Evaluation and Comparison of Air Pollution Governance Performance: An Empirical Study Based on Jiangxi Province

Abstract

This paper takes the air pollution governance performance as the research object, establishes the evaluation index system of air pollution governance performance using the pressure-state-response (PSR) model, and uses the data of 11 prefecture-level cities in Jiangxi Province from 2014–2017 to carry out empirical tests. The results show that, in terms of indicator weights, the state and pressure categories have higher weights than the response category, further highlighting the importance of reducing pollution emissions rather than post-pollution treatment. Regarding regional comparisons, only a few regions show a good balance between “stress-state-response”, while most regions show a “loss of balance”. In terms of annual changes, the performance of most regions in several categories rose and showed a wave-like upward trend, reflecting the intermittent improvement characteristics of air pollution governance performance in most regions of Jiangxi. Finally, combined with the evaluation results, this paper proposes policy suggestions, such as improving the performance evaluation index system of air pollution governance, promoting the comprehensive governance of air pollution, focusing on regions with weaker air pollution governance, and strengthening the regional collaborative governance of air pollution.

1. Introduction

Over the past 40 years of reform and opening up, China has lived up to expectations and developed as the world’s second-largest economy. However, increasing pressure on resources and environmental protection, as well as social trends, such as the urgent need to transform the market to an efficient, clean, and sustainable one [1,2], have forced the Chinese government to rethink how to deal with the challenges of air pollution management. As early as the National Conference on Ecological Protection of the Environment in 2018, General Secretary Xi Jinping emphasized the need to address outstanding ecological and environmental issues as a priority area of people’s livelihood. He stressed that the top priority is resolutely winning the battle for blue skies. Moreover, both the Three-Year Action Plan to Win the Blue Sky Defense War released by the State Council in 2018 and the government work reports of the two national conferences in recent years have demonstrated China’s determination to improve air quality and keep more blue skies continuously.

In recent years, China’s air pollution governance practices have been gradually adjusted and improved as the air pollution problem has changed. In general, the focus of air pollution governance has shifted from individual problems, such as industrial emissions and smoke (dust), to comprehensive and regional problems. The means of governance have shifted from reliance on administrative and legal enforcement-type policy tools to the integrated use of various market-based policy tools. Out of the consideration of responding to the practical needs and ensuring real results in the battle of pollution prevention and control, the performance assessment of the air environment governance work of governments at all levels has become the necessary point. The Law of the People’s Republic of China on the Prevention and Control of Air Pollution specifies the responsibility of provincial governments to conduct assessments and make public the final results of air pollution governance in areas under their jurisdiction. The Action Plan for the Prevention and Control of Air Pollution also includes PM and other binding indicators in the economic and social development assessment. In this context, as a visual display and important feedback on the effectiveness of governance, air pollution governance performance evaluation has increasingly become a hot topic of research by experts and scholars.

There has been much beneficial discussion of air pollution governance performance in the literature. Specifically, the discussion focused on the following three areas. First, from a multidisciplinary perspective, scholars have used Data Envelopment Analysis (DEA) models, multiple linear regression (MLR), fuzzy ranking method, life cycle assessment (LCA) method, and other mathematical and theoretical models to empirically study the air environment state [3,4,5,6]. Second, for considering air mobility and other characteristics, some scholars assess the performance of collaborative air pollution governance among local governments [7,8,9]. Third, some scholars have drawn on the PSR system dynamics model to construct an air environment PSR and conduct a theoretical exploration of the evaluation index system [10,11]. Most of these studies focus on the national or provincial level, while few relevant studies are mentioned at the city level. Moreover, the study samples are mostly from the key regions of air pollution prevention and control, such as the Beijing–Tianjin–Hebei Region, the Yangtze River Delta, the Yellow River Economic Belt, and the Guangdong–Hong Kong–Macao Greater Bay Area, while less attention is paid to the central provinces. This two-tiered, unbalanced pattern is not conducive to the overall promotion of national air pollution management [12].

