Whether the Establishment of National High-Tech Zones Can Improve Urban Air Pollution: Empirical Evidence from Prefecture-Level Cities in China

Abstract

Promoting the construction of national-level high-tech zones as green pioneer zones is a necessary condition for achieving high-quality development. Based on panel data from 254 prefecture-level cities from 2006 to 2018, in this paper, the difference-in-difference (DID) is used to empirically examine the influence effect of the establishment of national high-tech zones on local urban air pollution, and the spatial difference-in-difference (SDID) is used to explore its spatial spillover effect. It was found that the annual average ${PM}_{2.5}$ concentration in cities with national high-tech zones decreased by about 1.8% compared with cities without national high-tech zones, and there was a positive spillover effect on the annual average ${PM}_{2.5}$ concentration in nearby cities. Industrial structure upgrading and technological innovation effects are two important transmission paths for the establishment of national high-tech zones to influence urban air pollution; the heterogeneity analysis shows that the establishment of national high-tech zones has more significant implications for the improvement of air pollution in non-resource cities and less developed areas in the west, and the air pollution improvement effect of “growing” national high-tech zones is more desirable than that of “mature” national high-tech zones. Our empirical results conclude that we should continue to encourage the promotion of national high-tech zones, optimize the business environment, improve preferential policies, and design a combination of policy instruments scientifically according to local conditions in order to give full play to the effect of national high-tech zones on urban air-pollution improvement.

1. Introduction

At present, China’s economy has changed from the stage of high growth to the stage of high quality development. Promoting the green development of the social economy is a key link in high quality development, while continuous structural adjustment, especially industrial restructuring, is a necessary condition for green economic growth; this is important to guarantee in order to realize the dual integration and promotion of the economy and the environment. The most widely used industrial policy to promote industrial restructuring is the establishment of development zones [1].

Development zones have long been responsible for stabilizing growth, promoting employment, and increasing exports. Many scholars have conducted quantitative studies on the economic performance of the establishment of development zones, focusing mainly on macroeconomic and micro-enterprise aspects. Many scholars believe that the preferential policies and institutional arrangements of the development zones have had a significant impact on economic development [2,3], industrial structure upgrading [4,5], technological innovation [6], foreign investment [7], and total factor productivity, and firm growth has been identified [8,9]. However, the construction and development of development zones have contributed to both problems and solutions. Some scholars believe that although the policies of development zones have promoted regional economic growth, they have widened the economic gap between cities [10]. On the other hand, at the enterprise level, the preferential policies of development zones have inhibited the innovation capacity of enterprises [11,12].

A review of the existing literature reveals that quantitative studies on development zones have mostly focused on the impact on economic indicators, with fewer studies related to green development, especially the impact on the environmental performance of national high-tech industrial development zones (hereinafter referred to as national high-tech zones), which are capital- and technology-intensive, is still lacking in quantitative assessment. In the literature related to the research question of this paper, some scholars argue that the construction of development zones can reduce pollution emission intensity and improve urban environmental performance [13,14,15,16]. Among them, there are many studies on the quantitative evaluation of environmental performance established by national high-tech zones, focusing on green innovation efficiency, green economic growth, and green total factor productivity [17,18,19]. Therefore, this paper attempts to take the microscopic perspective of environmental performance as the starting point to further explore the influence effect and mechanism of the establishment of national high-tech zones on urban air pollution.

According to the 2020 China Ecological Environment Bulletin, about one-third of cities’ air quality does not meet the national secondary standard, and the frequency of regional air pollution from weather continues to be high, which means that China is currently facing many problems and challenges in air-pollution management. Pollution management is an important part of high-quality economic development; thus, we ask whether national high-tech zones, which gather high technology and new industries, suppress urban air pollution? Based on useful insights from previous scholars, their role in suppressing air pollution should not be ignored. Theoretically, on the one hand, the construction of national high-tech zones can promote the upgrading of industrial structure and thus play a role in improving air pollution; on the other hand, the construction of national high-tech zones can improve the technological innovation capacity of cities and thus have a suppressive effect on air pollution. Therefore, this paper considers the establishment of national high-tech zones as a “quasi-natural experiment” to assess the net effect of the establishment of national high-tech zones on urban air pollution, and tries to verify the suppression effects of industrial structure upgrading and technological innovation on air pollution, as well as to investigate the spatial spillover effect using SDM-DID.

The marginal contribution of this paper is divided into two aspects: firstly, there are a great number quantitative studies on the economic indicators of cities in development zones, but there are few studies on environmental performance, especially from a micro perspective, with air pollution as the entry point, making this study a much needed contribution; secondly, it clarifies how national high-tech zones affect urban air pollution through the “industrial structure upgrading effect” and the “technological innovation effect”, which enrich the relevant studies on national high-tech zones in terms of empirical evidence.

Firstly, scholars have mostly studied the economic and environmental performance of national high-tech zones from 285 prefecture-level cities, but related scholars found that the government prefers to set up development zones in provincial capitals [20], which will lead to more serious endogenous problems, resulting in large deviations in the estimation results. Based on this, the current paper excludes the national high-tech zone of the provincial capital city, and uses PSM-DID as the robustness test, which alleviates the endogenous problem and provides estimation results which are more objective. Secondly, based on the proximity and geographic distance matrices, the spatial double difference method is used to investigate the spatial spillover effects of the establishment of national high-tech zones on urban air pollution, providing empirical evidence for the “positive externalities” of national high-tech zones.

2. Materials and Methods

2.1. The Direct Effect of the Establishment of National High-Tech Zones on Urban Air Pollution

The direct impact of the approval of the establishment of national high-tech zones on urban air pollution mainly comes from two aspects of park assessment and park construction. (i) After the national high-tech zone is approved and established, it will face a more strict assessment mechanism, especially during environmental performance assessment, regarding the comprehensive evaluation index system of national high-tech zones issued by the Ministry of Science and Technology. The level of green development is an important assessment index for performance assessment, under the promotion tournament mechanism, which means officials must deal with the relationship between environmental protection and economic development, thus impacting urban air pollution positively. (ii) In terms of park construction, backward industries such as “three highs” and overcapacity are restricted from entering the park, and the high-tech attribute products of the enterprises in the park occupy an absolute position in the product structure of the enterprises; therefore, the enterprises in the park have a natural “green” attribute. On the other hand, the establishment of national high-tech zones can enhance the level of public service facilities, especially the level of environmental protection infrastructure, through financial and fiscal support. By providing unified pollution-control facilities and services, the cost of pollution control can be reduced and enterprises in the park can be effectively motivated to reduce pollution [21]. In addition, environmental protection facilities in the park will also be shared by enterprises near the park, generating a spillover effect of environmental regulation and achieving a reduction in overall urban pollutants and energy consumption per unit of output value [22]. This will lead to a reduction in pollutants and energy consumption per unit of output and thus to an improvement in urban air pollution.

