The Impact of Environmental Regulation on EGIE

The theory of environmental regulation suggests that governmental environmental regulatory policies are an effective means of addressing market failures in green technological innovation. The existing literature has gradually developed two different views in the process of studying the impact of environmental regulation on the efficiency of green innovation: the pollution paradise hypothesis and Porter’s hypothesis.

From a static point of view, the neoclassical economic school believes that strict environmental regulation policies will increase the production costs of enterprises and form a pollution paradise in regions with low environmental regulation, so environmental regulation cannot effectively promote regional green innovation efficiency. Some studies have found that due to the decentralized governance structure and performance appraisal mechanism, the government tends to give up the idea of long-term green economic growth in order to achieve short-term economic development, i.e., the existence of race-to-the-bottom at all levels of local governments, which then leads to the spatial spillover of environmental pollution. Around this view, Zhang et al. [4] point out that local governments often reduce the “compliance cost” of local enterprises by weakening the intensity of environmental regulation to achieve the goal of promoting economic growth, so even if the region’s environmental regulatory policies are very strict, the benefits of environmental governance inputs may be very limited. Baumeister et al. decomposed the environmental regulation stringency index, the OECD regulatory indicators, and the green total factor productivity growth index using 19 EU countries as examples [5]. The empirical results found that market-based environmental regulations have a negative impact on the pure and scale efficiency indices of green total factor productivity. Guo et al. argued that the scale of taxes on environmental protection in environmental regulatory instruments is too small to incentivize firms to invest in green eco-innovation [6]. Jin and Shen find that the interaction of environmental regulation enforcement in geographically neighboring municipalities exhibits asymmetry, with a race to the bottom and a race to the top co-existing, forcing less productive polluting firms to relocate between municipalities [7]. These observations support the existence of a “cost effect” of environmental regulations on the efficiency of green innovation. Li &Yang pointed out that the strengthening of environmental regulations makes the government and enterprises increase the cost of environmental protection and pollution control, which inhibits the output performance of enterprises and economic development [8].

The Porter hypothesis [9] posits the dynamic view that strict environmental regulation policies can actually have a net positive impact on the innovation capacity of regulated firms [5-9]. As the policy promotes cost cutting and higher levels of efficiency, which in turn reduces or completely offsets the costs of environmental regulation and facilitates the application of innovation in new technologies, it can help firms achieve international technological leadership. Sterlacchini et al. [10] empirically found that the stringency of market-incentivized environmental regulatory instruments in the EU region had a more positive facilitating effect on green innovation patents in 19 OECD countries. Li found that environmental regulation and government R&D funding have a positive impact on enterprises’ green technology innovation, and the coupling of the two has a stronger effect on green technology innovation [11]. Bao and Chai [12] empirically found that environmental regulation is positively correlated with regional green innovation efficiency, and the government improves the efficiency of green innovation by improving the allocation of financial resources [10-12]. Therefore, the study of Porter’s hypothesis further proves that there is an “innovation effect” on the impact of environmental regulation on green technology.

In summary, there is no consensus on the impact of environmental regulation on the efficiency of green innovation. Generally speaking, environmental regulation acts on enterprises’ green innovation activities mainly through two mechanisms: the negative “compliance cost effect” and the positive “innovation compensation effect”. Some scholars believe that the technological innovation effect of environmental regulation is sufficient to offset the cost of environmental governance and has a facilitating effect on the improvement of green development efficiency. However, some scholars hold the opposite view that the “cost compliance effect” of environmental regulation is greater than the “innovation incentive effect”, thus inhibiting urban green development. Some scholars hold the opposite view that the “cost compliance effect” of environmental regulation is greater than the “innovation incentive effect”, thus inhibiting urban green development. Based on the above analysis, the following hypotheses are proposed:

H1: Environmental regulation has a facilitating effect on regional cross-border e-commerce green innovation efficiency.

The Intermediary Effect of Industrial Structural Upgrading

Environmental regulation has a direct impact on green technology innovation efficiency, but it also has an indirect impact through certain factors. To study the impact of industrial structure upgrading on environmental regulation on the green innovation efficiency of regional cross-border e-commerce, we mainly analyze two aspects: one is the role of environmental regulation on the impact of industrial structure upgrading, and the other is the role of industrial structure upgrading on the green innovation efficiency of regional cross-border e-commerce.