As a large province in the central region, Jiangxi Province continues to deepen the creation of ecological civilization demonstration work and has achieved significant results in environmental governance. The Ministry of Ecology and Environment organized the selection of the “Lucid waters and lush mountains are invaluable assets” practice innovation base, the number of counties (cities) in Jiangxi Province is the highest in China. The data show that from January to August 2022, seven of the top 10 counties (cities and districts) in Jiangxi province ranked in PM2.5 were ecological demonstration creation areas. In 2021, the proportion of good days in the province reached 96.1%, ranking sixth in the country. Therefore, it is not only of great practical significance to dig deeper into the air pollution governance situation in Jiangxi Province, promote its beneficial experience, and help build a national typical model, but also an important supplement to improve the diversity of existing environmental governance research samples. This paper establishes the air pollution control performance evaluation index system with the help of the PSR model. It uses the entropy value method to conduct a comparative empirical analysis of the air pollution control performance of 11 prefecture-level cities in Jiangxi Province from 2014 to 2017. The purpose is to objectively present the effect of air pollution governance and provide reference to help air pollution governance and the ecological environment develop for the better.

2. Literature Review

International research on air pollution problems has been proliferating for decades. Due to a hundred years of industrialization, the developed countries in the West have an early start in air pollution management and have accumulated rich and valuable experience in pollution governance. Their research topics range from macro to micro, from single to multiple, and they constantly expand new research perspectives [13]. With the growing air pollution problem in developing countries and their active responsibility to reduce emissions, many scholars have started focusing on air pollution governance research in China. Following the practice of Gayialis et al. [1], we screened and analyzed the literature on “air pollution governance” in the Web of Science (WOS) core collection database and the China Knowledge Network (CNKI) database in the past five years. The relevant studies can be broadly classified into the following three categories.

The first type of research focuses on the causes of air pollution and its influencing factors and has shifted from a single perspective to an integrated, dynamic perspective. Earlier studies independently analyzed the impact of natural conditions, human production activities, and other factors on air quality and considered air pollution as a regional and enterprise problem. In contrast, later studies take natural and social factors into consideration at the same time, and the air pollution problem is considered more as a complex, multi-faceted problem that requires multi-subject and cross-regional collaborative governance [14,15]. Regarding research methods, scholars mostly use Geographic Information System (GIS), neural network analysis, backward trajectory models, exploratory spatial data analysis methods, multiple linear regression models, fixed effects panel models, spatial econometric models, etc. [16,17,18,19,20]. It was found empirically that weather conditions such as weaker winds, shorter sunshine hours, lower temperatures, and reduced precipitation are more likely to produce air pollution [21,22,23]. In addition, socio-economic factors [24,25] such as economic growth, industrial structure, urbanization level, energy efficiency, traffic pressure, urban greening, and construction, as well as political factors [26,27,28], such as decentralization system and inter-governmental cooperation, are also important drivers of changes in air quality.

The second research category focuses on governance tools and examines air pollution governance technologies and policies [29,30]. Liu et al. (2017) studied different air purification technologies and air pollution control and published several articles [31]. Khezami et al. (2021) focused on the hybrid treatment systems of air pollution control [32]. Phua et al. (2019) promote the further development of waste incinerator residue treatment processes [33]. Such research is based on a multidisciplinary intersectional perspective. It continues to deepen and expand the development and application of new technologies for air pollution control, but there is a relative lack of research from the humanities and social sciences fields. On the other hand, Gong (2018) summarized China’s development characteristics of air pollution protection policies based on a policy network theory perspective [34]. Urrutia-Goyes et al. found that traffic control during peak hours in urban centers ameliorated air pollution [35]. Yang et al. conducted a content analysis of air pollution governance regulations and policies since the founding of China [15]. Such studies have produced more fruitful results by analyzing different policy instruments from a policy perspective.

In addition, a few studies narrow the perspective to the study of air pollution governance performance. International scholars have relatively mature theoretical standards on the definition and content of environmental performance assessment. Therefore, the operable PSR model is mostly used in constructing the evaluation index system [36]. In contrast, Chinese scholars have started their research on air pollution governance performance relatively late. By 31 October 2022, a total of 69 articles were searched on the CNKI using the subject terms of “air pollution” and “governance performance”, including 35 journal articles, of which 22 were core journals and above, and 14 were CSSCI journal articles. Scholars have mostly used DEA and other empirical methods to verify the effectiveness of regional governance [37], and a mature and perfect evaluation index system has still not been established. Such studies focus their research and attention on the results, such as whether and to what extent the air quality is improved, while the research on the treatment process is obviously insufficient. Some other scholars follow the PSR model of the environmental governance evaluation system and try to build up a performance evaluation index system covering the whole process and all elements of air pollution governance [38]. However, this part of the research mostly stops at the level of theoretical research and lacks relevant empirical support, which is weak in applicability and operability. In the existing empirical literature, most of the samples were selected at the national or provincial level [27,39], and the studies focused on key regions and cities [40,41] such as the Beijing–Tianjin–Hebei Region, the Yangtze River Delta, and the Guangdong–Hong Kong–Macao Greater Bay Area, with less attention to the central and western regions. Even fewer studies delve into the performance of air pollution governance in prefecture-level cities.