From the spatial effect, air pollution and other forms of pollution exhibit great differences; air pollution has a high circulation and is cross-regional, air pollution in cities is not only affected by the local socio-economic environment but also by the cross-regional spread of air pollution from cities in close geographical proximity. Due to the mutual influence of economic activities between cities, there must be spatial interconnections, and this influence may force the surrounding cities to absorb the positive external influence brought by the national high-tech zone to produce an air-pollution improvement effect. Based on the above analysis, Hypothesis 1 is proposed:

Hypothesis 1.

The establishment of a national high-tech zone can reduce air-pollution levels, not only in local cities but also in neighboring and geographically close cities.

2.2. Indirect Effects of the Establishment of National High-Tech Zones on Urban Air Pollution

In terms of the industrial structure upgrading effect, national high-tech zones, as the main position of the high-tech industry, can raise the industry threshold of the whole city and produce an extrusion effect, which can drive local industrial structure upgrading at the aggregate level. Secondly, with its preferential tax incentives and talent introduction policies, a national high-tech zone reduces the R&D cost and endogenous financing constraints for independent innovation, which can effectively enhance the enthusiasm and initiative of enterprises in independent innovation, improve labor productivity, optimize factor allocation, and drive the industrial structure from low level to high level [23]. Finally, as a high-tech industry cluster, national high-tech zones have certain technological innovation effects and there are relatively perfect service systems of science and technology, finance, and management in the zones, which create a good technological innovation environment for enterprises in the zones. At present, technological innovation has an extremely important role in promoting industrial structure [24]. At the same time, relevant studies have confirmed that the upgrading of industrial structure produces a significant improvement in urban air quality [25,26]. On the one hand, the upgrading of urban industrial structure can reduce the proportion of traditional high-polluting industries, thus reducing the level of urban air pollution [27], and also has a significant emission reduction effect [28]. On the other hand, the upgrading of industrial structure will also improve the energy-use efficiency of the “three high” industries, thus reducing the energy-use intensity and ultimately reducing the air-pollution level in cities [29]. Therefore, the establishment of national high-tech zones can promote the improvement of urban air quality by adjusting and optimizing the regional industrial structure.

From the perspective of technological innovation, the national high-tech zone mainly reduces urban air pollution through the effect of technological innovation in two aspects: Firstly, the national high-tech zone park itself has the green and green technology research capabilities of enterprises, which are conducive to the generation of clean production processes, driven by the “learning effect”, such that clean-energy pollution treatment equipment is widely used; this can directly improve the level of green technology innovation in the city and subsequently produce a direct impact on urban air pollution. Secondly, due to the limited space-carrying capacity of the national high-tech zone, the industries and innovation elements associated with the enterprises in the zone are not only concentrated in the zone but also distributed around the zone and even in the whole city. Furthermore, the knowledge spillover and diffusion between enterprises and employees inside and outside of the zone form an innovation network covering the whole city, thus driving the whole city to improve the innovation level [30]. At the same time, the effect of technological innovation can suppress air pollution in the city [31]. On the one hand, technological innovation has environment-friendly characteristics, which can effectively reduce environmental pollution by saving energy, promoting industrial structure upgrading, and enhancing population concentration [32]. On the other hand, the improvement of technological innovation has given rise to more high-end pollution monitoring technologies, which enhance the monitoring of pollution emissions from the “three high” industries and improve the efficiency of energy use. Therefore, both clean and non-clean technological advances have an improving effect on environmental quality [33]. Based on the above analysis, Hypothesis 2 is proposed:

Hypothesis 2:

The establishment of national high-tech zones can improve urban air pollution through the industrial structure upgrading effect and the technological innovation effect.

3. Model Setting and Variable Description

3.1. Model Setting

In this paper, the difference-in-difference (DID) is used to assess the policy effects of national high-tech zones on urban air pollution. By the end of 2018, 169 national high-tech zones have been approved in China, which provides a good “quasi-natural experiment” for using the DID method. A total of 119 national high-tech zones were selected as the experimental group and 135 cities without approved national high-tech zones were chosen as the control group after screening and matching.

When using the DID method, the dummy variable of the experimental group and the dummy variable of the control group are set according to whether they are affected by the policy or not: the group affected by the policy is assigned as the experimental group with the value of one, and the group not affected by the policy is assigned as the control group with the value of zero. Meanwhile, the dummy variable of the experimental stage time is set according to the time of policy implementation, and the time of year of policy implementation, and is assigned with a value. Accordingly, the sample can be divided into four groups: the control group before the policy implementation (treat = 0, time = 0), the control group after the policy implementation (treat = 0, time = 1), the experimental group before the policy implementation (treat = 1, time = 0), and the experimental group after the policy implementation (treat = 1, time = 1). Among them, the interaction term treat × time for the two dummy variables of the experimental group and experimental staging is the net effect from the policy implementation.

Since the establishment of national high-tech zones has been approved year by year, not in the same year of unified planning and implementation, this paper assigns a value of one to the 119 prefecture-level cities in the experimental group that are approved as national high-tech zones and a value of zero to the 135 prefecture-level cities in the control group that are not approved as national high-tech zones. Considering the time difference in setting up a national high-tech zone, we assign a value of one to the year of setting up a national high-tech zone and a value of zero to the year before setting up; we then generate the dummy variable DID of setting up a national high-tech zone (DID = treat × time). Finally, a multi-period DID model was constructed to test the net effect of the establishment of national high-tech zones on urban air pollution, based on the practice of Fan et al. (2021). The specific model settings are as follows [34]:

$$Y_{i,t}={\alpha }_0+{\alpha }_1{DID}_{i,t}+{\alpha }_2{X}_{i,t}+{\lambda }_i+{\mu }_t+{\epsilon }_{i,t}$$

(1)

where $Y_{i,t}$ is the explanatory variable indicating the air-pollution level of city i in year t. ${DID}_{i,t}$ is the dummy variable for the approval of the establishment of the national high-tech zone, and ${\alpha }_1$ is the core estimation coefficient, which indicates the net effect of the establishment of national high-tech zones on urban air pollution, refer to Fang et al. (2022). If ${\alpha }_1$ is negative, it means that the establishment of national high-tech zones helps to reduce urban air-pollution levels [35]: there is an elevating effect. $X_{i,t}$ is a set of control variables, including real GDP per capita (Lnpgdp), a quadratic term of real GDP per capita (Lnpgdp)², R&D investment (Rd), population density (Pden), level of service development (Service), level of urbanization (Urban), and environmental regulation intensity (ER). ${\lambda }_i$ indicates urban fixed effects, ${\mu }_t$ reflects the time fixed effects, and ${\epsilon }_{i,t}$ is the random error term.