(1) The impact of environmental regulation on industrial structure upgrading. As one of the economic tools used by regional managers to control environmental pollution, the impact of environmental regulation on the upgrading of regional industrial structure needs to be realized through the two paths of advancement of industrial structure and rationalization of industrial structure. Based on the development trend of industrial green transformation, many scholars have studied the impact of environmental regulation on industrial structure upgrading at different scales, such as national, city cluster, watershed economy, specific industries, micro enterprises, etc., and their research conclusions are inconsistent. Some scholars hold the “promotion theory” on the impact of environmental regulation on industrial structure upgrading, believing that environmental regulation can play the “innovation compensation effect”, enhance the overall competitive advantage of the industry, and promote industrial structure upgrading. Some scholars hold the “inhibition theory”, believing that the “cost of compliance effect” of environmental regulation increases the burden of enterprises, raises the entry barriers of the industry, and inhibits the upgrading of industrial structures. Another scholar holds the “uncertainty theory” and believes that the impact of environmental regulation on industrial structure upgrading is related to the size of the “innovation compensation effect” and the “cost of compliance effect”. In addition, most scholars start from the supply perspective and believe that environmental regulation mainly promotes industrial structure upgrading by stimulating technological innovation in enterprises, attracting foreign direct investment, and facilitating factor mobility. Some scholars also pay attention to the demand side and take the growth of social demand as an important way for environmental regulation to affect industrial structure upgrading. Huang [13] argued that environmental regulation helps to stimulate the rationalization of industrial structure by improving the efficiency of factor replacement and also helps to enhance the advancement through industrial structure of factor upgrading [6, 10, 13].

(2) In terms of the impact of industrial structure upgrading on regional green innovation efficiency. Local governments should formulate industrial policies according to the industrial structure, factor endowment, and green transformation development strategy of different regions. Through the agreement on the division of labor and cooperation of green products and green technologies, green technology diffusion and inter-provincial green knowledge spillover can be realized so as to promote the spatial concentration of green innovation activities in the region. This not only reduces the uncertainty and risk of green innovation activities, but also lowers the cost of green science and technology research and development and the transformation of green results, which ultimately brings about a change in regional green innovation efficiency. In addition, as the industrial structure tends to be advanced, the production factors are transferred from the low-efficiency industrial sectors to the high-efficiency industrial sectors, and the spatial reset of innovation resources in this process will inevitably have an impact on the performance of green technological innovation. Green leading industries with high green production efficiency will drive the growth of other emerging industries, thus promoting the innovation of upstream and downstream industries, expanding the scale of regional green innovation, and promoting the enhancement of regional green innovation capacity and green innovation efficiency. Scholar She et al. empirically found that low-carbon pilot policies indirectly enhance the green total factor productivity of low-carbon pilot cities through the channel of promoting industrial upgrading [14]. Lu et al. used a finite mixture model to empirically analyze that industrial structure upgrading and technological innovation have a significant promotional effect on regional green total factor productivity and that there are obvious “leading-type”, “catching-up-type” and “lagging-type” green total factor productivity [15, 16]. Based on the above analysis, the following hypothesis is proposed:

H2: Environmental regulation promotes industrial structure upgrading, thereby improving regional cross-border e-commerce green innovation efficiency.

The Regulatory Effect of Intellectual Property Protection

In dealing with the contradiction between environmental protection and economic growth, it is necessary for the government to take the lead in adopting binding instruments for environmental control. At the same time, it is also necessary to adopt incentives and utilize intellectual property rights to protect the results of innovations in order to maintain and increase the incentives of enterprises to innovate.

Weak awareness of intellectual property protection in cross-border e-commerce transactions, negligence, and insufficient supervision by the relevant departments have led to numerous cases of intellectual property infringement in cross-border e-commerce, violating the interests and intellectual property rights of the relevant people, destroying the good business environment, and hindering the innovative development of cross-border e-commerce. In China’s theoretical research, the issue of intellectual property protection in e-commerce has also attracted more and more attention from the academic community. For example, Cui [17] believes that the current intellectual property rules have many problems, such as incomplete rules and unclear infringement liability, especially in the current e-commerce legal system in China, which is mainly based on internet information as the object of control and lacks a specification of the commercial activities of the platform concerned. Luo [18] believes that intellectual property rights are the driving force for enterprises to develop cross-border e-commerce businesses. Intellectual property is at the core of the development of cross-border e-commerce businesses, which is particularly important in the era of the knowledge economy, and the legal protection of intellectual property should be strengthened to improve the legal system of cross-border e-commerce [17, 18].

