创新低碳之城:创新型城市建设的碳排放绩效评估

创新决定城市未来，低碳引领未来城市。中国政府自2008年开始实施创新型城市试点政策，并将绿色低碳作为创新型城市建设的原则和目标。因此，准确评估创新型城市建设对碳排放绩效的影响效应，对于进一步推广这一试点政策具有重要意义。文章利用中国城市的面板数据，借助于创新型城市试点政策在不同城市、不同试点时间上的变异，使用双重差分法估计了创新型城市建设对碳排放绩效的影响。研究发现，相比于非试点城市，创新试点城市的碳排放绩效平均增加2.47%，意味着创新型城市建设显著提升碳排放绩效。经过共同趋势、工具变量、干扰政策、安慰剂等一系列稳健性检验后，上述结论依然成立。机制分析表明，创新型城市建设能够强化创新政策支持、提高创新要素集聚、增加创新投入和改善创新环境，从而促进技术创新、优化产业结构和转变发展方式，这有利于降低碳排放水平，进一步提升碳排放绩效。从动态效应上看，创新型城市建设对碳排放绩效的提升效应具有持续性，并且随时间不断增强。文章是国内首篇从绿色低碳发展的角度探讨创新型城市试点政策效应的文献，不仅丰富了碳排放绩效的相关研究，也为在全国范围内推广创新型城市建设提供了经验证据，同时为打造创新、低碳等新型特色城市提供了政策参考。

Innovative and Low-carbon City: The Impact of Innovative City Construction on Carbon Emission Performance

As the world's largest carbon emitter, China is facing tremendous pressure to reduce carbon emissions. In theory, improving carbon emission performance is widely regarded as one of the most cost-effective ways to combat climate change. For this reason, when discussing the issue of controlling greenhouse gas, it is no doubt to improve carbon emission performance from the perspective of technological innovation. Since 2008, the Chinese government has implemented the pilot policy of innovative city, and has taken green and low-carbon as the principle and goal of innovative city construction. The existing literatures mainly evaluate the policy effect of innovative city construction from the perspectives of urban innovation level, enterprise innovation behavior and FDI quality. However, they pay less attention to the effect of innovative city construction on carbon emission performance, which provides space for this study. Using the pilot policy of innovative city as a quasi-natural experiment, this paper investigates the causal impact of innovative city construction on carbon emission performance, based on the panel dataset of Chinese cities from 2005 to 2016 and the difference-in-differences (DID) method. The empirical results show that compared with non-pilot cities, the carbon emission performance in innovative pilot cities increased by about 2.47%, suggesting that innovative city construction significantly improves the carbon emission performance. After a series of robustness tests such as common trends test, instrumental variables test, confounding policies test and placebo test, this conclusion still holds. Furthermore, the mechanism analysis shows that innovative city construction can strengthen innovation policy support, improve innovation factor concentration, increase innovation investment and ameliorate the innovation environment, thereby promoting technological innovation, optimizing industrial structure and transforming the development mode, which is conducive to reducing carbon emissions and thus enhancing carbon emission performance. From the perspective of dynamic effect, the effect of innovative city construction on carbon emission performance is continuous and increases with time. Compared with the existing literature, the marginal contributions of this paper mainly reflect in the following three aspects. Firstly, to the best of our knowledge, this paper is the first paper in China to explore the policy effects of innovative city construction from the perspective of green and low-carbon development, providing empirical evidence for the promotion of innovative city construction nationwide and enriching research on carbon emission performance. Secondly, based on the estimation framework of the DID method, this paper identifies the net effect of innovative city construction by taking advantage of the variation of pilot policy across different cities and different periods. The advantage of this method is that it effectively avoids the endogeneity problem caused by omitted variables. Thirdly, this paper addresses the endogeneity of innovative pilot cities due to non-random selection. Specifically, on the one hand, the interaction term of the city attribute variables with the time trend polynomial is incorporated in the regression equation to control the impact of inherent attribute differences between cities on carbon emission performance over time. On the other hand, the endogeneity of the pilot policy is further addressed by constructing instrumental variables based on the number of universities in each city.