3.2. Variable Selection

(1)

Explained variables: ${PM}_{2.5}$ as a kind of respirable particulate matter, it is extremely harmful to human health and is a key indicator of concern for the air-pollution status. Therefore, we use the ${PM}_{2.5}$ logarithm of the annual average of surface concentration ${LnPM}_{2.5}$ to measure urban air-pollution levels. The data were obtained from the Columbia University Center for Socioeconomic Data and Applications in 2018, which published global ${PM}_{2.5}$ concentration mean raster data.

(2)

Core explanatory variables: The core explanatory variable in this paper is the national high-tech zone dummy variable DID, which is compiled and assigned according to the list of national high-tech zones in the China Torch Statistical Yearbook of previous years, combined with the approval and establishment time of national high-tech zones, and finally the core explanatory variable DID is obtained.

(3)

Control variables: Based on the existing literature, the following control variables are selected to influence the level of urban air pollution: economic development level using the quadratic term of real GDP per capita and real GDP per capita after deflating the base period of 2006; R&D investment using the proportion of local budget expenditure on science and technology; population density using the number of people per unit of administrative area; service industry development level using the proportion of tertiary industry to GDP; urbanization level using the ratio of non-agricultural population to total regional population at the end of the year; and the intensity of environmental regulation is expressed by referring to Xin et al. (2018), who selected the comprehensive index of environmental regulation intensity by measuring industrial wastewater, sulfur dioxide, and smoke (dust) emissions per unit of GDP [36].

3.3. Mechanism Variables

(i) This paper analyzes the effect of industrial structure upgrading in terms of advanced industrial structure and rationalization of industrial structure, respectively. The advanced measure of industrial structure is as follows: Firstly, the GDP is divided into three parts according to the three industrial divisions, and the proportion of each part to the GDP is taken as a component of the spatial vector, thus forming a set of three-dimensional vectors, $X_0=(x_{1,0},x_{2,0},x_{3,0})$ Then, we calculate $X_0$ The vectors with industries from low level to high level are then calculated separately. The vector of industries from low level to high level: $X_1=(1,0,0)$, $X_2=(1,0,0)$, $X_3=(0,0,1)$ including angles of ${\theta }_1$, ${\theta }_2$, and ${\theta }_3$ [37]:

$${\theta }_j=arccos\frac{\sum _{i=1}^3(x_{i,j}·x_{i,0})}{\sqrt{\sum _{i=1}^3{x}_{i,j}^2}·\sqrt{\sum _{i=1}^3{x}_{i,0}^2}}$$

(2)

Finally, the formula for defining the Industrial Structure Advancement Index (ISA) is as follows:

$$ISA=\sum _{k=1}^3\sum _{j=1}^k{\theta }_j$$

(3)

That is, the angle between the vector of the proportion of three industries and the corresponding coordinate axis reflects the advanced industrial structure. Among them, the larger the ISA value is, the higher the level of advanced industrial structure.

The industrial structure rationalization index refers to the structure-deviation index of Gan et al. (2011) and the Hamming closeness-evaluation method in fuzzy mathematics and combines them to construct the industrial structure rationalization index [38]. The specific formula is as follows:

$$ISR=1-\frac{1}{3}\sum _{i=1}^3\left|\frac{Y_{i,t}}{Y_t}-\frac{L_{i,t}}{L_t}\right|$$

(4)

where ISR denotes the structural deviation degree, i.e., the rationalization indicator of industrial structure, and the larger its value, the more the economy deviates from the equilibrium state. Therefore, a larger value of ISR represents a better match between the output structure and employment structure, i.e., a higher degree of rationalization.

(ii) The technological innovation effect is measured using the number of green patent applications, drawing on Zang and Sun (2021) to examine whether the establishment of national high-tech zones can promote the level of green technological innovation in cities and thus provide technical support for urban air-pollution control fundamentally [39]. The green patent list provided by the World Intellectual Property Organization (WIPO) was searched in the database of the State Intellectual Property Office of China (SIPO) and the relevant data were compiled, considering that some cities have zero green patents, the number of green patent applications was increased by one and then the logarithm taken as the proxy variable of technological innovation.

3.4. Data Sources and Descriptive Statistics

This paper uses panel data from 254 prefecture-level cities in China from 2006–2018 to study the impact of the establishment of national high-tech zones on urban air pollution. The data from national high-tech zones were obtained from the China Torch Statistical Yearbook in previous years; the data of city-level economic indicators were obtained from the China City Statistical Yearbook and the China Regional Economic Statistical Yearbook. In addition, several scholars have used these data in their studies [23], suggesting that they have a high degree of credibility. Some missing data were filled in by consulting the statistical yearbooks of each province or by interpolation. In addition, the sample was selected to exclude the cities that had undergone administrative reorganization at the prefecture-level city level during the study period, such as Chaohu, Bijie, and Tongren, and the Tibetan region was excluded from the study due to the poor quality of the data in Tibet.

Table 1. Definition and calculation methods of variables.

Variable	Meaning	Calculation Method
LnPM2.5	Air pollution	Logarithm of annual average of PM2.5 surface concentration
DID	National high-tech zone	Dummy variable 0, 1
Lnpgdp	GDP per capita	Real GDP per capita after deflating with 2006 as the base period
Lnpgdp2	Squared GDP per capita	Squared real GDP per capita after deflating from the base period of 2006
Rd	R&D	Investment in science and technology expenditure/local budget expenditure
Pden	Population density	Total population at the end of the year/area of the administrative district
Service	Service industry development	Tertiary industry output/GDP
Urban	Urbanization level	Non-agricultural population/total regional population at the end of the year
ER	Environmental regulation	Measurement of industrial wastewater, sulfur dioxide and smoke (dust) emissions per unit of GDP
Lnpatent	Technological innovation level	Ln(1 + number of green patent applications)
ISA	Advanced industrial structure	See introduction in the text
ISR	Industrial structure rationalization	See introduction in the text

and

Table 2. The descriptive statistics of each variable.