As an important institutional arrangement to encourage the production of new knowledge, intellectual property protection policy has attracted extensive attention from scholars. Zheng and Sun [19] found that the incentive effect of the establishment of intellectual property courts on the quality of firms’ innovation was more significant for state-owned enterprises and enterprises in the Pearl River Delta Economic Circle, but not for private enterprises [15, 17, 19]. While some scholars are against it, Zhang et al. found that strengthening Intellectual Property Rights protection would inhibit international technology spillovers, which in turn would have a negative impact on technological innovation [20]. Based on the above analysis, the hypothesis is formulated:

H3: The interaction of environmental regulation and intellectual property protection has a negative moderating effect on regional cross-border e-commerce green innovation efficiency.

The Threshold Effect of Industrial Agglomeration

Industrial agglomeration is the main manifestation and means of realizing industrial allocation, and at the same time, it is also a necessary condition for guaranteeing the green development of the region. The impact of industrial agglomeration on green innovation efficiency can be analyzed in terms of the “industrial synergy effect” and the “pollution transfer effect”.

(1) Industrial synergy. Geospatial agglomeration is conducive to the transmission of information, and with the deepening of the division of labor and cooperation among enterprises, a complete industrial chain is gradually formed, and enterprises in the agglomeration area can share information, infrastructure, labor markets, and specialized services, so as to be able to effectively control production costs. Therefore, under a certain level of industrial agglomeration, environmental regulation can promote the spillover and dissemination of green knowledge and promote green technological innovation.

(2) Pollution transfer effect. When the scale of industrial agglomeration reaches a certain level, problems such as environmental degradation, insufficient resources, population congestion, traffic congestion, and other problems gradually appear, the advantages of the agglomeration gradually dissipate, and the scale effect of the agglomeration area is no longer significant. Therefore, industrial enterprises with low green technology content and high pollution emissions will gradually shift to adjacent areas due to the early loss of comparative advantages, and the negative environmental effects of adjacent areas will continue to increase, which will hinder the green development of the region.

Accordingly, this paper formulates the hypothesis:

H4: There is a threshold effect of environmental regulation on regional cross-border e-commerce green innovation efficiency under different levels of industrial agglomeration.

System GMM Model Settings

The most commonly used estimation methods for panel data models are the fixed effects model and the random effects model, but there is an endogeneity problem in the model when the lagged terms of the explanatory variables are introduced as explanatory variables in the regression model, making the model dynamically explanatory. Therefore, Arellano et al. proposed the so-called Generalized Moment Method (GMM), and the GMM estimation method of the dynamic panel data model was divided into DIF-GMM and SYS-GMM [21]. Roodman argued that although the differential GMM method reduces the impact of endogeneity on model estimation, it suffers from a serious “weak instrumental variables” problem under limited sample conditions, which leads to poor accuracy of coefficient estimation results [22]. For this reason, Blundell et al. proposed a GMM estimation method to solve the problems of autocorrelation, heteroskedasticity, endogeneity, and weak instrumental variables [23]. Meanwhile, Roodman pointed out that the necessary condition for applying the system GMM model was that the number of cross sections should be larger than the time dimension. As the number of periods increases, the system GMM generates by default a large number of instrumental variables that may outweigh the endogenous variables and weaken the model setting test [7, 15, 21]. Therefore, to make the results more reliable, we prefer a two-step system GMM approach for estimating panel data.

In this paper, we take regional cross-border e-commerce green innovation efficiency as an explained variable and introduce the first-order lag term of green innovation efficiency and environmental regulation as a core explanatory variable to satisfy the condition setting that the explained variables are dynamic and not all explanatory variables are strictly exogenous. At the same time, the individual and temporal double fixed effects are also controlled to satisfy the conditions for the use of the system GMM model. In addition, there are many factors affecting green innovation efficiency, so other factors affecting green innovation efficiency are introduced as control variables. Among them, the control variables include level of economic development (AGDP), level of urbanization (UR), foreign direct investment (FDI), energy consumption structure (ECS), and level of human capital (HCL). In the equation, i denotes the industry individual, t denotes the year, and ε denotes the error term. Meanwhile, the model also controls the region fixed effect 𝛿, year fixed effect 𝜇, and 𝜀i as the random error term, 𝛽₀ is a constant, and 𝛽_i is the regression coefficient of the corresponding variable. Considering that the transformation of regional cross-border e-commerce green innovation efficiency results requires a time cycle, the green innovation efficiency of the current year not only depends on the performance of the current period, but is also affected by the green innovation efficiency stock of previous years. Therefore, the green innovation efficiency (EGIE_it-1) with a lag of one period is added to the model to control its own inherent shocks. The resulting dynamic panel regression model is shown below:

