The sources of air pollutants and CO2 are basically the same, hence the reduction of air pollutants will affect CO2 emissions. Considering the regional integration of economic development as well as air pollution control, it is necessary to analyze the impact of air pollutants reduction in a region on CO2 emissions in its surrounding regions. Furthermore, as different stages of air pollutants reduction have different effects on CO2emissions, it is also important to study the heterogeneity of this impact. In this article, we took China as the research case and built a spatial panel model based on the data of 240 cities above the prefecture level from 2005 to 2016 to study the impact of two different stages of air pollutants reduction—front reduction of air pollutants (FRAP) and end-of-pipe treatment of air pollutants (EPAP) on CO2 emissions—and their spatial spillover effects. On this basis, we further modified traditional spatial weight matrix and constructed the matrices of cities in the same and different provinces to discuss the influence of provincial administrative boundaries on the spillover effect between cities. The results show that FRAP affects CO2emissions mainly through the local synergistic effect, and its spatial spillover effect is not significant. The local effect of EPAP on CO2emissions is antergic, and the spatial spillover effect is significant. The increase of a city’s EPAP will increase the CO2 emissions in surrounding regions. Besides, provincial boundaries weaken the spatial spillover effects of FRAP and EPAP on CO2 emissions in prefecture-level cities. There is a significant spatial spillover effect between cities in the same province, but the spillover effect does not exist for cities in different provinces nearby.
Research Article
Spatial effects of air pollutants reduction on CO2 emissions
https://doi.org/10.21203/rs.3.rs-1984572/v1
This work is licensed under a CC BY 4.0 License
Journal Publication
published 22 May, 2023
Read the published version in Environmental Science and Pollution Research →
You are reading this latest preprint version
Synergetic reduction
Air pollution
CO2
Spatial spillover effects
CO2 and conventional air pollutants, such as sulfur dioxide and nitrogen oxides, come from the same source—the combustion of fossil fuels. Theoretically, there is a synergistic effect between policies and measures aimed at reducing air pollutants and CO2 emissions (Dong et al., 2015; Nam et al., 2014; Xu et al., 2021). Therefore, it is possible to achieve the goals of improving air quality and reducing carbon emissions simultaneously and at a lower cost.
Air pollutants reduction not only affects local CO2 emissions, but also has spatial spillover effects. There are two main reasons. First, economically developed regions often have higher requirements for environmental quality, thus pollution-intensive industries may migrate to areas with weaker policies (Chen et al., 2018; Zhang et al., 2020; Zhou et al., 2021b). Consequently, air pollutants reduction in a region has a negative spillover effect on pollutants and CO2 emissions in surrounding regions (Feng et al., 2020). Second, stronger governance can affect enterprises’ production and pollution control technologies (Porter and Vanderlinde, 1995). Owing to geographical proximity, regional integration, as well as industrial agglomeration, the air pollutants reduction in a region may indirectly affect CO2 emissions in neighboring regions through the flow of factors such as capital and labor (Meng et al., 2017; You and Lv, 2018; Zhou et al., 2021a). Analyzing the spatial spillover effects of air pollutants reduction on CO2 emissions helps formulate more efficient coordinated reduction policies of CO2 and pollution thus realize regional air quality and climate goals.
China is one of the most air polluted countries in the world. In the initial years, China’s air pollution was mainly due to primary pollutants such as SO2 (Lu et al., 2011; Song et al., 2017). In recent years, the emission of primary pollutants has reduced, but the problem of complex air pollution remains serious (Huang et al., 2014; Wang et al., 2017). Facing serious air pollution, the energy conservation and emission reduction policies of developing countries like China have long been aimed at improving air quality (Hasselmann et al., 2003; Lelieveld et al., 2015). Through a series of air pollutants reduction policies, China has achieved remarkable results in air pollution prevention and control (Jiang et al., 2015; Yue et al., 2020; Zheng et al., 2018), but the air quality is still far from international standards. Apart from air pollution, China also emits the most CO2 in the world. According to World Bank, China’s CO2 emissions experienced rapid growth from 1978 to 2013, an increase of approximately 5.8 times. After 2013, the growth rate of China’s CO2 emissions slowed down, but total emissions still showed an upward trend. In response to the increasingly severe CO2 emissions problem, China has proposed several CO2 reduction goals, including medium and long-term goals of achieving peak carbon dioxide emissions by 2030 and carbon neutrality by 2060. Facing the dual tasks of reducing air pollutants and CO2 emissions, China’s efforts on the collaborative reduction has strong reference significance for other countries.
There are lots of studies exploring the impact of China’s air pollutants reduction on CO2 reduction. Nam et al.(2014) used the CGE model to compare the synergistic effects between air pollution control and climate control in the United States and China. They found that as the strictness of SO2 and NOX control increases, the resulting CO2 reduction effect also increases in the two countries. Further, the CO2 reduction effect in China was higher than in the United States. Peng et al.(2017) found that air pollutants reduction measures in China’s industrial sector can effectively reduce CO2 emissions. For example, a 10% increase in energy efficiency in the industrial sector can reduce CO2 emissions by 440 Mt. If 80% of small coal-fired power plants are replaced by large high-efficiency coal-fired power plants, CO2 emissions will reduce by 53 Mt. Qian et al.(2021) used enterprises’ data at a micro level and set up different scenarios to simulate the impact of air pollutants reduction measures on CO2 reduction. They found that improving energy intensity, phasing out small firms, and electrification are effective for achieving the synergistic effect of CO2 and air pollutants reduction. Some scholars have studied the synergistic effect of CO2 and air pollution in regions of China. Chen et al.(2006) used the MARKAL model to predict the synergistic emission reduction of air pollutants and CO2 in Shanghai. They found that Shanghai’s environmental policy reduced the annual growth rate of CO2 emissions from 2.7% in 2000 to 1.1–1.2% in 2020. Liu et al.(2013) utilized the GAINS model to study the synergistic effect of air pollutants reduction and CO2 reduction under different scenarios in Beijing. The study found that starting from 2005, air quality management measures can effectively reduce CO2 emissions by 5–22% by 2030. Xu et al.(2021) found that Air Pollution Prevention and Control Action Plan significantly reduced CO2 emissions in Beijing, Tianjin, and Hebei by 20.7 Mt (19.7%), 6.8 Mt (3.8%), and 80.2 Mt (9.2%), respectively.