Var. Name	Obs	Mean	Std. Dev.	Min	Max
${\mathrm{Y}}_{\mathrm{i},\mathrm{t}}$	3302	3.77	0.329	2.599	4.687
DID	3302	0.26	0.439	0	1
Lnpgdp	3302	10.16	0.72	7.926	12.470
Lnpgdp2	3302	103.8	14.75	62.81	155.5
Rd	3302	0.013	0.0136	0.001	0.207
Pden	3302	0.04	0.0285	0.0019	0.134
Service	3302	0.372	0.0861	0.0207	0.775
Urban	3302	0.36	0.214	0.0759	1.713
ER	3302	0.207	0.443	0.0011	3.037
Lnpatent	3302	4.157	1.621	0.693	8.231
ISA	3302	6.379	0.327	5.634	7.325
ISR	3302	0.864	0.067	0.527	0.995

show the definition and calculation of the variables and the descriptive statistics of the variables, respectively.

4. Empirical Study and Robustness Test

4.1. Baseline Regression Results

Table 3. Baseline regression estimation results.

Variables	(1)	(2)	(3)
DID	−0.018 **	−0.018 **	−0.018 **
	(−2.25)	(−2.28)	(−2.28)
Lnpgdp		−0.062	−0.062
		(−0.70)	(−0.70)
(Lnpgdp)2		0.001	0.001
		(0.14)	(0.14)
Rd		−0.759 ***	−0.759 ***
		(−3.09)	(−3.09)
Pden		−4.451 ***	−4.451 ***
		(−5.64)	(−5.64)
ER		0.008	0.008
		(1.51)	(1.51)
Urban		0.023	0.023
		(0.66)	(0.66)
Service		−0.005	−0.005
		(−0.08)	(−0.08)
Year_trend			−0.020 ***
			(−5.96)
Time fixed effects	YES	YES	YES
Regional fixed effects	YES	YES	YES
Constant	3.840 ***	4.539 ***	4.558 ***
	(836.81)	(9.02)	(9.10)
Observations	3302	3302	3302
R-squared	0.647	0.656	0.656

Note: Standard errors of clustering to city level are in parentheses; **, *** indicate significance at the 5%, and 1% significance levels, respectively, as in the table below.

reports the net effect of the establishment of national high-tech zones on urban air pollution estimated by the DID method, the explanatory variable for the model in Columns (1)–(3) is the logarithm of the annual mean of the surface concentration of ${PM}_{2.5}$. It can be seen that the estimation results are significantly negative at the 5% level with or without the inclusion of control variables, indicating that the establishment of national high-tech zones can significantly reduce urban air-pollution levels in general. In terms of estimated coefficients, the annual average ${PM}_{2.5}$ surface concentration in cities with national high-tech zones decreased by 1.8% compared to cities without national high-tech zones. Although we adopted certain control variables to mitigate the effects of irrelevant economic factors on air pollution, further robustness tests are needed to ensure whether the policy effects are overestimated or underestimated to ensure the truth of Hypothesis 1. In addition, considering that air pollution in different cities may show different time trends over time, this may lead to the erroneous assumption that the establishment of national high-tech zones has led to a reduction in urban air-pollution levels, resulting in a pseudo-regression phenomenon. Therefore, this paper further controls for each city-specific linear time trend in the baseline regression model, and the results are shown in Column (3) of Table 3. These results show that the estimated coefficients are still significantly negative at the 5% level, indicating that the findings of the study still hold.

From the results of the control variables, the primary term of real GDP per capita in the table is negative and the secondary term is positive, and the regression results are not significant, indicating that the environmental Kuznets curve hypothesis does not exist during the study sample period, which is consistent with the findings of Sun et al. (2019) and Shao et al. (2019) [40,41]. Both R&D investment and population density are significantly negative at the 1% level, indicating that the increase in R&D investment is beneficial to technological progress, thus helping to improve energy utilization and pollution control technology; the increase in population concentration is also beneficial to reduce air-pollution levels, similar to the findings of Lu and Feng (2014) [42]. The estimated coefficients of environmental regulation, urbanization level, and service industry development level do not pass the significance test, and their effects on urban air pollution are not yet clear.

4.2. Robustness Test

4.2.1. Parallel Trend Test

Estimation using the DID method is predicated on satisfying the parallel trend test, i.e., there is no systematic difference in the trend of urban air-pollution levels between cities that established a national high-tech zone first and cities that established and did not establish a national high-tech zone at different times before the establishment of the national high-tech zone, i.e., cities in the experimental group and cities in the control group have a common trend in urban air-pollution levels. Therefore, this paper draws on Beck et al. (2010). We use the event study method to test whether the hypothesis of parallel trends is valid by setting dummy variables for several years before and after the establishment of the national high-tech zone [43], and the regression model is as follows:

$$Y_{i,t}=\alpha +\sum _{k\ge -10}^9{\beta }_k{D}_{i,t}^k+\gamma X_{i,t}+{\lambda }_i+{\mu }_t+{\epsilon }_{i,t}$$

(5)

where $D_{i,t}^{k<0}$ ($D_{i,t}^{k>0}$) means that if the sample is the city where the national high-tech zone is established and it is the kth year before (after) the establishment of the national high-tech zone, then it takes the value of one; otherwise, it takes the value of zero. For $D_{i,t}^{k=0},$ the rest of the variables are the same as above. If the parallel trend is satisfied and the establishment of the national high-tech zone does alleviate the urban air-pollution level, this paper expects $D_{i,t}^{k<0}$. The coefficient estimate of $D_{i,t}^{k\ge 0}$ is insignificant, while the coefficient estimates of the national high-tech zone are significantly negative and show variability. As shown in Figure 1, there is no significant difference in urban air-pollution levels among cities before the establishment of national high-tech zones, and they show significant differences after the establishment, which proves that the DID method in this paper satisfies the common trend hypothesis.