In the process of analyzing the impact of environmental regulation on regional cross-border e-commerce green innovation efficiency，the interaction between environmental regulation and intellectual property protection should be considered at the same time as a measure of the circular cumulative effect caused by the impact of knowledge protection behaviors other than environmental regulation. Therefore, the moderating effect model is:

Selection of Variables

Data Sources

Due to the lack of data for Hong Kong, Macao, Taiwan, and Tibet, as well as the availability of the sample, panel data for 30 provinces (municipalities and autonomous regions) in China were used for the analysis from 2011-2021, with a sample observation of 330. The data come from the China Urban Statistical Yearbook, the China Financial Yearbook, the China Science and Technology Statistical Yearbook, the China Statistical Yearbook, the China Energy Statistics Yearbook, and the statistical yearbooks of prefectural-level municipalities. Based on the fact that the outliers in the sample years may cause the regression results to be less robust, this paper adopts the “shrinking method” to shrink the data samples by the upper and lower 1%. All data are logarithmically processed for each variable to ensure the robustness of the variable data, and some of the missing data is filled in by linear interpolation. The descriptive statistics of each indicator are shown in Table 2.

Pre-Testing and Data Transformations

We use the panel unit root test to determine that the continuous variables used in the regression model are stationary. The unit root test proposed by Maddala and Wu [28] uses the Fisher’s test, which defines the null, and alternative hypotheses of this test are defined as:

H0: All panels contain unit roots;

H1: At least one panel is stationary.

The results of the panel unit root test on the relevant variables are reported in Table 3. As can be readily seen, Table 3 shows that for the variables except HCL, the null hypothesis that all panels contain unit roots is strongly rejected by the test. The null hypothesis of unit root is rejected for the other tested panel variables. However, the null hypothesis was not rejected for the variable HCL, and we performed a second difference before rejecting the null hypothesis.

The Pearson correlation matrix and variance impact factor between the study variables were then utilized to check for multicollinearity between the independent variables. As shown in Table 3, the VIF values of all variables except variable UR are less than 5, indicating that the problem of multicollinearity among variables is limited. Therefore, we can continue to estimate the panel model.

Main Effect and Regulatory Effect Regression Analysis

The results of model regression are shown in models (1) and (2) in Table 4. The use of system GMM needs to determine the instrumental variables validity over-identification test. The Hansen test of the P-value is greater than 0.1 and does not reject the original hypothesis, indicating that the selection of instrumental variables in the system GMM model is valid and there is no over-identification problem. In addition, the system GMM model needs to utilize the Arellano-Bond test for the autocorrelation of the random perturbation term, and the P-values of the AR (1) test of the model are all less than 0.1; the P-values of the AR (2) test are all greater than 0.1, which indicates that there is no problem of second-order autocorrelation of the residual term in the system GMM model. Therefore, the model setting of environmental regulation on regional cross-border e-commerce green innovation efficiency is reasonable and effective.

Table 4 Model (1) shows the effect of environmental regulation on regional cross-border e-commerce green innovation efficiency after considering the first-order lag term of green innovation efficiency. According to model (1), the first-order lagged term of the explained variable has a significant promotion effect on the current green innovation efficiency at the 1% level, with an impact coefficient of 0.179, which indicates that the regional cross-border e-commerce green innovation efficiency has inertia with a significant cumulative growth effect, and it is necessary to continuously accumulate green innovation technology and take green innovation capacity building as the focus of regional strategic development. The effect of environmental regulation on green innovation efficiency is significantly positive at the 1% level, with a coefficient of 0.105, and hypothesis H1 is verified. This indicates that under the facilitating effect of environmental regulation, the innovation main body in the region achieves greater effectiveness in energy saving, emission reduction, and green production technology, which significantly and positively guides and promotes the generation of regional green innovation efficiency.