By analyzing existing research, we find that studies on synergistic emission reduction generally focuses on the local impact, while studies on the spatial effects of air pollutants reduction looks at its spillover effect on SO2, PM2.5, and other air pollutants. However, few studies explore the spatial spillover effects of air pollutants reduction on CO2 emissions. Additionally, not all air pollutants reduction measures can synergistically reduce CO2 emissions, especially for end-of-pipe treatment of air pollution (EPAP), which may increase CO2 emissions due to higher energy consumption (Qian et al., 2021; Tan et al., 2016). Existing research mainly focused on the CO2 reduction effect caused by the overall treatment of air pollution. Nevertheless, few studies paid attention to the heterogeneity impacts of different stages of air pollutants reduction on CO2 emissions.
To fill the above research gap, this study built a spatial panel model based on the panel data of 240 cities above the prefecture level in China from 2005 to 2016. We divided air pollutants reduction into two stages—FRAP (front reduction of air pollutants) and EPAP, and studied the impact of FRAP and EPAP on CO2 emissions and their spatial spillover effects. Simultaneously, combined with the regional characteristics of China’s air pollution prevention and control, this study incorporated provincial administrative boundaries into the spatial panel model to further discuss the impact of administrative boundaries on the spatial spillover effects between cities.
The remainder of this paper is organized as follows. In Section 2, we describe the construction method of the spatial panel model and related indicators and the source of panel data. Further, we present our research results on the local and spatial spillover effects of FRAP and EPAP on CO2 emissions in Section 3. Subsequently, in Section 4, we discuss the impact of provincial administrative boundaries on the spatial spillover effects to provide supports for the formulation of cross-region policies to reduce air pollutants and CO2. Finally, we draw our conclusions in Section 5.
2.1 Methodology
FRAP in this paper refers to reducing the amount of pollutants generated from the source through measures such as shutdown of outdated production facilities, improvement of production technologies, and optimization of production processes. EPAP refers to reducing the amount of pollutants emitted into the environment through the use of end-of-pipe treatment technologies. Considering the representativeness and availability of data, we selected the SO2 production per unit of GDP in each city to represent FRAP and the removal rate of SO2 to represent EPAP.
This study used Moran’s I to investigate the spatial correlation of related indicators.
$${\text{M}\text{o}\text{r}\text{a}\text{n}}^{{\prime }}\text{s} \text{I}=\frac{N}{{S}_{0}}\frac{\sum _{i=1}^{N}\sum _{j=1}^{N}{w}_{ij}({x}_{i}-\stackrel{-}{x})({x}_{j}-\stackrel{-}{x})}{\sum _{i=1}^{N}{{(x}_{i}-\stackrel{-}{x})}^{2}}$$
1
where \(N\) is the number of individuals, \({x}_{i}\) and \({x}_{j}\) are the observed values of individuals \(i\) and \(j\), respectively, and \(\stackrel{-}{x}\) is the average of all observations. \({S}_{0}\) is the sum of all elements of the spatial weight matrix, and \({w}_{ij}\) is the element in the \(i\)-th row and \(j\)-th column of the weight matrix.
We further constructed a spatial panel model based on the STIRPAT theory, in which the level of economic development was represented by GDP per capita, the population effect by population density, and the technology effect by the proportion of science and technology expenditure in government expenditure. Moreover, combined with the results of existing literature, we added other relevant control variables. Fossil fuel combustion in the secondary industry is a major source of CO2 emissions. Therefore, this study selected the proportion of the value added by the secondary industry to GDP in each city to represent the role of the secondary industry within the industrial structure. The CO2 emissions from transportation account for approximately 10% of the total CO2 emissions in China and the CO2 emissions from road transportation account for 74% of the total CO2 emissions from transportation. Hence, this study selected the number of cars in each city to represent transportation effect. The degree of foreign trade openness also has an impact on CO2 emissions, and there is no consensus on the effect of trade openness on CO2 emissions. On the one hand, the “pollution halo” hypothesis suggests that foreign investment can promote a country’s introduction of green technologies to reduce CO2 emissions (Xie et al., 2020; Zhang and Zhou, 2016). On the other hand, the “pollution haven” hypothesis contends that foreign investment may increase a country’s CO2 emissions through the transfer of energy-intensive and pollution-intensive industries (Baek, 2016; Ren et al., 2014). Therefore, this study selected foreign direct investment (FDI) to represent the degree of openness to the outside world. On this basis, we added the core explanatory variables of this study—FRAP and EPAP in the model. Furthermore, using Grossman’s EKC theory, this study included the level of economic development as a quadratic term in the model and analyzed the nonlinear relationship between CO2 emissions and the level of economic development.
The commonly used spatial panel models in existing research mainly include the spatial lag model (SLM), the spatial error model (SEM), and the spatial Durbin model (SDM). The first two can be seen as a simplification of SDM. SDM can be expressed as follows:
where \({W}_{ij}{X}_{jt}\theta\) represents the spatial lag term of each control variable, which is not expanded here. In this paper, the spatial weight matrix in the model adopts the K nearest neighbor matrix, where K equals 10.