4.2.2. Placebo Test

We next further verified that the reduction in urban air-pollution levels is indeed influenced by the establishment of national high-tech zones and not caused by other unobservable factors. In this paper, the chance of the effect of the establishment of a national high-tech zone on urban air pollution is identified, and there are 119 cities in the sample for the establishment of a national high-tech zone. A total of 119 cities were randomly selected from all samples as the new policy intervention group, and it is assumed that these samples are the sample of cities for the establishment of a national high-tech zone, and the other samples are taken as the policy non-intervention group. The sampling is repeated 1000 times, such that 1000 policy effects are obtained. The coefficient and P value distributions are shown in Figure 2. The estimated coefficients of the new policy intervention group are symmetrically distributed on both sides of zero. The distribution of the estimated coefficients is close to the normal distribution, and most of the samples are larger than the baseline regression coefficients; additionally, the p-value is greater than 10% in most samples, i.e., not significant at the 10% level. This suggests that the conclusion that the establishment of national high-tech zones can reduce urban air-pollution levels was not reached by chance and is unlikely to be influenced by other unobservable factors, passing the placebo test and further demonstrating the robustness of the regression results above.

4.2.3. Propensity Score Matching Estimation Results

Since the regression results may be subject to systematic errors due to self-selection problems in the use of DID, the explanatory variables are not random but are the result of a selection process that biases the estimation of the main effects of our study. Propensity score matching is used to reduce such endogeneity problems. In this paper, the variables of economic development level (real GDP per capita, quadratic term of real GDP per capita), urbanization level (Urban), R&D investment (Rd), population density (Pden), environmental regulation intensity (ER), and service industry development level (Service) are selected for matching, while the experimental and control groups are calibrated in the ratio of 1:3, and the calibrated matching results of the equilibrium test are shown in

Table 4. Balance test results.

Control Variables		Average Value		Standard Deviation (%)	Error Reduction (%)	p-Value
		Experimental Group	Control Group
Lnpgdp	Before matching	10.43	9.93	73.8		0.000
	After matching	10.33	10.32	1.2	98.3	0.733
Lnpgdp2	Before matching	109.29	99.04	73.7		0.000
	After matching	107.02	106.84	1.3	98.2	0.713
Urban	Before matching	0.401	0.323	37.5		0.000
	After matching	0.379	0.368	5.2	86.2	0.181
Rd	Before matching	0.018	0.008	74.9		0.000
	After matching	0.015	0.016	−7	96.9	0.248
Pden	Before matching	0.048	0.033	56		0.000
	After matching	0.045	0.045	−1.7	96.9	0.668
ER	Before matching	0.162	0.247	−19.3		0.000
	After matching	0.173	0.19	−3.9	79.8	0.266
Service	Before matching	0.376	0.368	9.7		0.005
	After matching	0.369	0.361	9.1	6.6	0.02

. It can be seen that the standard deviations of the control variables after matching have decreased substantially compared with those before matching, and the p-values of the matched variables are not significant or the standard deviations are less than 10%, which indicates that the model matching results are better; therefore, it is feasible to use the PSM-DID method in this paper.

Based on the analysis above, in this paper, the PSM-DID method is used to improve the randomness of the explanatory variables, reduce the self-selection problem, and make the estimation results more robust. Meanwhile, for the robustness of matching regression results, caliper matching without a 1:3 ratio is listed as a robustness test. The specific empirical results are shown in

Table 5. PSM-DID estimation results.

Variables	1:3 Proximity Matching		Caliper Matching
	(1)	(2)	(3)	(4)
DID	−0.019 **	−0.018 **	−0.022 ***	−0.020 **
	(−2.33)	(−2.18)	(−2.67)	(−2.36)
Controls	NO	YES	NO	YES
Time fixed effects	YES	YES	YES	YES
Regional fixed effects	YES	YES	YES	YES
Constant	3.864 ***	4.753 ***	3.837 ***	4.685 ***
	(743.25)	(7.56)	(807.75)	(8.18)
Observations	2712	2712	3137	3137
R-squared	0.671	0.678	0.646	0.655

Note: standard errors of clustering to city level are in parentheses; **, *** indicate significance at the 5% and 1% significance levels, respectively.

, where the estimated results of PSM-DID are significantly negative at least at the 5% level, regardless of whether control variables are added or what matching method is adopted, thus indicating that the approved establishment of national high-tech zones can significantly reduce urban air-pollution levels, further proving the robustness of the benchmark regression results.

4.2.4. Other Robustness Tests

First, after sorting through the relevant literature, we found that the policies of “emissions trading pilot”, “atmosphere Article 10”, “low-carbon city”, and “smart city” may have an impact on air quality [44,45,46,47], which is incorrectly assumed to be the improvement of air quality brought by the establishment of national high-tech zones. In order to exclude the interference of other policies, this paper constructs a regression model with the intersection of city dummy variables and time dummy variables of “emissions trading pilot”, “atmosphere Article 10”, “low-carbon city”, and “smart city” policies to control the interference of their policies on the results of the baseline regression model, and the regression results are shown in Column (1) of

Table 6. Estimation results of other robustness tests.

Variables	(1)	(2)	(3)	(4)	(5)
DID	−0.017 **	−0.022 ***	−0.017 **	−0.019 **	−0.015 *
	(−2.32)	(−2.81)	(−2.21)	(−2.42)	(−1.69)
Controls	YES	YES	YES	YES	YES
Time fixed effects	YES	YES	YES	YES	YES
Regional fixed effects	YES	YES	YES	YES	YES
Constant	4.458 ***	4.819 ***	4.520 ***	4.977 ***	4.363 ***
	(8.41)	(8.80)	(8.65)	(9.23)	(6.72)
Observations	3302	3003	3302	3048	2286
R-squared	0.661	0.663	0.659	0.669	0.193

Note: standard errors of clustering to city level are in parentheses; *, **, *** indicate significance at the 10%, 5%, and 1% significance levels, respectively.

.

Second, since the government may prefer to set up national high-tech zones in cities with better economic development, this may incur the problem of sample selection bias, i.e., the sample in the study may not be random, even though the study of the article has excluded the provincial capitals with better economic development levels, and the cities with the second highest economic development levels other than the provincial capitals have been removed in this paper for the sake of more pure research results. The regression results are shown in Column (2) of Table 6.

Third, as a result of possible anomalies in the data publication itself and problems in the data collection process, the data were regressed to the 1–99% quantile in order to exclude outlier interference, and the regression results are shown in Column (3) of Table 6.

Fourth, to exclude the potential endogeneity problem of control variables, all control variables were regressed with one period lag in this paper, and the regression results are shown in Column (4) of Table 6.

Fifth, in order to exclude that the policy effect happens to occur only during the time window chosen in the article, this paper chooses different time-window periods for the study, and only the regression results from 2007 to 2015 are presented here. The regression results are shown in Column (5) of Table 6.