Through the literature, we found that the existing papers have rich and useful discussions on air pollution governance, and more results have been achieved in impact factors, governance technologies, and policy approaches, which laid the foundation for the study of our study. However, the domestic quantitative evaluation of air pollution governance performance is still lagging behind, and academics commonly use the traditional PSR model for environmental governance performance evaluation in selecting evaluation methods. Few studies in this part of the literature have evaluated air pollution governance performance in central prefecture-level cities. Based on this, this paper tries to make improvements and additions in the following aspects. On the one hand, in the research method, the entropy value method is combined with the PSR model to build up a whole-factor and whole-process air pollution governance performance evaluation system. The method not only reflects the key process and parameters of air environment governance but also covers the process and result indicators of air governance. On the other hand, the study uses panel data from 11 cities in Jiangxi Province from 2014–2017 for the empirical study. As a model of successful air pollution management in China, Jiangxi Province has been recognized by the national environmental protection authorities for its achievements, and an in-depth analysis of its underlying mechanisms and mechanisms is useful to promote and learn from its management experience.

3. Theoretical Model and Index System

3.1. PSR Model

Since the 1970s, the Organization for Economic Co-operation and Development (OECD) has used the modified PSR model in environmental reports. Subsequently, the PSR model has been frequently used in other environmental assessment modules such as sustainable use of environmental resources and environmental systems assessment. The PSR model framework is shown in Figure 1. Stress, state, and response are linked and interact with each other. They interact with each other so that people’s socio-economic life inevitably puts pressure on the environment. If people increase the efficiency of resource use, reduce the output rate of environmental pollutants and strengthen the investment in environmental pollution governance, environmental quality will continue to improve. On the contrary, if people make the resource use efficiency decrease, the output rate of environmental pollutants increases, and the investment in environmental pollution governance weakens, the environmental quality will continue to decline, as shown in Figure 2.

3.2. Indicator System Based on PSR Model

3.2.1. Indicator System

The evaluation index system of Jiangxi air pollution governance performance aims to present the actual situation and development trend of Jiangxi air environment governance as a whole and objectively. Therefore, the below principles should be followed. First is the principle of systemic. This principle requires the index system to reflect the current situation and level of air pollution governance in Jiangxi in a comprehensive, accurate and multi-angle manner. Second is the principle of objectivity. Both theoretical research and scientific methods should support the selection of indicators to measure them so that they can accurately reflect the reality of the air environment and thus objectively reflect the current situation of air pollution governance. The third is the principle of operability. On the one hand, the index data can be obtained from bulletins, reports, annual reports, statistical yearbooks, etc., published by national or local related departments. On the other hand, the index data can be quantified and taken in a consistent statistical caliber. Fourth is the principle of representativeness. The data of indicators that have a more significant impact on the governance performance of air pollution are representative and closely integrated with the management strategy of Jiangxi Province are selected.

In this paper, according to the above basic principles, the fundamentals of the PSR model, and relevant policy documents, an evaluation index system for the governance performance of air pollution is established. Among them, pressure-based indicators should reflect the pressure generated by pollutant emissions on the air environment. Currently, the primary pollutants in China’s air environment are sulfur dioxide (SO₂), nitrogen oxides (NO_X), smoke (dust) and industrial waste gas. As a result, the pressure-type index is calculated by using the emissions of SO₂, NO_X, smoke (dust), and industrial waste gas per unit of GDP (SO₂, NO_X, and smoke (dust) emissions refer to the emissions from industry and urban life). These are inverse indicators, and the smaller the value, the better. Specifically, the smaller the value of the indicator, the lower the pressure on the air due to emissions in the process of economic development, and the fewer pollutants need to be produced per unit of GDP increase. The state improvement index is the central monitoring index to judge whether the air environment has improved since the government conducted air pollution governance. The specific indicators were selected as “Increase in the proportion of days with good or better air quality,” “Reduction in nitrogen dioxide concentration,” “Reduction in sulfur dioxide concentration,” and “Reduction of PM10 per unit space.” This indicator contains the difference in pollutants per unit of air compared to the previous period, which is a positive indicator. In other words, the higher the value of the indicator, the better the degree of air quality improvement. The reaction type indicator is calculated using the air pollution control investment in each region of Jiangxi Province. Some scholars believe that the number of exhaust gas treatment facilities represents the fixed asset input of air pollution governance, and the operating cost of exhaust gas treatment represents the mobile factor input of air pollution governance, so they are able to use the values of these two types of indicators to calculate the air pollution treatment input [42]. Generally speaking, the larger the indicator, the higher the degree of attention paid to air pollution governance and the stronger the action to combat air pollution in each region of Jiangxi Province. By analyzing the logical relationship between the three elements of pressure, state, and response, and based on the research of scholars such as Li [12], this paper finally established the evaluation index system of air pollution governance performance in Jiangxi Province, which contains 10 indicators (see