Model (2) in Table 4 shows the regulatory effect of intellectual property protection in environmental regulation on regional cross-border e-commerce green innovation efficiency. Among them, the regression coefficient of the cross-multiplier term of intellectual property protection and environmental regulation negatively affects the regional cross-border e-commerce green innovation efficiency at the 1% significance level, with a coefficient of -0.270, which indicates that the stronger the intellectual property protection is, the more obvious is the inhibiting effect of environmental regulation on the regional cross-border e-commerce green innovation efficiency, and the H3b hypothesis is verified. The reason may be that when the intensity of intellectual property protection is too high, in order to avoid higher innovation risks and costs, enterprises will choose to carry out non-environmentally friendly technological innovations with low technological difficulty and fast returns to ensure a certain level of returns. This will largely crowd out the government’s R&D investment in environmental regulation innovation, forming a crowding-out effect on the enterprise’s green innovation capacity, which is not conducive to green technological innovation.

Regression Analysis of Intermediary Effects

In order to explore the role of environmental regulation on the path of industrial structure upgrading, this paper constructs the mediation effect model in the following equation: Industry is the intermediary variable, representing industrial upgrading. The c is the total effect coefficient of environmental regulation affecting the efficiency of green innovation. The a is the coefficient of environmental regulation on the mediator variable. The b is the coefficient of the mediator variable’s effect on the efficiency of green innovation, and the c′ is the coefficient of the direct effect of environmental regulation affecting the efficiency of green innovation, formula (5).

In the mechanism of industrial structure upgrading shown in Table 4, the coefficients c, a, and b are 0.1820, 0.1538, and 0.1581, respectively, all of which pass the test of significance at least 90% confidence level, and the coefficient c′ is also significant, which suggests that industrial structure upgrading has a partially mediating effect on the influence of environmental regulation on green innovation efficiency. Model (4) shows that environmental regulation has a significant positive effect on industrial structure upgrading (coefficient = 0.1538), and model (5) shows that industrial structure upgrading also has a significant positive effect on the green innovation efficiency of cross-border e-commerce (coefficient=0.1581). Combined with the theoretical analysis, it can be seen that the impact of environmental regulation on regional cross-border e-commerce green innovation efficiency can be realized through the path of empowering industrial structure upgrading, which supports hypothesis H2.

Threshold Regression

This paper draws on the new method of threshold regression proposed by Hansen [30] to verify whether there is a threshold effect of government subsidy-based environmental regulation on regional cross-border e-commerce green innovation efficiency and constructs a threshold model as in Equation (6):

Where I () is the indicator function, AGGi is the three threshold variables, respectively, and γi is the threshold value, when the values of the threshold variables in parentheses satisfy the conditions, the indicator function takes the value of 1, and 0 anyway. Table 5 shows the number of thresholds and the threshold values obtained by repeated sampling 300 times using bootstrap. The results show that the current stage of industrial agglomeration passes the single threshold and double threshold tests at 1% significance, so the double threshold test is used to analyze it. The LR trend is shown in Fig. 1, and the LR values corresponding to the first and second thresholds of 0.0072 and 0.0202 fall below the critical value of 5% and tend to be close to 0, i.e., the estimated thresholds are equal to the real thresholds, and the industrial agglomeration can be more accurately classified as three stages.

The results of the parameter estimation of the double-threshold model are shown in Table 5. Environmental regulation has a positive effect on regional cross-border e-commerce green innovation efficiency. When the level of financial agglomeration is lower than the threshold (AGG≤0.0072), the influence coefficient of industrial agglomeration on green innovation efficiency is 0.2891, and the significance test of 1% indicates that industrial agglomeration has a positive significant impact on green innovation efficiency under this condition. When the level of industrial agglomeration is greater than the threshold (AGG>0.0202), the effect of industrial agglomeration at the 1% level on the promotion of green innovation efficiency decreases significantly (coefficient = 0.1277). Therefore, industrial agglomeration has different effects on the regional cross-border e-commerce green innovation efficiency at different levels. When the level of industrial agglomeration increases, the industrial ecological environment and the corresponding supporting facilities in the region where it is located are also significantly improved, and the positive externality of industrial agglomeration takes the advantage. When the scale of industrial agglomeration exceeds the threshold, environmental degradation, lack of resources, population congestion, traffic jams, and other problems gradually appear, the scale effect of the agglomeration area is no longer significant. Thus, industrial enterprises with lower green technology content and higher pollution emissions will lose their comparative advantages due to earlier losses, which in turn hinders the green development of the region.