In summary, as can be seen from Table 6, all other robustness tests were significantly negative at least at the 10% level, proving the robustness of the baseline regression results.

4.3. Spatial Double Difference Method Regression Results

The analysis above describes how national high-tech zones can improve local air-pollution levels. However, where the establishment of national high-tech zones has an impact on the air quality of neighboring areas—whether due to the mobility of the atmosphere or because a region takes active measures to reduce pollution—it may also be emulated and learned by neighboring regions, thus achieving regional environmental optimization and improving the air-pollution levels of neighboring regions. Considering the above spatial correlation, this paper adopts a spatial difference-in-difference model (SDID) to further explore the spillover effect of national high-tech zones on neighboring cities. Another advantage of adopting SDID is that it can fully control the possible spatial correlation between variables, i.e., control the part of the treatment effect on one individual that changes with other individuals or not, and the potential omitted variables with spatial influence, while also decomposing the average treatment effect in terms of direct effect, spatial spillover effect, and total effect through parameter testing. In this way, we can effectively compensate for the inadequate observation of the strength and direction of the effect of policy implementation on the control group.

4.3.1. Spatial Correlation Test

The global Moran’s index is used to reflect the degree of correlation difference between the attribute values of spatially adjacent or neighboring regions at the global level. The results of the preliminary spatial correlation determination by measuring the global Moran’s index to confirm the existence of spatial dependence of urban air pollution are shown in

Table 7. Moran index results.

Projects	Proximity Matrix	p-Value	Geographical Distance Matrix	p-Value
2006	0.824	0.00	0.205	0.00
2007	0.822	0.00	0.205	0.00
2008	0.785	0.00	0.180	0.00
2009	0.822	0.00	0.201	0.00
2010	0.806	0.00	0.212	0.00
2011	0.798	0.00	0.191	0.00
2012	0.791	0.00	0.181	0.00
2013	0.799	0.00	0.190	0.00
2014	0.761	0.00	0.187	0.00
2015	0.806	0.00	0.226	0.00
2016	0.806	0.00	0.225	0.00
2017	0.772	0.00	0.204	0.00
2018	0.789	0.00	0.220	0.00

. From the calculated results, it can be seen that the global Moran’s index of air pollution in all 254 cities is significantly positive at the 1% level. This indicates that there is a significant positive spatial dependence of air pollution in 254 cities, i.e., cities with strong (weak) pollution levels are surrounded by cities with strong (weak) pollution levels, showing a “high–high” clustering (“low–low” clustering) in the spatial organization pattern of the city. Therefore, it is reasonable to consider the spatial effect in this study.

4.3.2. Estimation Results of the Spatial Double Difference Model

Table 8. Spatial double difference estimation results.

Variables	SDM-SDID		SAR-SDID		SEM-SDID
	(1)	(2)	(1)	(2)	(1)	(2)
DID	−0.001	−0.011 **	−0.001	−0.012 ***	0.001	−0.010 **
	(−0.47)	(−2.56)	(−0.55)	(−2.87)	−0.57	(−2.22)
W × DID	−0.011 **	−0.448 ***
	(−2.51)	(−7.53)
W × Y	0.881 ***	0.982 ***	0.884 ***	0.983 ***
	−137.15	−194.5	−140.28	−212.27
W × E					0.887 ***	0.983 ***
					−142.93	−209.51
Direct effect	−0.008 *	−0.126 **
	(−1.78)	(−2.30)
Indirect effects	−0.087 **	−28.825 **
	(−2.39)	(−2.08)
Total effect	−0.095 **	−28.951 **
	(−2.35)	(−2.08)
Controls	YES	YES	YES	YES	YES	YES
Time fixed effects	YES	YES	YES	YES	YES	YES
Regional fixed effects	YES	YES	YES	YES	YES	YES
Observations	3302	3302	3302	3302	3302	3302
R-squared	0.18	0.043	0.298	0.026	0.284	0.199

Note: (1): geographical proximity matrix; (2): geographical distance matrix; (3): standard errors of clustering to city level are in parentheses; (4) *, **, *** indicate significance at the 10%, 5%, and 1% significance levels, respectively.

reports the spatial model regression results based on geographic proximity and distance weight matrices. Prior to this, this paper sequentially identified the two weight matrices as the optimal choice using the SDM-SDID model with spatio-temporal values fixed by Hausman, LM, LR, and Wald tests. In addition, for the robustness and richness of the results, the regression results of the SAR-SDID model and SEM-SDID model are presented in this paper.

From the regression coefficients in Table 8, the local and spatial spillover effects of the establishment of national high-tech zones on air pollution are significant in three aspects: (i) In the SDM-SDID model based on the estimation of two spatial weight matrices, the W × DID coefficients are significantly negative at the 1% level, indicating that the establishment of national high-tech zones has a spatial effect, which can reduce the air-pollution levels in neighboring cities and cities with a similar geographical distance. (ii) For the air pollution spatial lag term variable W × Y, the spatial double difference model estimates of both spatial weight matrices are significantly positive at the 1% level, indicating that air pollution has a strong positive spatial correlation; the more serious the air pollution in the surrounding areas, the more serious the air pollution in the region, and vice versa—that is, air pollution in the surrounding areas has a spatial spillover effect on the region. (iii) For the spatial lag term variable W × E of the error term, the SEM-SDID model estimates of the two spatial weight matrices are significantly positive at the 1% level, indicating that there is a spatial spillover effect of the random disturbance factors, and the unobserved factors in neighboring cities and geographically close cities will have a negative impact on the local air-pollution level.

In order to more accurately and comprehensively reflect the spatial effect characteristics of the establishment of national high-tech zones on air pollution, the regression coefficients are decomposed into direct effect, spatial spillover effect, and total effect based on the spatial Durbin double difference model (SDM-SDID) by means of partial differentiation. The results of the effect decomposition of the SDM⁃SDID model with two spatial weight matrices can be seen from Table 8: (i) Regarding the direct effect, the direct effect of the establishment of national high-tech zones on urban air pollution is significantly negative at least at the 10% level, indicating that the establishment of national high-tech zones can suppress urban air pollution, which is consistent with the baseline regression results in the previous paper. (ii) Regarding the spatial spillover effect, the spatial spillover effect on urban air pollution driven by the establishment of national high-tech zones accounts for more than 90% of the total effect, which further confirms the important contribution of the spatial spillover effect brought by the establishment of national high-tech zones to the reduction in air-pollution levels in other cities.