Table 1. Evaluation index system of air pollution governance performance in Jiangxi.

Type of Indicators	Symbols	Content of Indicators	Unit	Calculation Method	Description of Indicators	Nature of Indicators
Pressure Performance	P1	Sulfur dioxide (SO2) emissions per unit of GDP	Ton/Million Yuan	SO2 emissions/GDP	SO2 emissions from the creation of one unit of GDP, indicating the degree of pressure	Reverse indicator, the lower, the better
	P2	Smoke (dust) emissions per unit of GDP	Ton/Million Yuan	Smoke (dust) emissions/GDP	The amount of smoke (dust) emissions generated to create one unit of GDP, indicating the degree of pressure	Reverse indicator, the lower, the better
	P3	Industrial waste gas emissions per unit GDP	10 million cubic meters/million RMB	Industrial waste gas emissions/GDP	The industrial emissions generated for the creation of one unit of GDP, indicating the degree of pressure	Reverse indicator, the lower, the better
	P4	Nitrogen oxide (NOX) emissions per unit of GDP	Billion cubic meters/million yuan	NOX emissions/GDP	NOX emissions from the creation of one unit of GDP, indicating the degree of pressure	Reverse indicator, the lower, the better
State Performance	S1	Increase in the proportion of days with good or better air quality	%	The proportion of days with good or better air quality in the current year-the proportion of days with good or better air quality in the previous year	The degree of increase in the proportion of days with good or better air quality, indicating the improvement of the air environment	Positive indicators, the higher, the better
	S2	Reduction in nitrogen dioxide (NO2) concentration	µg/m3	NO2 concentration in the air in the previous year—NO2 concentration in the current year	The degree of reduction in NO2 concentration, indicating improvement in the air environment	Positive indicators, the higher, the better
	S3	Reduction in sulfur dioxide (SO2) concentration	µg/m3	SO2 concentration in the previous year—SO2 concentration in the current year	The degree of reduction in SO2 concentration, indicating improvement in the air environment	Positive indicators, the higher, the better
	S4	Reduction of PM10 per unit space	µg/m3	Previous year PM10 per unit space—Current year PM10 per unit space	The degree of reduction of PM10 in the year, indicating the improvement of the air environment	Positive indicators, the higher, the better
Reaction Performance	R1	Number of exhaust gas treatment facilities	Hundreds of sets	Number of exhaust gas treatment facilities invested by each entity in the current year	Number of exhaust gas treatment facilities, indicating the response to air pollution	Positive indicators, the higher, the better
	R2	Operating costs of exhaust gas treatment facilities	10 million yuan	Total operating costs of exhaust gas treatment facilities invested in the current year	Operating costs of exhaust gas treatment facilities, indicating the response to air pollution	Positive indicators, the higher, the better

Source: Based on the research of Li and other scholars compiled by the authors.

for details).

3.2.2. Data Source

This paper selects panel data of each prefecture-level city in Jiangxi Province as the research object to be measured. In order to ensure the authority and accuracy of the data, the original data were obtained from the Jiangxi Statistical Yearbook, the Jiangxi Environmental Statistics Annual Report, the Jiangxi Environmental Statistics Bulletin, the “Environmental Quality Monthly Report” in the Jiangxi Department of Ecology and Environment, and the “Atmospheric Environment of Jiangxi Province” in the EPS Database, etc. Since promulgated the “Ambient Air Quality Standards” in 2012, Jiangxi Province has promoted the construction of 180 automatic ambient air monitoring stations covering all cities, counties, and districts. At the same time, taking into account the national model to monitor the air authority, entrusted to a third-party company responsible for the operation and maintenance to ensure the accuracy and stability of monitoring data. In order to compare the performance of air pollution governance in different regions and years, a total of 44 sets of data (ten indicators for each set of data) from 2014 to 2017 in 11 prefecture-level cities of Jiangxi Province were selected for comparative analysis.

