Could the Environmental Regulation Promote Green Innovation? Evidence From China


 Could the environmental regulation promote green innovation? This is a very controversial issue. In view of the fact that the existing literature only studies the relationship between the two, lacks effective heterogeneity research, and pays less attention to the deeper analysis mechanism between the two. This study fills the gap. This paper selects the panel data of 285 prefecture level cities in China from 2000 to 2019 for empirical research. The results show that environmental regulation has a significant and continuous positive impact on green innovation.From the perspective of heterogeneity, we find that cities with higher level of green innovation are suitable to improve the intensity of environmental regulation; Cities with low level of green innovation can not formulate high-intensity environmental regulation policies. The intermediary mechanism shows that under the situation of stricter environmental regulations, producers will pay more attention to the promotion and accumulation of human capital, and provide strong intellectual support for green innovation activities. The adjustment mechanism shows that the cities with high degree of marketization and financial R&D investment are conducive to strengthening the promotion of environmental regulation on green innovation. On the contrary, it weakens the role of environmental regulation in promoting green innovation. In addition, this paper uses SYS-GMM model and selects appropriate instrumental variables to solve the endogeneity problem of the model. We find that after reducing the endogeneity of the model, improving the intensity of environmental regulation can still promote the level of green innovation. Using SDM decomposition model, we find that environmental regulation has spatial spillover effect on green innovation, and the formulation of environmental regulation strategy is conducive to the coordinated development of regional green innovation.


11
Could the environmental regulation promote green innovation? This is a very controversial issue. 12 In view of the fact that the existing literature only studies the relationship between the two, lacks 13 effective heterogeneity research, and pays less attention to the deeper analysis mechanism between 14 the two. This study fills the gap. This paper selects the panel data of 285 prefecture level cities in 15 China from 2000 to 2019 for empirical research. The results show that environmental regulation 16 has a significant and continuous positive impact on green innovation.From the perspective of 17 heterogeneity, we find that cities with higher level of green innovation are suitable to improve the 18 intensity of environmental regulation; Cities with low level of green innovation can not formulate 19 high-intensity environmental regulation policies. The intermediary mechanism shows that under 20 the situation of stricter environmental regulations, producers will pay more attention to the 21 promotion and accumulation of human capital, and provide strong intellectual support for green 22 innovation activities. The adjustment mechanism shows that the cities with high degree of 23 marketization and financial R&D investment are conducive to strengthening the promotion of 24 environmental regulation on green innovation. On the contrary, it weakens the role of 25 environmental regulation in promoting green innovation. In addition, this paper uses SYS-GMM 26 model and selects appropriate instrumental variables to solve the endogeneity problem of the 27 model. We find that after reducing the endogeneity of the model, improving the intensity of 28 environmental regulation can still promote the level of green innovation. Using SDM 29 decomposition model, we find that environmental regulation has spatial spillover effect on green 30 innovation, and the formulation of environmental regulation strategy is conducive to the 31 coordinated development of regional green innovation. China has rapidly promoted the process of industrialization, laying the foundation for the 38 development of "world factory". But the rapid economic growth is at the cost of environmental 39 pollution. As a result, China has quickly become the world's largest energy consumer and 40 greenhouse gas emitter. According to the 2018 global environmental performance index jointly 41 compiled by Yale University and other authoritative institutions in the United States, China's air 42 quality ranks fourth from the bottom among 180 countries and regions, only surpassing India, 43 Bangladesh and Nepal.Among them, the concentrations of PM2.5 and PM10 are far behind the 44 international average annual standard. Climate problems such as "black confusion" and "haze on 45 ten sides" aggravate health inequality, but also bring significant economic losses and reduce social 46 welfare. According to the 2020 "Lancet Countdown China Report", the number of high 47 temperature heat wave related deaths in China has risen by four times since 1990. The economic 48 losses caused by high temperature related deaths in 2019 reached 13.6 billion US dollars, which is 49 equivalent to the economic income of more than 1.3 million people in the same year.Focusing on 50 the protection of ecological environment and promoting the sustainable development of economy 51 and society is the proper meaning. 52 53 In order to curb the continuous deterioration of the ecological environment, the government 54 continues to increase the intensity of environmental regulation, and strive to transform the 55 denotative growth into the connotative growth of economic development mode. Strict 56 environmental regulation is conducive to improving the ecological environment, but supervision 57 will cause distortion of local government incentives. Driven by the "Official Promotion" 58 assessment mechanism with economic growth as the core, local governments reduce 59 environmental standards in order to attract the inflow of capital and production factors. By 60 reducing environmental standards, producers can save "compliance costs" , leading to "bottom-up 61 competition" (Ahlers & Shen, 2018). Under the guidance of the concept of scientific development, 62 local governments include environmental quality indicators in their performance appraisal. As an 63 important link between government environmental regulation and sustainable development of 64 green economy, green innovation has become an important consideration of high-quality 65 economic development (Zhang & Zhu, 2019). Different from traditional innovation, green 66 innovation is an important development strategy of producers, which is related to the improvement 67 of their ability and competitive position (Przychodzen, 2020). Green innovation takes into account 68 both economic and environmental effects, and achieves the expected environmental and economic 69 benefits by increasing green human capital for R&D of green products and new technologies (Li 70 & Zeng, 2020). 71