In summary, the local effect of the establishment of national high-tech zones is significant, and the spatial effect is obvious for the effect of national high-tech zones. This spatial effect is expressed in the proximity and geographical distance of the similarity. This proves Hypothesis 1.

5. Mechanism Analysis

According to the previous analysis, the establishment of national high-tech zones reduces urban air-pollution levels through the “industrial structure upgrading effect” and “technological innovation effect”. Therefore, in order to verify the existence of the mechanism of action, this paper adopts the test method of mediating effects [48]. The recursive regression equation is used to test the model as follows:

$$M_{i,t}={\gamma }_0+{\gamma }_1{DID}_{i,t}+{\gamma }_2{X}_{i,t}+{\lambda }_i+{\mu }_t+{\epsilon }_{i,t}$$

(6)

$$Y_{i,t}={\chi }_0+{\chi }_1{DID}_{i,t}+{\chi }_2{M}_{i,t}+{\chi }_3{X}_{i,t}+{\lambda }_i+{\mu }_t+{\epsilon }_{i,t}$$

(7)

Among them, $M_{i,t}$ is the mediating variable, and the industrial structure upgrading effect includes industrial structure advancement and industrial structure rationalization; the technology innovation effect includes the logarithm of the number of green patent applications. The remaining formula variables are consistent with the previous benchmark model. If ${\gamma }_1$ and ${\chi }_1$ are significant, and ${\chi }_1$ becomes smaller or less significant as ${\alpha }_1$ becomes smaller or less significant, it indicates a partial mediation effect; if ${\alpha }_1$ and ${\gamma }_1$ are significant, while ${\chi }_1$ is insignificant, it indicates a full mediation effect; if at least one of ${\gamma }_1$ and ${\chi }_2$ is not significant, then a bootstrap test is required. The regression results are shown in

Table 9. Mechanism test estimation results.

Variables	(1)	(2)	(3)	(4)	(5)	(6)
	ISA	Y	ISR	Y	Lnpatent	Y
DID	0.019 **	−0.016 **	0.001	−0.018 **	0.097 *	−0.017 **
	(2.46)	(−2.06)	(0.13)	(−2.27)	(1.82)	(−2.14)
ISA		−0.096 ***
		(−3.70)
ISR				−0.184 ***
				(−3.11)
Lnpatent						−0.016 ***
						(−3.65)
Controls	YES	YES	YES	YES	YES	YES
Time fixed effects	YES	YES	YES	YES	YES	YES
Regional fixed effects	YES	YES	YES	YES	YES	YES
Constant	−2.309 ***	4.317 ***	−2.034 ***	4.164 ***	−7.348 *	4.418 ***
	(−3.65)	(8.49)	(−5.40)	(7.80)	(−1.83)	(8.78)
Bootstrap test	Z = 2.41	p = 0.016	Z = 3.76	p = 0.000	Z = 6.73	p = 0.000
Observations	3302	3302	3302	3278	3302	3302
R-squared	0.865	0.659	0.308	0.695	0.841	0.659

Note: standard errors of clustering to city level are in parentheses; *, **, *** indicate significance at the 10%, 5%, and 1% significance levels, respectively.

.

Columns (1) to (4) in Table 9 show the results of the mediating effect test for the industrial structure upgrading effect, designed to test whether the establishment of national high-tech zones reduces urban air-pollution levels by promoting an advanced and rationalized industrial structure. The regression results show that the establishment of national high-tech zones significantly promotes the industrial structure upgrading, and the coefficient value of the double difference term decreases after adding the indicator term in Column (2). In addition, the regression coefficient of industrial structure upgrading is significantly negative at the 1% level, i.e., industrial structure upgrading plays a part in the mediating effect of the establishment of national high-tech zones in reducing urban air pollution, and the indirect effect passes the bootstrap test. Further, it is calculated that the proportion of the mediating effect in this path is 10.13% of the total effect. In the regression results of Column (3), the establishment of national high-tech zones promotes the rationalization of industrial structure, but it is not significant at the 10% level, and the coefficient of industrial structure rationalization is significantly negative after adding the term of industrial structure rationalization in Column (4). At this point, a bootstrap test is needed, and it is found that the mediating effect of industrial structure rationalization is significant, i.e., the establishment of national high-tech zones can reduce urban air-pollution levels by helping in industrial structure rationalization. Further, it is found that the proportion of the mediating effect in this path is 18.58% of the total effect. Therefore, the establishment of national high-tech zones can suppress urban air pollution through the effect of industrial structure upgrading.

Columns (5) to (6) in Table 9 show the results of the mediating effect test for the technological innovation effect, designed to test whether the establishment of national high-tech zones reduces urban air-pollution levels by enhancing the level of technological innovation. The regression results show that the establishment of national high-tech zones significantly promotes the level of green technology innovation, and the coefficient value of the double difference term decreases after adding the indicator term in Column (6). In addition, the regression coefficient of the green technology innovation level is significantly negative at the 1% level—that is, technology innovation plays a part in the mediating effect of the establishment of national high-tech zones in reducing urban air pollution, and the indirect effect passes the bootstrap test. Further, the proportion of the mediating effect in this path is 8.62% of the total effect. Therefore, the establishment of national high-tech zones improves urban air pollution through the effect of technological innovation, which proves Hypothesis 2.

6. Heterogeneity Analysis

This paper examines whether there are differences in air-pollution effects following the establishment of national high-tech zones in cities in terms of city location, city type, and heterogeneity of the growth cycle of national high-tech zones, in turn enriching the core findings.

6.1. Urban Location Heterogeneity

The above paper analyzed the overall effect of the establishment of national high-tech zones on urban air pollution; however, this analysis based on the overall sample may hide the potential regional differences. On the one hand, there are significant differences in geographical location and resource endowment among cities; on the other hand, the development level of national high-tech zones varies from region to region, which leads to a differentiation of the policy effect of national high-tech zones on urban air pollution. Therefore, this paper examines the regional heterogeneity of the policy effects of national high-tech zones. The regression results are shown in Columns (1) and (2) of

Table 10. Heterogeneity test estimation results.