4. Performance Evaluation of Air Pollution Governance: An Empirical Analysis of Jiangxi Province

4.1. Entropy Method

The entropy method is an objective weighting method based on the theory that the weights are judged based on the level of differences between all evaluation index data. And the original data are normalized using the extreme value processing method. The former is superior to the standardized advection method’s standardization of the raw data. Therefore, when the entropy value method is applied, standardization of the initial data by the extreme value method is the optimal choice [43]. Therefore, in the following, we adopt a combination of both the extreme value method standardization and the entropy method to evaluate the performance index system of air pollution governance.

4.1.1. Construction of the Original Matrix

In the original matrix, X_ij is the value of the jth indicator in the city i.

4.1.2. Standardization of Index Data

This paper selects ten indicators for Jiangxi air pollution governance, whose economic significance, role, and magnitude vary greatly. Therefore, it is imperative to adopt dimensionless processing for these indicators. In order to facilitate the calculation of the entropy method, the data of the selected ten indicators need to be pre-processed. More specifically, the P-indicator is taken as the reciprocal, and the data is polarized so that all the indicators have the same force and eliminate the dimensionality difference.

In order to eliminate the influence of the index outline on the evaluation results, to make it obey the normal distribution with the mean value of 0 and variance of 1, so that the indicators are comparable with each other, the original data need to be standardized. The standardization formula is as follows.

$$X^{\prime }{}_{ij}=\frac{X_j-X_{min}}{X_{max}-X_{min}}$$

where i indicates the year, X_j represents the value of the j_th index, X_max represents the maximum value of the jth index. X_min represents the minimum value of the jth index, $X^{\prime }{}_{ij}$ indicates the standardized value.

4.1.3. Metrics Homogenization

Calculate the characteristic weight $P_{ij}$ of the jth indicator.

$$P_{ij}=\frac{X^{\prime }{}_{ij}}{{{\displaystyle \sum }}_{i=1}^n{X}^{\prime }{}_{ij}}$$

4.1.4. Calculate the Entropy Value

Calculate the entropy value of the jth indicator.

$$e_j=-k{{\displaystyle \sum }}_{i=1}^n{P}_{ij}\ln \left(P_{ij}\right)$$

where n represents the number of samples. If we set k = 1/ln(n), we will have 0 ≤ ej ≤ 1.

4.1.5. Calculate the Coefficient of Variability

The coefficient of variation g_i is calculated for indicator j. In general, the greater the variation between X’_ij, the smaller the entropy value e_j, indicating that the effect of indicator j is also more significant.

$$g_i=1-e_j$$

The importance of this indicator increases as the coefficient of variation g_i becomes larger.

4.1.6. Determine the Weights

Calculate the indicator weights of the jth item.

$$V_j=\frac{g_i}{{{\displaystyle \sum }}_{i=1}^n{g}_i}$$

4.1.7. Calculation of Air Pollution Governance Performance in Jiangxi

The performance of air pollution governance in each city of Jiangxi is comprehensively derived.

$$E_j={{\displaystyle \sum }}^{}{V}_j{X}^{\prime }{}_{ij}$$

E_j indicates the comprehensive evaluation index of the sample. The larger the index, the higher the city’s air pollution governance performance in the province.

4.2. Weighting Results of the Indicator System

This paper uses the entropy value method to calculate the indicators weights of the evaluation system of the air pollution governance performance, as detailed in

Table 2. Evaluation indicators weights of the air pollution governance performance in Jiangxi.

Type of Indicators	Symbols	Indicators	Weight	General
Pressure Performance	P1	Sulfur dioxide emissions per unit of GDP	0.0632	0.2801
	P2	Smoke (dust) emissions per unit of GDP	0.0613
	P3	Industrial waste gas emissions per unit GDP	0.0362
	P4	Nitrogen oxide emissions per unit of GDP	0.1194
State Performance	S1	Increase in the proportion of days with good or better air quality	0.1394	0.5257
	S2	Reduction in nitrogen dioxide concentration	0.0404
	S3	Reduction in sulfur dioxide concentration	0.0243
	S4	Reduction of PM10 per unit space	0.3216
Reaction Performance	R1	Number of exhaust gas treatment facilities	0.0369	0.1942
	R2	Operating costs of exhaust gas treatment facilities	0.1573

Data source: Compiled by the author, indicator weights are retained to 4 decimal places.