72
In fact, there are still many controversies about whether environmental regulation can promote 73 green innovation, and there is a lack of clear conclusions in academic circles. There are two main 74 issues in the debate: the "Restriction Theory" and the "Porter Hypothesis". The "Restriction 75 Theory" represented by neoclassical school believes that strict environmental regulation will 76 reduce the competitiveness of producers. Because environmental regulation has greatly increased 77 the operating costs of producers, hindered the desire of producers to actively participate in green 78 innovation research and development activities, and reduced the production of green innovation 79 products. This view is also known as "Compliance Cost Theory". Supporters of the compliance 80 cost theory look for theoretical and empirical evidence around the constraint theory ( the relationship between the two, and the discussion of the important internal mechanism has not 95 attracted the attention of the academic circles. What is the impact mechanism of environmental 96 regulation and green innovation? To clarify the above problems, we can clarify the significant 97 effects and channels between the two, and provide policy enlightenment for China to implement 98 environmental regulation to promote green innovation. 99 100 The existing literature lays a theoretical foundation and logical starting point for this study. Based 101 on this, the marginal contribution of this paper is as follows: first, the research perspective. Due to 102 the problem of data availability at the level of prefecture level cities in China, there are few 103 literatures on this topic at the level of cities. City is the concentrated embodiment of modern 104 economic and social civilization. By obtaining urban micro data, we can ensure that the research  105  results are more reliable and reasonable, and open up a new research perspective. Second, research  106 experience. In this paper, quantitative research is used to provide more abundant empirical 107 information, and dynamic situation is used to investigate the long-term application effect of 108 conclusions. Due to the heterogeneity of the implementation intensity of environmental regulation 109 and the degree of green innovation in different cities, the implementation standards of 110 environmental regulation policies may not meet the requirements of each city. Therefore, this 111 paper strengthens the analysis and discussion of the heterogeneity. In addition, this paper uses 112 SYS-GMM model to solve the endogeneity problem in the existing literature, and reduces the 113 estimation error caused by endogeneity by selecting appropriate instrumental variables. We choose 114 the spatial Dubin model to investigate the spatial effects of environmental regulation and green 115 innovation. Third, research mechanism. What are the specific channels to influence environmental 116 regulation and promote green innovation? There is a lack of discussion on the internal mechanism 117 of this topic in the existing literature. This paper will further introduce the mediating and 118 moderating mechanisms to examine the influence channels of research topics. This study helps to 119 clarify the direction of environmental regulation on green innovation. Through strengthening the 120 analysis of heterogeneity and endogeneity, and clarifying the internal mechanism, it can guide the 121 application of environmental regulation policy tools in green innovation in China. 122