LeSage and Pace (2009) pointed out that using partial derivatives to calculate the direct and indirect effects of explanatory variables can reduce estimation errors, we thus used this method to analyze the direct, indirect, and total effects of the factors. The maximum likelihood estimation was used to avoid the biased estimation of spatial parameters and standard deviation. We calculated Moran’s I using the STATA 16.0 software and performed spatial regression analysis using the Matlab 2020b software.
2.2 Data
This study exploited panel data of 240 cities above the prefecture level in China from 2005 to 2016. Data on the amount of production, removal, and emission of SO2 in each city are from China City Statistical Yearbook 2006–2017 and the statistical bureau of each city. Data on the dependent variable, CO2 emissions, are from China Emission Accounts and Datasets (Chen et al., 2020), in which the CO2 emissions of each city is the sum of the emissions of the districts and counties within its jurisdiction. Data on other control variables are from the China City Statistical Yearbook, the statistical bureau and statistical bulletin of each city and the province where the city is located. The GDP per capita data were deflated using 2000 as the base year. The FDI data were converted from USD to RMB according to the exchange rate of each year and deflated using 2000 as the base year.
3.1 Spatial correlation
The Moran’s I of core variables are calculated based on equation 1, and the results are shown in Table 1. The table shows that the Moran’s I of CO2 emissions in each city from 2005 to 2016 fluctuates between 0.193 and 0.253, and all Moran’s I are significant at the 1% level. Its average value in the study period equals 0.231. This result shows that the CO2 emissions of various cities have a positive and significant spatial correlation—the higher the CO2 emissions in a city, the higher the CO2 emissions in its surrounding cities, and vice versa. Therefore, the spatial distribution of CO2 emissions in cities in China is not random but indicative of the phenomenon that some cities tend to agglomerate in space.
Table 1. Results of Moran analysis.
|
Year |
Y |
End |
Front |
|
2005 |
0.239*** （9.19） |
0.053** （2.12） |
0.047** （2.11） |
|
2006 |
0.237*** （9.09） |
0.071*** （2.81） |
0.034 （1.57） |
|
2007 |
0.248*** （9.49） |
0.071*** （2.81） |
0.042* （1.89） |
|
2008 |
0.253*** （9.68） |
0.108*** （4.18） |
0.016 （0.87） |
|
2009 |
0.240*** （9.21） |
0.110*** （4.26） |
0.035* （1.77） |
|
2010 |
0.235*** （9.01） |
0.094*** （3.67） |
0.040** （2.04） |
|
2011 |
0.220*** （8.45） |
0.013 （0.64） |
0.043** （2.11） |
|
2012 |
0.220*** （8.45） |
0.008 （0.47） |
0.075*** （3.92） |
|
2013 |
0.196*** （7.57） |
0.026 （1.11） |
0.053*** （2.90） |
|
2014 |
0.193*** （7.43） |
0.044* （1.81） |
0.063*** （3.31） |
|
2015 |
0.208*** （7.97） |
0.077*** （3.02） |
0.073*** （4.56） |
|
2016 |
0.208*** （7.98） |
0.151*** （5.80） |
0.098*** （4.03） |
Note: ***, **, and * indicate significance levels of 1%, 5%, and 10%, respectively. The z values are in parentheses.
We further calculate the Moran’s I of FRAP and EPAP. Table 2 shows that, except for the period 2011–2013, the Moran’s I of the variable End is significant at the 10% level or lower in the hypothesis testing. The fluctuation range in the significant years is from 0.053 to 0.151, and the mean value is 0.096. Except for 2006 and 2008, the Moran’s I of variable Front is significant at the 10% level or lower in the hypothesis testing. The fluctuation range in the significant years is from 0.035 to 0.098, and the mean value is 0.057. Therefore, there is a significant positive spatial correlation between the levels of FRAP and EPAP in cities in China, and there is a cluster phenomenon in the spatial distribution of air pollutants reduction. An increase in the level of air pollutants reduction in one region can increase that in its surrounding regions, and vice versa.
3.2 Estimation results of the spatial panel model
First, we tested the model selection. The results show that the LR test results reject the null hypothesis that the spatial fixed effect is not significant, and the time fixed effect is not significant at the 1% level. Therefore, this study adopts the spatial and time fixed effect model. The LM-lag and LM-error of the spatial and time fixed effects model reject the null hypothesis at the 1% level, the Robust LM-lag rejects the null hypothesis at the 5% significance level, and the Robust LM-error rejects the null hypothesis at the 1% significance level. This result indicates a significant spatial effect in the model. Simultaneously, according to Moran’s I, the core explanatory variables of this study also have a significant spatial correlation. As this study focuses on how a city’s FRAP and EPAP affect the CO2 emissions of surrounding cities, we need to consider the spatial effect of independent variables in the model. In summary, the SDM is a more suitable choice.
In terms of the SDM, equation 5 shows that when , it can be simplified to SLM, and when , it can be simplified to SEM. is the matrix composed of the regression coefficients of independent variables. Therefore, before estimating the SDM, we should perform a Wald test on the hypotheses: and . If the test rejects both hypotheses, it is more reasonable to select the SDM (Elhorst, 2012). In addition to the Wald test, this study adds a likelihood ratio test to further verify whether the SDM can be simplified. Table 4 shows that both LR-SLM and Wald-SLM reject the null hypothesis that “the model can be simplified to a spatial lag model” at the 1% significance level. Simultaneously, LR-SEM and Wald-SEM also reject the null hypothesis that “the model can be simplified to a spatial lag model” at the 1% significance level. Hence, SDM cannot be simplified to SLM or SEM in this study. Considering the result of model testing and the purpose of this research, this study will use the SDM for estimation.