.

Table 2 shows that in terms of indicator types, the final order of weighting obtained by using the entropy method is as follows. First is the state indicator, followed by the pressure and reaction indicators. The contribution of reaction indicators to air pollution governance performance is lower than pressure and state indicators. The calculation results show that the “NOx emission per unit GDP” indicator contributes the highest to air pollution governance performance among the pressure indicators. The weights of the “SO₂ emissions per unit of GDP” indicator and the “Smoke (dust) emissions per unit of GDP” indicator are equal. The weight of the “Industrial waste gas emissions per unit GDP” indicator is the lowest. Among the state indicators, the “Reduction of PM10 per unit space” indicator accounts for the highest percentage, followed by the “Increase in the proportion of days with good or better air quality” indicator, and then the “Reduction in NO₂ concentration” indicator and “Reduction in SO₂ concentration” indicator. Among the response indicators, the “Operating costs of exhaust gas treatment facilities” indicator contributes more to the governance performance of air pollution than the “Number of exhaust gas treatment facilities” indicator. According to the results of the ten indicators, it can be found that the contribution to the governance performance of air pollution is ranked as follows. The “Reduction of PM10 per unit space” indicator is the highest reduction indicator compared with last year. “NO_X emissions per unit of GDP,” “Increase in the proportion of days with good or better air quality,” and “Operating costs of exhaust gas treatment facilities” indicators are the next highest. “Reduction in SO₂ concentration,” “Smoke (dust) emissions per unit of GDP,” and “Reduction in NO₂ concentration” indicators are the second highest. “Industrial waste gas emissions per unit GDP,” “SO₂ emissions per unit of GDP,” and “Number of exhaust gas treatment facilities” indicators are the lowest. It can be seen that to improve the environment and obtain higher air pollution governance performance and more attention must be paid to controlling pollution sources and reducing pollution emissions rather than post-pollution governance.

4.3. A Comprehensive Evaluation Results of Air Pollution Governance Performance

The entropy method measured the air pollution governance performance of 11 prefecture-level cities in Jiangxi Province and their indicator performance from 2014 to 2017, and the results are detailed in

Table 3. Comprehensive results of air pollution governance performance in Jiangxi (2014–2017).