123
The rest of this paper is arranged as follows: the second part puts forward the theoretical 124 hypothesis of environmental regulation and green innovation. The third part describes the use of 125 empirical analysis methods, including standard empirical model, estimation method, index 126 measurement and data sources. The fourth part introduces the estimation results and empirically 127 analyzes the relationship between environmental regulation and green innovation, including 128 heterogeneity analysis and endogenous test. The fifth part discusses the internal mechanism of 129 their relationship, including the mediating effect and regulatory effect. The sixth part is the 130 summary of this study, and puts forward a series of policy recommendations. 131 132 2. Literature review and theoretical hypothesis 133 The impact of environmental regulation on green innovation is mainly based on the 134 comprehensive effect of "Compliance Cost Effect" and "Innovation Compensation Effect". 135 Environmental regulation will produce greater innovation compensation effect to offset the 136 adverse factors brought by the cost of compliance. It mainly comes from the following three 137 reasons: (1) Cost saving. For producers, the cost of not complying with environmental regulations 138 is high. In order to avoid the increase of production cost caused by higher environmental 139 regulation, producers tend to make appropriate changes to reduce the compliance cost. Producers 140 should carry out green process improvement and technology research and development to reduce 141 production costs and improve production efficiency (Meng et al, 2020).
(2) Promote competition. 142 Producers are faced with three kinds of external environmental pressures: the pressure of 143 environmental regulations, the pressure of customers and suppliers' stakeholders, and the pressure 144 of imitation competition to maintain their market share (Cai & Li, 2018). Environmental 145 regulation promotes producers to accept new ideas, stimulate their creative thinking, improve 146 product quality and environmental performance, and gain market competitive advantage (Li et al,147 2019).
(3) Social responsibility. Producers do not always pursue profit maximization, and meeting 148 social needs and goals is also an important factor for them to consider. In order to maintain the 149 legitimate needs of consumers and the government, producers are more willing to carry out 150 transformation, which can help producers gain a sense of social responsibility and achieve 151 additional profits by establishing a green image (Du et al, 2020). In the face of more stringent environmental regulations, producers' pursuit of compliance depends 156 on their ability to understand, absorb and commercialize external knowledge (Leblebici et al, 157 1991). Therefore, having a high level of green innovation will increase organizational flexibility 158 and make it possible to adapt to the higher pressure of environmental regulation. On the one hand, 159 with the improvement of individual or regional green innovation level, the public awareness of 160 environmental protection is constantly enhanced, and the government supervision is 161 correspondingly increased. Regions with higher level of green innovation have higher level of 162 green absorptive capacity and will have the opportunity to get more preferential tax treatment 163 from the government. High intensity of environmental regulation will encourage regions to 164 increase green innovation investment, and promote producers, especially those with high pollution 165 and high emission, to obtain competitive advantage in green transformation (Delmas,  H2: Strong green innovation area is suitable for improving the intensity of environmental 176 regulation, while weak green innovation area is not suitable for making high-intensity regulation. 177 178 Under the pressure of environmental laws and policies, producers must take active measures to 179 deal with environmental challenges (Chan, 2005). When producers implement green innovation 180 strategy, managers integrate producer resources through green performance management and 181 compensation practice, cultivate internal staff and introduce innovative talents, so that they can 182 make greater efforts to produce new ideas, methods and actions, and effectively enhance human 183 capital (Ma et al, 2019). According to Romer's human capital theory, human capital is of great 184 significance to productivity growth (Romer, 1990). Human capital is the basic element of 185 knowledge economy and the core strategic resource of sustainable competitive advantage (Song & 186 Yu , 2018). Regions with high human capital have stronger technology attraction and diffusion 187 ability, which is conducive to the promotion and application of new technologies, and tend to 188 accept the concept of environmental protection and comply with environmental regulations, which 189 is conducive to reducing environmental pollution (Bano, 2018). Human capital should increase the 190 use of green technology, effectively avoid the "Resource Curse" and transform it into "Resource 191 Gospel" by giving full play to skilled equipment operation and technology application ability, so 192 as to realize the knowledge driven sustainable economic development mode (Singh, 2020). More and more attention has been paid to market-oriented green innovation in China. The market 201 mechanism can stimulate the enthusiasm of producers, and the market price can reasonably reflect 202 the scarcity of resources and the supply and demand of products, both of which provide reasonable 203 guidance for production allocation (Filipovic, 2019). Effective market is conducive to cross 204 internalize the externality of pollution, reduce the marginal cost of pollution control, and obtain 212 more compensation for emission reduction . In addition, under the imperfect market 213 mechanism, due to unclear property rights, regions with more abundant natural resources are more 214 likely to face rent-seeking, corruption and opportunism, which will hinder the development of in green innovation activities. Endogenous growth theory shows that the intensity of financial 229 R&D investment has a positive impact on the production sector, and efficient public R&D system 230 can make better use of private R&D funds (Conte, 2013). Financial R&D investment can send a 231 "Signal" to the market, guide producers and society to participate in green R&D, cultivate green 232 innovative talents, alleviate the problem of lack of funds for relevant producers, and reduce the 233 cost and risk of green R&D (Yi, 2020). The data of environmental regulation comes from the work report of the city government, and is 245 mined by Python technology. The instrumental variable air circulation coefficient is based on era 246 interim meteorological data published by European Center for weather forecasting (ECMWF), and 247 is extracted by overlaying grid data with urban base map through ArcGIS. The other indicators are 248 from the statistical yearbook of Chinese cities and the statistical yearbook of local cities.In order 249 to reduce the heteroscedasticity of data, we use logarithm to deal with the data. We use 1% 250 winsorize to deal with the main continuous variables. 251 252