Table 2. Estimation results of spatial panel models.
|
Variable |
SLM |
SEM |
SDM |
|
|
0.010*** （5.98） |
0.010*** （5.84） |
0.009*** （5.53） |
|
|
0.008*** （5.07） |
0.005*** （3.02） |
0.006*** （4.00） |
|
|
0.123*** （9.11） |
0.142*** （9.95） |
0.136*** （9.54） |
|
|
-0.026*** （-7.57） |
-0.032*** （-8.49） |
-0.030*** （-7.87） |
|
|
0.042*** （3.56） |
0.036*** （2.98） |
0.035*** （2.97） |
|
|
-0.003 （-0.90） |
0.001 （0.20） |
0.001 （0.28） |
|
|
0.077*** （4.48） |
0.122*** （6.77） |
0.121*** （6.71） |
|
|
0.017*** （3.23） |
0.020*** （3.59） |
0.021*** （3.85） |
|
|
0.004*** （3.00） |
0.006*** （4.53） |
0.006*** （4.53） |
|
|
|
|
-0.006 （-1.47） |
|
|
|
|
0.025*** （5.65） |
|
|
|
|
0.039 （1.16） |
|
|
|
|
0.015* （1.91） |
|
|
|
|
0.023 （0.75） |
|
|
|
|
-0.013* （-1.82） |
|
|
|
|
-0.249*** （-5.79） |
|
|
|
|
0.007 （0.51） |
|
|
|
|
-0.010** （-2.97） |
|
|
0.687*** （38.48） |
|
0.709*** （37.19） |
|
|
|
0.761*** （45.07） |
|
|
R-squared |
0.995 |
0.993 |
0.995 |
|
Log-likelihood |
4258.71 |
4279.57 |
4325.83 |
|
LR (simplified to sar) |
|
|
134.23*** |
|
LR (simplified to sem) |
|
|
92.51*** |
|
Wald (simplified to sar) |
|
|
126.61*** |
|
Wald (simplified to sem) |
|
|
81.51*** |
Note: ***, **, and * indicate significance levels of 1%, 5%, and 10%, respectively. The t values are in parentheses.
The estimation results of the spatial panel model are shown in Table 2. The spatial lag term of CO2 emissions is positive and significant at the 1% level, and the regression coefficient is approximately 0.70. This further verifies the significant spatial correlation of CO2 emissions between various cities in China. A 1% increase in CO2 emissions in a city will increase the average CO2 emissions of its surrounding cities by about 0.70%. Simultaneously, the spatial terms in the SLM and the SEM are significant at the 1% level, which further verifies the necessity of including the spatial effect in the model.
The local effects of FRAP and EPAP are both significantly positive at the 1% level, which indicates that the less SO2 produced per unit of GDP, the lower the CO2 emissions. The higher the SO2 removal rate, the higher the CO2 emissions. As shown in Fig.A1, the trends of FRAP and EPAP in China from 2005 to 2016 indicate that FRAP reduces CO2 emissions, and EPAP increases CO2 emissions. Currently, the methods of FRAP mainly include the following: (1) shutting down outdated production facilities through industrial structure adjustment, thereby reducing energy use; (2) improving energy efficiency through energy technology development or energy quality improvement; (3) optimizing production processes to reduce energy use. These FRAP methods can reduce the generation of air pollutants by reducing energy use and improving energy efficiency. As the main source of CO2 and air pollutants such as SO2 is the combustion of fossil fuel, the abovementioned methods can reduce the amount of air pollutants and CO2 emissions at same time, which shows synergistic effect.
The removal rate of air pollutants is an important indicator to monitor the level of air pollutants reduction of various enterprises in China as well as evaluate the implementation of air pollutants reduction by governments at all levels. Compared to FRAP, EPAP has a faster effect and lower cost in reducing pollutant emissions, making it the preferred solution for many industrial enterprises. However, pollutants removal equipment consume electricity during operation and generates additional CO2, besides, some pollutants removal processes also generate CO2 (Mao et al., 2012; Qian et al., 2021; Shi et al., 2017; Tan et al., 2016). For example, if all sintering machines in China achieve ultra-low emissions, when 70% of machines adopt the SCR denitration process, the additional CO2 emissions from this process can reach 32 million tons a year (Ministry of Ecology and Environment of China, 2020). Therefore, although EPAP can reduce pollutants emissions, it increases CO2 emissions in the pollutant removal process. In this context, air pollutants reduction and CO2 emissions have an antagonistic relationship, and the increase in CO2 emissions caused by this effect cannot be ignored.
In terms of control variables, the linear coefficient of GDP per capita is positive and significant at the 1%. level, and the quadratic coefficient is negative and significant at the 1% level. This result shows that the relationship between CO2 emissions and economic development of each city presents an inverted U shape. The calculated inflection point of the curve is 96,800 yuan (2000 constant price). According to this result, we find that in the study period, most cities in China belong to the left side of the curve, and only a few cities such as Shenzhen have reached the right side of the curve. The growth rate of CO2 emissions in cities on the left side of the curve is slowing down as GDP per capita increases. The results indicated that the decoupling of CO2 emissions from economic growth is expected to be achieved in the future. However, currently, China’s economic development still leads to an increase in CO2 emissions.
The regression coefficients of population density, the number of cars, and industrial structure are positive and significant at the 1% level. This indicates that an increase in population, the number of cars, and the proportion of the secondary industry leads to an increase in CO2 emissions. The regression coefficient of FDI is positive and significant at the 1% level, which indicates that FDI increased China’s CO2 emissions in the study period.
3.3 Spillover effects of the drivers
When analyzing the spatial spillover effects, this study mainly considers the direct and indirect effects based on equation 6. The estimated results are shown in Table 3.