Variables	(1)	(2)	(3)	(4)	(5)	(6)
DID	−0.016 *	−0.031 *	−0.035	−0.017 **	−0.009	−0.020 **
	(−1.92)	(−1.69)	(−1.38)	(−2.00)	(−0.68)	(−2.26)
Controls	YES	YES	YES	YES	YES	YES
Time fixed effects	YES	YES	YES	YES	YES	YES
Regional fixed effects	YES	YES	YES	YES	YES	YES
Constant	3.014 ***	5.630 ***	5.995 **	4.293 ***	5.364 ***	4.691 ***
	(4.74)	(5.63)	(2.18)	(8.49)	(8.04)	(8.61)
Observations	2457	845	312	2990	2119	2925
R-squared	0.656	0.747	0.606	0.673	0.619	0.655

Note: standard errors of clustering to city level are in parentheses; *, **, *** indicate significance at the 10%, 5%, and 1% significance levels, respectively.

, which show that the estimated coefficients of the effect of the establishment of national high-tech zones on urban air pollution are significantly negative, and the suppression effect of the establishment of national high-tech zones on air pollution in less developed cities in the west is more obvious than that in more developed cities in the east and central regions. The reason for this is that the policy benefits from the establishment of national high-tech zones in developed cities in east and central China reached the upper threshold and showed an obvious law of diminishing margins, while the effect on less developed cities has not yet reached the upper threshold and is still in the ascending channel, so the marginal benefits of the establishment of national high-tech zones are greater for them.

6.2. Test for Heterogeneity of City Types

In this paper, cities are divided into resource-depleted cities and non-resource-depleted cities to analyze the heterogeneity of the air-pollution effects of the establishment of national high-tech zones on different types of cities. In total, 24 resource-depleted prefecture-level cities in three batches in 2008, 2009, and 2011 are collated based on the “National Plan for Sustainable Development of Resource-based Cities” for sub-sample regression estimation, as shown in Columns 3 and 4 of Table 10. It can be found that the sample of resource-depleted cities is not significantly negative, while the sample of non-resource-depleted cities is significantly negative, which indicates that the establishment of national high-tech zones has no significant effect on the improvement of air pollution in resource-depleted cities; this is because the industrial structure of resource-depleted cities is homogeneous and the phenomenon of “low-end locking” makes it difficult for the national high-tech zones in resource-depleted cities to improve air pollution through the industrial structure. It is difficult for national high-tech zones to improve regional air pollution through industrial structure upgrading.

6.3. Growth Cycle Heterogeneity Test

From 1988 to now, national high-tech zones have been constructed and developed for more than 30 years, during which there were two approval climaxes. The first approval climax has been constructed and developed over more than 20 years, while the second approval climax has been constructed and developed over less than 10 years. Considering that different national high-tech zones are in different growth cycles, the national high-tech zones are divided into two categories: the first category is the “mature” national high-tech zones approved in 2010; the second category is the “growth” national high-tech zones approved after 2010. The test results of the impact of the national high-tech zones in different growth cycles on urban air pollution are shown in Columns (5) and (6) of Table 10. It can be found that the effect of “mature” national high-tech zones on urban air pollution is not significant, while the effect of “growth” national high-tech zones on urban air pollution is significantly negative, because the “mature” national high-tech zones have been established for a longer time. The “mature” national high-tech zones were established earlier; thus, the policy effect shows a diminishing marginal effect and reaches the threshold value. In addition, the policy orientation of the early national high-tech zones is mainly based on economic benefits, resulting in the environmental effect of the establishment of the national high-tech zones being poor. In contrast, while in the “growth” period of the national high-tech zone after being approved for establishment, the “growth” national high-tech zone can learn the lessons from the “mature” national high-tech zone and can coordinate the economic development and environmental benefits of the national high-tech zone as early as possible. This makes the policy effect of “growth” national high-tech zones more obvious.

7. Conclusions and Insight

Based on panel data from 254 prefecture-level cities in China from 2006–2018, the net effect of the establishment of national high-tech zones on urban air pollution was analyzed using multi-period double difference and spatial double difference models. The results found that: firstly, the air-pollution level in cities with national high-tech zones is reduced by 1.8% compared to cities without national high-tech zones, and this finding still holds after a parallel trend test, placebo test, endogeneity treatment, and other robustness tests. The analysis of spatial effects shows that national high-tech zones also have a significant positive spillover effect on air pollution in geographically close and neighboring cities. Second, the mechanism analysis shows that national high-tech zones reduce urban air pollution through the industrial structure upgrading effect and the technological innovation effect. Third, from the urban location heterogeneity, the national high-tech zones established in less developed cities in the west have a better effect on reducing urban air pollution than more developed areas in the east and central regions; from the city type heterogeneity, compared with non-resource-based cities, the national high-tech zones established in resource-based cities have a better effect on reducing urban air pollution than non-resource-based cities. In terms of city-type heterogeneity, the effect of reducing urban air pollution is not significant in resource-based cities compared with non-resource-based cities; in terms of the growth cycle of national high-tech zones, the effect of reducing urban air pollution is more significant in growing national high-tech zones compared with mature national high-tech zones.

This paper finds that the establishment of national high-tech zones has complex effects on urban air pollution but, in general, the establishment of national high-tech zones significantly reduces urban air pollution in China, which provides policy ideas for practicing the concept of green development and exploring high-quality development that harmonizes ecological civilization with economic prosperity. In response to the construction of national high-tech zones as a pioneer zone of high-quality development, this paper proposes the following insights:

(1)

There should be reasonable use of the national high-tech zone “demonstration first, radiation driven” effect of the spatial layout, expanding the scope of the pilot to achieve a multi-regional policy and explore the full range of radiation-driven effects, to cement the role of the national high-tech zone in the new era of high-quality development of green pioneer areas.

(2)

The government should strengthen top-level design, further optimize the business environment, create a favorable R&D environment for enterprises and research institutions in the park, increase the enthusiasm in social innovation, improve the quality of green technological innovation, and provide technical support and guarantees for achieving green development, in addition to taking the initiative to support relevant tax, financial, land, and other preferential policies in the process of national high-tech zone construction to absorb domestic and foreign high-tech enterprises. This should actively guide the transformation and upgrading of the city’s industrial structure.

(3)

The construction of national high-tech zones needs to be constantly adjusted and improved in practice, gradually exploring programs suitable for different development stages and different development modes, implementing differentiated policies for the characteristics of different types and regional cities, and at the same time combining national high-tech zones with their own location advantages, industrial development goals, and technology development levels, designing a combination of policy tools according to local conditions and scientific conditions to achieve green development and economic prosperity The policy tools are scientifically designed to achieve green development and economic prosperity in a coordinated and unified manner.