3.2.Model construction 253
Considering the impact of environmental regulation on green innovation, this paper takes the level 254 of green innovation as the explained variable and environmental regulation as the explanatory 255 variable. In addition, there may be inertia in the level of green innovation, that is, the previous 256 intensive level has an impact on the current intensive level. A dynamic model is constructed as 257 trend and a positive correlation. This means that the concept of sustainable development is deeply 301 rooted in the hearts of the people, the government continues to improve the environmental laws 302 and policies system, and green innovative technology has made a breakthrough. 303 304 Fig.1 Time sequence characteristics of ER and GI 305 Fig.2 shows the spatial distribution of environmental regulation and green innovation (Taking 306 2019 as an example). We find that environmental regulation presents the spatial distribution 307 characteristics of "Differentiation". This may be due to the wide coverage of China's land, 308 unbalanced carrying conditions of regional resources and environment, regional development 309 stage, industrial structure and environmental governance needs and other reasons, forming a 310 "Classified Guidance" policy system. 311

312
The spatial distribution of green innovation has two main characteristics: First, the "Cluster 313 Effect" is significant, forming the spatial pattern of "Four Cores, Three Clusters, Two Belts and 314 Many Points". With Beijing, Shanghai, Guangzhou and Shenzhen as the "Four cores", it radiates 315 the "Urban Agglomerations" of Beijing Tianjin Hebei, Yangtze River Delta and Pearl River Delta. 316 The eastern coastal economic belt and Yangtze River economic belt form a "Two Belt" distribution, 317 and Wuhan, Xi'an, Chongqing and other cities form a multi-point "Growth Pole". Cities rely on 318 their economic development, resource endowment, human capital, innovation factors and location 319 policies to promote the development of urban green innovation level. Second, "Matthew Effect" is 320 highlighted. The green innovation level of East China and South China is absolutely superior. In 321 addition, regional industries complement each other and have strong comprehensive 322 competitiveness, which further attract green innovation elements to accelerate agglomeration. 323 Therefore, the difference of green innovation development will be further expanded.