Table 3. Decomposition estimates of the direct, indirect, and total effects.
|
Variable |
Direct |
Indirect |
Total |
|
|
0.009*** （5.04） |
0.001 （0.04） |
0.010 （0.63） |
|
|
0.010*** （5.68） |
0.095*** （6.58） |
0.105*** （6.85） |
|
|
0.152*** （10.33） |
0.449*** （4.44） |
0.601*** （5.93） |
|
|
-0.031*** （-8.03） |
-0.019 （-0.80） |
-0.050** （-2.01） |
|
|
0.041*** （3.22） |
0.160* （1.67） |
0.202** （1.99） |
|
|
-0.001 （-0.19） |
-0.040* （-1.89） |
-0.041* （-1.82） |
|
|
0.102*** （5.61） |
-0.540*** （-4.08） |
-0.438*** （-3.18） |
|
|
0.024*** （4.04） |
0.071* （1.75） |
0.095** （2.22） |
|
|
0.006*** （3.87） |
-0.020* （-1.73） |
-0.014 （-1.19） |
Note: ***, **, and * indicate significance levels of 1%, 5%, and 10%, respectively. The t values are in parentheses.
The direct effect of FRAP is positive and significant at the 1% level, while the indirect effect is not significant. This shows that FRAP mainly affects local CO2 emissions but has no significant impact on other regions. From a practical perspective, FRAP requires a relatively higher technical threshold, a longer cycle of technological progress, and a larger amount of investment. Therefore, although some developed cities have made progress in FRAP technology, their neighboring cities still have difficulties achieving the same technological progress. Simultaneously, FRAP technology can not only reduce corporate pollutant emissions but also increase corporate profits. Thus, its spillover effect between regions will be affected by stronger corporate self-protection and local protection, which further increases the difficulty of spillover.
Both direct and indirect effects of EPAP are positive and significant at the 1% level. This shows that an increase in the EPAP level of a city increases the local CO2 emissions and significantly increases the CO2 emissions of surrounding cities. The pollutant removal rate, to a certain extent, can reflect the air pollution control efforts in a region (Ren et al., 2018; Wu et al., 2020). The desulfurization rate in a region increases means that the pollutants reduction efforts also increase. Under this condition, some energy-intensive and pollution-intensive enterprises will move to the surrounding regions with less stringent air pollutants control (Yin et al., 2015; Zhao et al., 2020), which increase their CO2 emissions. The regression result of the industrial structure variable further illustrates this phenomenon. In terms of the proportion of the secondary industry, the coefficient of the direct effect is positive and significant at the 1% level, and that of the indirect effect is negative and significant at the 1% level. This indicates that the reduction of the proportion of the secondary industry reduces local CO2 emissions but increases CO2 emissions in surrounding cities. Some developed cities transfer high energy-consuming and high-polluting industries to the surrounding regions to meet increasingly stringent environmental goals, which puts considerable environment pressure on surrounding regions (Zhou et al., 2021a) and increases their CO2 emissions.
Additionally, the estimation results also show that control variables such as GDP per capita, population density, the number of cars, science and technology expenditure, and FDI directly and indirectly affect CO2 emissions in local and surrounding regions.
To test the robustness of the estimation results of the above model, this study constructs other K-nearest neighbor spatial weight matrices where k equals 8 and 12 and estimates the model. The results are shown as Table A1 and Table A2. The robustness test results show that after modifying the spatial weight matrix, the direction and significance of the regression coefficient of each variable do not change significantly and are consistent with the benchmark model. This indicates that the econometric model in this article is robust.
According to the results of Section 3, the air pollutants reduction in one city can affect CO2 emissions in surrounding cities. Is there an administrative boundary to this spillover effect? We will discuss this question in this section.
China’s existing mode of administrative management is dominated by the management of higher-level governments, and its administrative decentralization feature is evident (Cheng et al., 2021; Zhang et al., 2017). Therefore, cities in the same province have strong consistency in policy measures and development modes. Simultaneously, due to the coordination of provincial governments, communication between prefecture-level cities in the same province is frequent. Conversely, due to provincial administrative boundaries, prefecture-level cities in different provinces often have certain barriers in achieving consistency in policy formulation and communication. Sometimes, although two prefecture-level cities are near, there are significant differences in their modes of economic development and urban governance levels. In addition to the abovementioned regional barriers, the air pollutants reduction strategies of various countries have strong regional characteristics (Wang and Hao, 2012). The formulation and implementation of air pollutants reduction policies are highly consistent within the same region and have strong heterogeneity between different regions. This may have an impact on the spatial spillover effects of air pollutants and CO2 reduction in various regions. Thus, does the impact of a city’s air pollutants reduction on CO2 emissions in local and surrounding regions differ due to the effect of administrative boundaries? In this section we modify the traditional spatial weight matrix and construct new matrices for cities in same province and different provinces. Further, we adopt spatial panel model to analyze the differences between a city’s spatial spillover effects on cities in the same province and neighboring cities outside the province and then explore the effect of provincial administrative boundaries.
Specifically, we modify the spatial weight matrix of the same province based on the spatial adjacency matrix. The equation is as follows:
$${w}_{ij}=\left\{\begin{array}{c}1,i and j are cities in tℎe same province\\ 0,i and j are cities in different provinces\end{array}\right.$$
3
We modify the weight matrix of different provinces based on the K-nearest neighbor weight matrix. The k nearest cities are selected which in different provinces and assigned a weight of 1, and other cities are assigned a weight of 0. The value of k is also 10. The direct, indirect, and total effects of each variable calculated by Eq. 6 are shown in Table 4.
