation, per capita manufacturing innovation and new product sales revenue as the explanatory variables, the regression model is

constructed to investigate the relationship between digital economy and manufacturing innovation, and the regional heterogeneity analysis is carried out by regions and urban agglomerations. Using lag variables, replacement explanatory variables and spatial panel model for robustness analysis, and regional heterogeneity analysis by region and urban agglomeration. The possible nonlinear effects of digital economy and manufacturing innovation are investigated by threshold regression. Innovation efficiency can reflect the transformation ability of innovation input and output. This paper calculates the manufacturing innovation efficiency of each province from 2010 to 2019 through the Three-stage DEA model, studies the impact of digital economy on manufacturing innovation efficiency through Tobit model, makes heterogeneity analysis by region, and tests the robustness through lag variable regression.

Through the previous research, there are several findings. First, through the construction of the digital economy index, it is found that the overall development of China's digital economy presents a continuous development situation, and the digital economy index of most provinces has been continuously improved. Most of the high degree of development of digital economy are concentrated in the eastern region and provinces, a small number of central provinces have a high degree of development, and some central provinces have increased significantly. Second, there are some regional differences in China's manufacturing innovation capability. The eastern manufacturing innovation capability is relatively strong, while the western provinces have relatively weak manufacturing innovation capability. The manufacturing innovation capability of some central provinces has developed rapidly and has entered the forefront of the country. The manufacturing innovation capability is improving as a whole, and some central provinces have developed rapidly, However, the innovation ability of manufacturing industry in some northeastern provinces tends to weaken. The difference of provincial manufacturing innovation capacity began to decrease after 2015. Third, the coupling and coordination between China's manufacturing industry and innovation is not high as a whole. At the same time, it reflects large regional differences and has great room for improvement. The degree of coupling and coordination between China's manufacturing innovation ability and digital economy is constantly improving, and some developed provinces have entered a highly coupled and coordinated range. Fourth, by constructing the panel data model, it is found that the digital economy significantly promotes the total amount of manufacturing innovation. Through the spatial panel data model, replacing the explained variable as the proportion of R & D and sales, and adopting the lag term of the core explanatory variable, the conclusion is stable. The role of the digital economy in promoting the total amount of manufacturing innovation over time can be more prominent. In terms of regional heterogeneity, the coefficients in the East and West are significantly positive, indicating that the digital economy has significantly increased the total innovation of manufacturing industry, especially in the East. From the regression results of urban agglomerations, the coefficients of Beijing Tianjin Hebei, Yangtze River Delta, Central Plains and Chengdu Chongqing urban agglomerations are positive and significant, reflecting that the digital economy promotes the total amount

of manufacturing innovation in these urban agglomerations. Through spatial autocorrelation analysis, it is found that digital economy and manufacturing innovation have a certain effect of spatial agglomeration. High value agglomeration areas are mainly concentrated in the southeast and low value agglomeration areas are concentrated in the West. Through threshold regression, it is found that per capita GDP can promote the impact of digital economy on manufacturing innovation. The role of digital economy reflects nonlinear characteristics, which also confirms the scale effect and network effect of digital economy. Sixth, through the Three-stage DEA innovation efficiency analysis, this paper analyzes the innovation efficiency of China's manufacturing industry. Through the analysis of manufacturing innovation efficiency, it is found that China's manufacturing innovation efficiency is not high as a whole. In the spatial distribution of innovation efficiency, the areas with high innovation efficiency are mainly concentrated in the East, and the areas with relatively low innovation efficiency are mainly concentrated in the southwest and northwest. However, some central provinces have good performance in improving innovation efficiency. The innovation efficiency of China's manufacturing industry has been improved to a certain extent. Through the analysis of relaxation variables by environmental factors, it can be found that the scientific research atmosphere represented by the industrial structure and the number of scientific research institutions can promote the role of scientific researchers in the innovation ability of manufacturing industry. Seventh, by constructing Tobit model to analyze the impact of digital economy on manufacturing innovation efficiency, it is found that digital economy can promote the improvement of manufacturing efficiency, the effect in the East is more significant, but the effect in the middle and West is not significant. By constructing the core explanatory variable lag term, the regression effect is still significant.

Based on the above analysis, this paper puts forward policy suggestions from six aspects. First, strengthen the top-level strategic design. Make concerted efforts to improve the policy environment through the coordinated linkage of industrial policy, financial policy, fiscal policy and education policy; Second, by strengthening training and university education, vocational and technical education, supporting the digital and intelligent transformation of industrial enterprises through financial means, strengthening publicity, guidance and promotion, encouraging human resources to join manufacturing enterprises, vigorously supporting the R & D and promotion of 5g and 6G networks, and playing the role of big data exchange, Activate industrial big data trading market and other segmentation policies to promote the development of digital economy and manufacturing industry; Third, in different regions, we should adopt methods in line with the actual situation, form a certain regional driving role through the leading role of developed regions, and promote exchanges between different regions; Fourth, strengthen the integration of manufacturing and digital economy. Fifth, according to the actual situation, vigorously promote the industrial Internet and intelligent manufacturing projects of manufacturing enterprises, and actively promote industrial intelligence and industrial big data; Sixth, support enterprises to explore and improve the management mechanism of promoting digital

economy innovation, break the data barriers within enterprises and form the digitization of the whole process.

The possible innovations of this paper are as follows: first, through text analysis, the keywords related to digital economy in the government report are extracted as the indicators to measure the development environment, and the digital economy index system is constructed by NRI method according to the four categories of indicators: digital economy development, digital economy infrastructure, digital economy application and digital economy development environment. The second is to analyze the development and regional heterogeneity of digital economy and manufacturing innovation through spatial autocorrelation analysis, coefficient of variation, coupling coordination, Three-stage DEA, panel data model and Tobit model, and study the relationship between digital economy and manufacturing innovation. The third is to comprehensively measure the innovation ability of manufacturing industry through the total innovation output, per capita innovation output, new product sales revenue and innovation efficiency, and investigate the innovation ability of manufacturing industry from multiple perspectives, such as innovation output, innovation market transformation ability and innovation input-output ratio. Fourth, in order to improve the robustness and solve the endogenous problems, the robustness and reliability of the conclusion are investigated from multiple angles by incorporating the panel data model of spatial effect, replacing the explained variables as the proportion of R & D sales and the lag regression of explanatory variables.