4.1.Benchmark regression 357
The baseline regression results are reported in Table 2

4.2.Endogeneity test 373
Because the level of R&D tendency and expenditure of producers will also affect the 374 environmental regulation faced by producers, there is a two-way causal relationship. In order to 375 solve the endogenous problem, this paper refers to Fisman & Svensson (2007) and Hering & 376 Poncet (2014), and takes the average level of environmental regulation (lnPREG) and air 377 circulation coefficient 3 (lnVC) as the instrumental variables of environmental regulation. Taking 378 the average value of environmental regulation in different cities as an instrumental variable is not 379 directly affected by the behavior of a single producer, but the average level of a city is directly 380 related to the explanatory variable, so it can be used as an instrumental variable of environmental 381 regulation. According to the ArcGIS software, the grid data is superimposed on the map of 382 Chinese cities, and the air circulation coefficient of the corresponding city and year is 383 matched.The air circulation coefficient only depends on natural phenomena such as regional 384 climate conditions. In addition to affecting the degree of environmental regulation, there is no 385 other mechanism between air circulation coefficient and green innovation, which meets the 386 requirements of "Correlation" and "Exogenous" hypothesis, so it can be used as a instrumental 387 variables of environmental regulation.  According to the first law of geography, there is correlation between things, and the correlation 409 between things closer is higher than that between things farther (Tobler, 1970). In order to enhance 410 the robustness of the model and whether there is spatial spillover, we use spatial model to measure 411 spatial dependence. The spatial Durbin model considers the correlation of independent variables 412 and the correlation between independent variables and dependent variables in adjacent areas. It is 413 a composite model of spatial autocorrelation model and spatial error model. Therefore, the spatial 414 Dobbin model is selected to analyze the spatial relationship between environmental regulation and 415 green innovation. Before spatial model analysis, we need to use Moran index to test the correlation of core 424 indicators. The results show that the overall Moran index of green innovation and environmental 425 regulation have passed the 1% confidence level test (Schedule 1), indicating that the overall 426 spatial correlation degree of the two is high. The local Moran index is positively correlated (Fig. 1  427 and Fig.2), which indicates that there is a local spatial correlation between them, and the spatial 428 model can be used for further analysis. Table 4 reports the regression results of the dynamic spatial 429 Durbin decomposition model, in which (1) is the direct effect model, (2) is the indirect effect 430 model, and (3) is the total effect model. The results show that environmental regulation has a 431 positive correlation with local green innovation level (β= 0440, P < 0.005, Model 1), and has a 432 significant role in promoting the level of green innovation in adjacent areas (β= 1670, P < 0.005, 433 Model 2). It shows that environmental regulation has spatial spillover effect. With the gradual 434 progress of ecological civilization construction, the central and local governments have frequent 435 strategic interaction in environmental protection. The formulation of environmental regulation 436 strategy has gradually changed into "Top-To-Top Competition", which is conducive to the 437 coordinated development of regional green innovation (Peng, 2020). 438 439

.Heterogeneity analysis 444
Each city has the characteristics of unbalanced development, so it is necessary to investigate the 445 policy differences of environmental regulation on local green innovation development in different 446 cities.In order to verify whether this difference exists, this paper takes the sample median of 447 environmental regulation and green innovation as the boundary, and divides the research objects 448 into strong regulation (RQ) and weak regulation cities (RR), strong green innovation (GQ) and 449 weak green innovation cities (GR), strong regulation and strong green innovation cities RQ&GQ), 450 weak regulation and weak green innovation cities (RQ&GQ). For the division of areas, refer to Fig.  451 3 and Fig. 4.