|
Variable |
\({W}_{same}\) |
\({W}_{different}\) |
||||
|---|---|---|---|---|---|---|
|
Direct |
Indirect |
Total |
Direct |
Indirect |
Total |
|
|
\(lnFront\) |
0.008*** (3.92) |
0.041*** (8.37) |
0.050*** (10.20) |
0.009*** (4.56) |
0.008 (1.37) |
0.017*** (2.96) |
|
\(lnEnd\) |
0.008*** (3.86) |
0.043*** (10.57) |
0.050*** (13.20) |
0.010*** (5.54) |
-0.007 (-1.23) |
0.003 (0.48) |
|
\(lnGDP\) |
0.146*** (7.95) |
0.290*** (11.03) |
0.436*** (18.03) |
0.219*** (14.23) |
0.106*** (-2.86) |
0.114*** (2.83) |
|
\({\left(lnGDP\right)}^{2}\) |
-0.030*** (-6.18) |
-0.078*** (-10.33) |
-0.108*** (-16.41) |
-0.046*** (-10.69) |
-0.003 (-0.33) |
-0.049*** (-5.09) |
|
\(lnPOP\) |
0.037** (2.42) |
0.247*** (7.80) |
0.285*** (9.03) |
0.056*** (4.05) |
0.192*** (4.32) |
0.248*** (5.41) |
|
\(lnTEC\) |
-0.001 (-0.35) |
-0.015** (-2.32) |
-0.017*** (-3.08) |
-0.002 (-0.54) |
-0.006 (-0.60) |
-0.007 (-0.76) |
|
\(lnSEC\) |
0.056** (2.23) |
-0.023 (-0.63) |
0.033 (1.12) |
0.061*** (2.93) |
0.015*** (0.29) |
0.076 (1.34) |
|
\(lnCAR\) |
0.017** (2.33) |
0.046*** (4.10) |
0.063*** (6.12) |
0.046*** (7.51) |
0.043** (2.40) |
0.089*** (4.62) |
|
\(lnFDI\) |
0.004** (2.01) |
-0.008*** (-2.79) |
-0.004 (-1.64) |
0.001 (0.73) |
-0.038*** (-7.96) |
-0.037*** (-7.29) |
| Note: ***, **, and * indicate significance levels of 1%, 5%, and 10%, respectively. The t values are in parentheses. | ||||||
The results in Table 6 show that in the weight matrix model of the same province (\({W}_{same}\)), the direct and indirect effects of FRAP and EPAP are positive and significant at the 1% level. This indicates that the impact of FRAP and EPAP on CO2 emissions in the same province has a positive and significant spillover effect—an increase in the level of FRAP in one city will lead to a decrease in CO2 emissions in other cities in the same province. While an increase in the level of EPAP in one city will lead to an increase in CO2 emissions in other cities in the same province. In the weight matrix model of different provinces (\({W}_{different}\)), the indirect effects of FRAP and EPAP are not significant. This result shows that the increase in the strength of FRAP and EPAP in a city does not have a significant impact on the neighboring cities outside the province.
The environmental and economic policies of cities in the same province are highly consistent, while the policies of cities in different provinces are considerably different. Based on the analysis of existing policy documents of various prefecture-level cities in China, we find that under the unified regulation of the provincial government, the air pollutants reduction policies and goals of cities in the same province are relatively consistent, and the emission reduction strategies and measures are also highly similar. Therefore, the results between cities in the same province are partially caused by top-down management. For different provinces, due to the existence of inter-provincial barriers, each province usually aims to maximize its own interests when formulating policies and always based on its own level of economic development, industrial structure, technical level, etc., which leads to relatively large heterogeneity in their air pollutants prevention and control strategies. Under this management model, cities in the same province have significant spatial spillover effects due to the consistency of policy objectives and governance measures. Although cities in different provinces are close in geographical distance, due to the differences in policies and measures, we cannot observe a significant spatial spillover effect across provinces in the model.
On the other hand, inter-provincial barriers hinder the diffusion of information, technology, and other factors (Poncet, 2003). The existence of inter-provincial barriers increases the cost of an enterprise to enter the market in a different province. Simultaneously, the barriers, to a certain extent, hinder the flow of technical factors such as information and knowledge between provinces. Further, the market segmentation caused by inter-provincial barriers may lead to restrictions on the entry of factors such as capital from other regions and a monopoly on the supply of factors (Su et al., 2018), thereby restricting the communication between provinces. With respect to this study, when a city’s emission reduction technology improves, the provincial government will extend this technological progress to other cities in the same province. Owing to the existence of inter-provincial barriers, it is difficult for cities in other provinces to benefit from this progress. Therefore, for FRAP and EPAP technology, there is a significant spatial spillover effect between cities in the same province, thereby affecting CO2 emissions. On the contrary, there is no significant spatial spillover effect between cities in different provinces in adjacent regions.
This study divided air pollutants reduction into two stages—FRAP and EPAP—and constructed a spatial panel model based on 240 cities above the prefecture level in China from 2005 to 2016 to explore the impact of different stages of air pollutants reduction on CO2 emissions and their spatial spillover effects. Further, it focused on the impact of provincial administrative boundaries on spatial spillover effects. The main conclusions are as follows:
- The CO2 emissions of the cities have a significant positive spatial correlation. The higher the CO2 emissions of one city, the higher the CO2 emissions of its neighboring cities. There exists a cluster phenomenon in the spatial distribution of FRAP and EPAP in various cities in China. An increase in the air pollutants reduction in one region often drives the air pollutants reduction in surrounding regions.
- The local and spatial spillover effects of FRAP and EPAP on CO2 emissions show obviously heterogeneity. FRAP affects CO2 emissions mainly through the local synergistic effect, and its spatial spillover effect is not significant. The local effect of EPAP on CO2 emissions is antergic, and the spatial spillover effect is significant. The increase of a city’s EPAP will increase the CO2 emissions in surrounding regions.
- Provincial boundaries weaken the spatial spillover effects of FRAP and EPAP in prefecture-level cities. The increase of FRAP in one region will reduce the CO2 emissions of other cities in the same province, and the increase of EPAP will increase the CO2 emissions of other cities in the same province. However, the spillover effects of the two do not exist for the neighboring cities outside the province.