Characterization of Spatial Differentiation

In order to realize China’s goal of crossing into the ranks of innovative countries, the Ministry of Science and Technology issued a circular in April 2016 on the guidelines for the construction of innovative provinces, with Jiangsu, Sichuan, Anhui, Hunan, Fujian, Shaanxi, Zhejiang, Hubei, Guangdong, Shandong, and Jilin as pilot innovative provinces [31]. In addition, Beijing, Shanghai, Chongqing, and Tianjin are considered pilot innovative municipalities. This paper explores the impact of environmental regulation on the regional cross-border e-commerce green innovation efficiency based on the innovative province and municipality pilots as the differentiation condition [2, 32, 33]. Among the innovative provinces and municipalities, eight eastern regions, four central regions, and three western regions are included.

Fig. 1.

Trend in LR for double-threshold estimates.

Clarifying the spatial trend change of the regional cross-border e-commerce green innovation efficiency helps to grasp the process and mechanism of cross-border e-commerce development evolution as a whole. In this paper, with the help of the ArcGIS 10.8 software trend analysis tool, the spatial trend maps of 30 provinces and municipalities in 2011, 2016, and 2021 are drawn, and the results of the trend surface fitting are shown in Fig. 2. From the trend surface, it can be seen that in the north-south direction, the green innovation efficiency shows a U-shaped pattern, indicating that the green innovation efficiency in the north-south region is higher than that in the central region during the sample period, and the north-south region is almost at the same level over time. In the east-west direction, the green innovation efficiency shows an “inverted U” pattern, characterized by high in the middle and low on both sides, indicating that the green development efficiency of the eastern and western regions is lower than that of the central region during the sample period. Over time, the trend curve rises from west to east, the green innovation efficiency of the eastern region increases significantly, and the trend curve evolves from an “inverted U” to a “zigzag pattern”, indicating that the gap between the central region and the northern and southern regions gradually decreases. During the study period, the transition of the trend surface in the north-south direction is steeper, while the transition of the trend surface in the east-west direction tends to flatten out, which means that the divergence of green innovation efficiency in the north-south direction is more obvious and strong. It is found that the regional cross-border e-commerce green innovation efficiency, as an important manifestation of the attractiveness and competitiveness of “Internet e-commerce + manufacturing” in each region, is highly consistent with the spatial configuration of the regional economic development level and the abundance of industrial clusters.

Fig. 2.

Trend surface analysis of EGIE.

Robustness Tests

The model replacement method is used to examine the stability of the evaluation methods and indicators, and if the same results can be obtained, it indicates that the data conclusions are stable and reliable, so this paper utilizes the differential GMM model to test the stability of the data in the robustness type test. As shown in Table 6, the regression coefficients and significance of the first-order lag terms of regional cross-border e-commerce green innovation efficiency, environmental regulation, and the cross-multiplier terms of environmental regulation and financial agglomeration in Models (7) and (8) are consistent with Models (1) and (2) in Table 4, which proves that the results of the main effects test are stable.

Heterogeneity Analysis

Model (9) is the sub-sample regression of non-innovative provinces (municipalities) regions, and model (10) is the sub-sample regression of innovative provinces (municipalities) regions. As shown in Table 6, the first-order lagged terms of the regional cross-border e-commerce green innovation efficiency are positively correlated with the current green innovation efficiency at the 1% significance level.

The environmental regulation in innovative provinces (municipalities) has a promotional effect on the regional cross-border e-commerce green innovation efficiency at the 1% significance level, with a coefficient of 0.060, which is consistent with the conclusion of the main effect test. The environmental regulations in non-innovative provinces (municipalities) have a promotional effect on the regional cross-border e-commerce green innovation efficiency, but there is no significant relationship. This may be due to the fact that the 15 non-innovative provinces (municipalities) include 12 provinces in the central and western regions, which have a relatively low degree of economic development and lag behind in the application of environmental regulations compared with the eastern regions, resulting in the weak ability to transform the results of green innovation and high-tech and the failure to form the competitiveness of green technological innovation [34, 35].