5.1.Mediation and regulatory effect model 480
In order to further explore the relationship between environmental regulation and green innovation, 481 we still need to study its internal mechanism. This paper draws on the intermediary and regulatory 482 effect mechanism model constructed by Chan (2021) to identify the mechanism of environmental 483 regulation on green innovation. On the basis of (1), model (2) (3) is added to the test process of 484 mediating effect, and the test process of moderating effect is (4) -(5). 485 regulation has a significant positive impact on green innovation (β= 0907, P < 0.001, Model 1). 507 Every 1% increase in environmental regulation, 0.0238% increase in human capital and 0.182% 508 increase in green innovation (β= 0238, P < 0.001, Model 2; β= 182, P < 0.001, Model 3). The 509 mediating effect accounted for 4.7757%, which verified hypothesis H4. The formulation of 510 environmental regulation will significantly improve the level of human capital, and the 511 improvement of human capital has a significant positive impact on green innovation. It shows that 512 human capital plays an important role in this process. Under the condition of stricter 513 environmental regulation, producers tend to pay more attention to the cultivation of green 514 innovative talents and encourage talents with high green innovation potential. The improvement of 515 human capital brings new knowledge and new technology to green innovation activities, and is 516 conducive to improving the level of green innovation (Song et al, 2021). 517 518  Table 7 reports the regression results of the regulatory effects of marketization and financial R&D 524 investment. The results show that marketization has a positive impact on green innovation (β= 525 0687, P < 0.001, Model 1). By introducing the cross product of environmental regulation and 526 marketization (lnREG×lnMar), the marketization of regulatory variables has a significant positive 527 effect on green innovation (β= 0295, P < 0.001, Model 2). It shows that when the degree of 528 marketization is high, the promotion of environmental regulation on green innovation is 529 strengthened. On the contrary, it weakens the role of environmental regulation in promoting green 530 innovation. In order to strengthen the role of the market, the guiding opinions on building a 531 market-oriented green technology innovation system issued by the national development and 532 Reform Commission and the Ministry of science and technology clearly states that enterprises 533 must participate in green innovation projects with clear market orientation, and the proportion of 534 green R&D projects supported by major national science and technology projects and national key 535 R&D plans led by enterprises shall not be less than 55%. 536 537 Many scholars believe that financial R&D investment crowns out private investment and ignores 538 the important role of public R&D expenditure.In the field of environmental governance and green 539 innovation, financial R&D expenditures crowd out producer investment, which is not conducive to 540 improving the level of green innovation (Chervier, 2019). But is that really the case? The results 541 show that financial R&D investment has a positive impact on green innovation (β= 298, P < 0.001, 542 Model 3). By introducing the cross factor of environmental regulation and financial R&D 543 investment (lnREG×lnPte), the impact of financial R&D investment on green innovation also has 544 a significant positive regulatory role (β= 0267, P < 0.001, Model 4). It shows that financial R&D  Note: In parentheses denote the standard error of the respective coefficients, ***/**/* indicates 558 the significance at the 1%/5%/10% levels, respectively. 559 560

561
The core of green innovation is to promote resource conservation and governance, environmental 562 friendliness and governance. Environmental regulation has become an important means to 563 promote green innovation. This study has important theoretical value and practical significance for 564 clarifying the relationship between environmental regulation and green innovation, and promoting 565 pollution control and environmental protection from the perspective of green innovation. We use SYS-GMM model, and select the explained variable of lag period, the mean value of 587 environmental regulation and air circulation coefficient as the instrumental variables. After 588 reducing the endogeneity of the model, improving the intensity of environmental regulation can 589 still promote the level of green innovation. Using the spatial Durbin decomposition model, we find 590 that environmental regulation has spatial spillover effect on green innovation. The formulation of 591 environmental regulation strategy has gradually changed into "Top To Top Competition", which is 592 conducive to the coordinated development of regional green innovation. technology to optimize the management system, improve the efficiency of green human capital 608 allocation mechanism. 609 610 Third, the combination of efficient market and promising government. We need to give full play 611 to the role of market mechanism, innovate the supply and demand mechanism of green technology, 612 and enhance the market value of green technology.The government should increase the R&D 613 investment support for green key technology, explore the transfer and benefit distribution 614 mechanism of funded achievements projects to inventors and small and medium-sized enterprises, 615 and strengthen the innovation guiding leverage of financial R&D investment.