Based on our conclusions, we propose the following policy recommendations for synergistically reducing air pollutants and CO2 at the regional level.
First, CO2 control should be incorporated into the joint regional air pollution prevention and control mechanism. Currently, the air pollution prevention and control mechanism in major regions of various countries is relatively mature and has achieved remarkable results. In the future, it is recommended to include CO2 control in the existing air pollutants reduction framework. For example, in the assessment of pollution control, CO2 emission reduction efficiency should be included in assessment indicators. Besides, assessments of air pollutants reduction should take both the pollutant removal rate and the inefficiency of increasing CO2 emissions in the removal process into consideration. Moreover, policies should be formulated to control the transfer of energy-intensive and pollution-intensive industries between regions and reduce the transfer of pollution and CO2 emissions caused by differences in governance.
Second, improve the efficiency and contribution of FRAP. Currently, EPAP is still the main measure for reducing air pollutants in developing countries. These countries should continue to strengthen emission reduction by improving the industrial and energy structure and reducing the production of pollutants and CO2. Additionally, increasing expenditure on key energy and production technologies is required. Incentives should be provided to enterprises with high energy utilization efficiency to encourage technological research and innovation and reduce the production of pollutants and CO2 through technological progress.
Third, the crafts of EPAP should be further optimized. Although EPAP can increase CO2 emissions and has limited contribution to air pollutants reduction, it still plays an important role in the short term in developing countries. It is suggested to increase the development of pollutant removal crafts to realize the replacement of the original carbon-producing removal craft with low carbon removal craft. At the same time, the developing of high energy efficiency removal equipment is also important for the purpose of reducing the power consumption in the process of pollutant removal, so as to reduce the CO2 generated by energy consumption.
Acknowledgments
This work was supported by funding from the National Natural Science Foundation of China (No.42171277) and the Ministry of Science and Technology of the People’s Republic of China (No.2021YFC3101704).
Author contribution
H Zhou collected data and drafted the manuscript. MD Jiang participated in the analysis and interpretation of the data. H Zhou, MD Jiang, YM Huang, Y Bai revised this manuscript. Q Wang designed the concept of the work and revised the manuscript. All authors read and approved the final manuscript.
Data availability
No supplementary information.
Ethics approval
Not applicable.
Consent to participate
Not applicable.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
- Baek J (2016) A new look at the FDI-income-energy-environment nexus: Dynamic panel data analysis of ASEAN. Energy Policy 91:22–27
- Chen CH, Wang BY, Fu QY, Green C, Streets DG (2006) Reductions in emissions of local air pollutants and co-benefits of Chinese energy policy: a Shanghai case study. Energy Policy 34:754–762
- Chen JD, Gao M, Cheng SL, Hou WX, Song ML, Liu X, Liu Y, Shan YL (2020) County-level CO2 emissions and sequestration in China during 1997–2017. Sci Data 7:391
- Chen Z, Kahn ME, Liu Y, Wang Z (2018) The consequences of spatially differentiated water pollution regulation in China. J Environemtal Econ Manage 88:468–485
- Cheng Y, Awan U, Ahmad S, Tan ZX (2021) How do technological innovation and fiscal decentralization affect the environment? A story of the fourth industrial revolution and sustainable growth. Technol Forecast Social Change 162:120398
- Dong HJ, Dai HC, Dong L, Fujita T, Geng Y, Klimont Z, Inoue T, Bunya S, Fujii M, Masui T (2015) Pursuing air pollutant co-benefits of CO2 mitigation in China: A provincial leveled analysis. Appl Energy 144:165–174
- Elhorst JP (2012) Dynamic spatial panels: models, methods, and inferences. J Geogr Syst 14:5–28
- Feng T, Du HB, Lin ZG, Zuo J (2020) Spatial spillover effects of environmental regulations on air pollution: Evidence from urban agglomerations in China. J Environ Manage 272:110998
- Hasselmann K, Latif M, Hooss G, Azar C, Edenhofer O, Jaeger CC, Johannessen OM, Kemfert C, Welp M, Wokaun A (2003) The challenge of long-term climate change. Science 302:1923–1925
- Huang RJ, Zhang YL, Bozzetti C, Ho KF, Cao JJ, Han YM, Daellenbach KR, Slowik JG, Platt SM, Canonaco F, Zotter P, Wolf R, Pieber SM, Bruns EA, Crippa M, Ciarelli G, Piazzalunga A, Schwikowski M, Abbaszade G, Schnelle-Kreis J, Zimmermann R, An ZS, Szidat S, Baltensperger U, Haddad E, Prevot I, A.S.H (2014) High secondary aerosol contribution to particulate pollution during haze events in China. Nature 514:218–222
- Jiang XJ, Hong CP, Zheng YX, Zheng B, Guan DB, Gouldson A, Zhang Q, He KB (2015) To what extent can China's near-term air pollution control policy protect air quality and human health? A case study of the Pearl River Delta region. Environ Res Lett 10:104006
- Lelieveld J, Evans JS, Fnais M, Giannadaki D, Pozzer A (2015) The contribution of outdoor air pollution sources to premature mortality on a global scale. Nature 525:367–
- LeSage JP, Pace RK (2009) Introduction to spatial econometrics. CRC Press, Boca Raton
- Liu F, Klimont Z, Zhang Q, Cofala J, Zhao LJ, Huo H, Nguyen B, Schopp W, Sander R, Zheng B, Hong CP, He KB, Amann M, Heyes C (2013) Integrating mitigation of air pollutants and greenhouse gases in Chinese cities: development of GAINS-City model for Beijing. J Clean Prod 58:25–33
- Lu Z, Zhang Q, Streets DG (2011) Sulfur dioxide and primary carbonaceous aerosol emissions in China and India, 1996–2010. Atmos Chem Phys 11:9839–9864
- Mao X, Xing Y, Hu T, Zeng A, Liu S (2012) An environmental-economic analysis of carbon,sulfur and nitrogen co-reduction path for Chinas power industry. China Environ Sci 32:748–756
- Meng B, Wang JG, Andrew R, Xiao H, Xue JJ, Peters GP (2017) Spatial spillover effects in determining China's regional CO2 emissions growth: 2007–2010. Energy Econ 63:161–173
- Ministry of Ecology and Environment of the People’s Republic of China, 2020.Steel Industry Environmental Assessment Report
- Nam KM, Waugh CJ, Paltsev S, Reilly JM, Karplus VJ (2014) Synergy between pollution and carbon emissions control: Comparing China and the United States. Energy Econ 46:186–201
- Peng W, Yang JN, Wagner F, Mauzerall DL (2017) Substantial air quality and climate co-benefits achievable now with sectoral mitigation strategies in China. Sci Total Environ 598:1076–1084
- Poncet S (2003) Measuring Chinese domestic and international integration. China Econ Rev 14:1–21
- Porter ME, Vanderlinde C (1995) Toward a New Conception of the Environment-Competitiveness Relationship. J Economic Perspect 9:97–118
- Qian HQ, Xu SD, Cao J, Ren FZ, Wei WD, Meng J, Wu LB (2021) Air pollution reduction and climate co-benefits in China's industries. Nat Sustain 4:417–
- Ren SG, Li XL, Yuan BL, Li DY, Chen XH (2018) The effects of three types of environmental regulation on eco-efficiency: A cross-region analysis in China. J Clean Prod 173:245–255
- Ren SG, Yuan BL, Ma X, Chen XH (2014) International trade, FDI (foreign direct investment) and embodied CO2 emissions: A case study of Chinas industrial sectors. China Econ Rev 28:123–134
- Shi WX, Lin C, Chen W, Hong JL, Chang JC, Dong Y, Zhang YL (2017) Environmental effect of current desulfurization technology on fly dust emission in China. Renew Sustain Energy Rev 72:1–9
- Song CB, Wu L, Xie YC, He JJ, Chen X, Wang T, Lin YC, Jin TS, Wang AX, Liu Y, Dai QL, Liu BS, Wang YN, Mao HJ (2017) Air pollution in China: Status and spatiotemporal variations. Environ Pollut 227:334–347
- Su YQ, Tesfazion P, Zhao Z (2018) Where are the migrants from? Inter- vs. intra-provincial rural-urban migration in China. China Econ Rev 47:142–155
- Tan QL, Wen ZG, Chen JN (2016) Goal and technology path of CO2 mitigation in China's cement industry: from the perspective of co-benefit. J Clean Prod 114:299–313
- Wang SX, Hao JM (2012) Air quality management in China: Issues, challenges, and options. J Environ Sci 24:2–13
- Wang T, Xue LK, Brimblecombe P, Lam YF, Li L, Zhang L (2017) Ozone pollution in China: A review of concentrations, meteorological influences, chemical precursors, and effects. Sci Total Environ 575:1582–1596
- Wu HT, Hao Y, Ren SY (2020) How do environmental regulation and environmental decentralization affect green total factor energy efficiency: Evidence from China. Energy Econ 91:104880
- Xie QC, Wang XY, Cong XP (2020) How does foreign direct investment affect CO2 emissions in emerging countries?New findings from a nonlinear panel analysis. J Clean Prod 249:119422
- Xu M, Qin ZF, Zhang SH (2021) Carbon dioxide mitigation co-effect analysis of clean air policies: lessons and perspectives in China's Beijing-Tianjin-Hebei region. Environmetal Res Lett 16:015006
- Yin JH, Zheng MZ, Chen J (2015) The effects of environmental regulation and technical progress on CO2 Kuznets curve: An evidence from China. Energy Policy 77:97–108
- You WH, Lv ZK (2018) Spillover effects of economic globalization on CO2 emissions: A spatial panel approach. Energy Econ 73:248–257
- Yue HB, He CY, Huang QX, Yin D, Bryan BA (2020) Stronger policy required to substantially reduce deaths from PM2.5 pollution in China. Nat Commun 11:1462
- Zhang CG, Zhou XX (2016) Does foreign direct investment lead to lower CO2 emissions? Evidence from a regional analysis in China. Renew Sustain Energy Rev 58:943–951
- Zhang K, Zhang ZY, Liang QM (2017) An empirical analysis of the green paradox in China: From the perspective of fiscal decentralization. Energy Policy 103:203–211
- Zhang XD, Chen LL, Yuan RS (2020) Effect of natural and anthropic factors on the spatiotemporal pattern of haze pollution control of China. J Clean Prod 251:119531
- Zhao XM, Liu CJ, Sun CW, Yang M (2020) Does stringent environmental regulation lead to a carbon haven effect? Evidence from carbon-intensive industries in China. Energy Econ 86:104631
- Zheng B, Tong D, Li M, Liu F, Hong CP, Geng GN, Li HY, Li X, Peng LQ, Qi J, Yan L, Zhang YX, Zhao HY, Zheng YX, He KB, Zhang Q (2018) Trends in China's anthropogenic emissions since 2010 as the consequence of clean air actions. Atmos Chem Phys 18:14095–14111
- Zhou H, Jiang MD, Huang YM, Wang Q(2021a) Directional spatial spillover effects and driving factors of haze pollution in North China Plain. Resources, Conservation and Recycling. 169, 105475
- Zhou Q, Zhong SH, Shi T, Zhang XL (2021b) Environmental regulation and haze pollution: Neighbor-companion or neighbor-beggar? Energy Policy 151:112183