Region	Type of Indicators	Year				Average
		2014	2015	2016	2017
Nanchang City	Pressure Performance	0.1077	0.1224	0.1431	0.1587	0.1330
	State Performance	0.2287	0.1598	0.1050	0.1117	0.1513
	Reaction Performance	0.1144	0.1285	0.1234	0.1295	0.1240
	Comprehensive Performance	0.4508	0.4107	0.3715	0.3999	0.4082
Jingdezhen City	Pressure Performance	0.0724	0.0732	0.0854	0.0899	0.0802
	State Performance	0.0702	0.1342	0.1254	0.1165	0.1116
	Reaction Performance	0.0597	0.0659	0.0622	0.0836	0.0679
	Comprehensive Performance	0.2023	0.2733	0.2730	0.2900	0.2597
Pingxiang City	Pressure Performance	0.0895	0.0685	0.0741	0.0764	0.0771
	State Performance	0.1188	0.0341	0.1205	0.1290	0.1006
	Reaction Performance	0.0840	0.1001	0.0772	0.0918	0.0883
	Comprehensive Performance	0.2923	0.2027	0.2718	0.2972	0.2660
Jiujiang City	Pressure Performance	0.0701	0.0705	0.0800	0.0873	0.0770
	State Performance	0.0634	0.1473	0.1224	0.1207	0.1135
	Reaction Performance	0.1044	0.1225	0.1637	0.1231	0.1284
	Comprehensive Performance	0.2379	0.3403	0.3661	0.3311	0.3189
Xinyu City	Pressure Performance	0.0696	0.0704	0.0739	0.0751	0.0723
	State Performance	0.0291	0.2087	0.0951	0.0879	0.1052
	Reaction Performance	0.1421	0.1436	0.1403	0.1381	0.1410
	Comprehensive Performance	0.2408	0.4227	0.3093	0.3011	0.3185
Yingtan City	Pressure Performance	0.0770	0.0811	0.1289	0.1388	0.1065
	State Performance	0.1132	0.1170	0.1325	0.1098	0.1181
	Reaction Performance	0.1660	0.1617	0.1313	0.1176	0.1442
	Comprehensive Performance	0.3562	0.3598	0.3927	0.3662	0.3687
Ganzhou City	Pressure Performance	0.0774	0.0791	0.0820	0.0937	0.0831
	State Performance	0.0841	0.1601	0.0792	0.0959	0.1048
	Reaction Performance	0.0684	0.0793	0.0803	0.0942	0.0806
	Comprehensive Performance	0.2299	0.3185	0.2415	0.2838	0.2684
Ji’an City	Pressure Performance	0.0787	0.0786	0.0895	0.0979	0.0862
	State Performance	0.0904	0.1516	0.8450	0.0734	0.2901
	Reaction Performance	0.0690	0.0599	0.0699	0.1024	0.0753
	Comprehensive Performance	0.2381	0.2901	1.0044	0.2737	0.4516
Yichun City	Pressure Performance	0.0693	0.0694	0.0727	0.0740	0.0714
	State Performance	0.0634	0.1981	0.0355	0.1245	0.1054
	Reaction Performance	0.1215	0.1540	0.1393	0.1447	0.1399
	Comprehensive Performance	0.2542	0.4215	0.2475	0.3432	0.3166
Fuzhou City	Pressure Performance	0.1098	0.1083	0.1000	0.1116	0.1074
	State Performance	0.1262	0.1269	0.1122	0.0972	0.1156
	Reaction Performance	0.0559	0.0683	0.0516	0.0776	0.0634
	Comprehensive Performance	0.2919	0.3035	0.2638	0.2864	0.2864
Shangrao City	Pressure Performance	0.0764	0.0786	0.0852	0.0902	0.0826
	State Performance	0.1199	0.1212	0.1005	0.0899	0.1079
	Reaction Performance	0.0825	0.0889	0.0929	0.0924	0.0892
	Comprehensive Performance	0.2788	0.2887	0.2786	0.2725	0.2797
Jiangxi Province	Pressure Performance	0.0816	0.0818	0.0923	0.0994	0.0888
	State Performance	0.1007	0.1417	0.1703	0.1051	0.1295
	Reaction Performance	0.0971	0.1066	0.1029	0.1086	0.1038
	Comprehensive Performance	0.2794	0.3301	0.3655	0.3131	0.3220

Data source: Compiled by the author, all scores are retained to four decimal places.

. Where the Pressure Performance is the total performance of pressure indicators, the state performance is the total performance of state indicators, and the reaction performance is the total performance of reaction indicators. The pressure, state, and reaction performance scores of Jiangxi Province are the average of 11 prefecture-level cities in that year. The composite performance is the total performance score of pressure, state, and reaction. According to the calculation results in Table 2, we can find that the combined performance = 0.2801 $\times$ Pressure indicator + 0.5257 $\times$ State indicator + 0.1942 $\times$ Response indicator.

As can be seen from Table 3, the pressure performance and response performance in Jiangxi Province increased year by year, while the state performance and comprehensive performance increased from 2014 to 2016 and decreased in 2017. This shows that Jiangxi Province is more active in controlling pollution emissions and investing in air pollution governance, but the effect of improving air quality in 2017 is unsatisfactory.

4.4. Inter-Regional Comparison of Air Pollution Governance Performance

In

Table 4. Ranking of 4-year average values of performance by category in each region.

Region	2014–2017 Average Ranking (High–Low)
	Pressure Performance	State Performance	Reaction Performance	Comprehensive Performance
Nanchang City	1	2	5	2
Jingdezhen City	7	6	10	11
Pingxiang City	8	11	7	10
Jiujiang City	9	5	4	4
Xinyu City	10	9	2	5
Yingtan City	3	3	1	3
Ganzhou City	5	10	8	9
Ji’an City	4	1	9	1
Yichun City	11	8	3	6
Fuzhou City	2	4	11	7
Shangrao City	6	7	6	8

Data source: Compiled by the author.

, the four-year stress performance, state performance, response performance, and comprehensive performance averages of 11 prefecture-level cities in Jiangxi Province are ranked in the order of highest and lowest, and the rankings are presented separately. The top five cities in the comprehensive performance ranking are Ji’an, Nanchang, Yingtan, Jiujiang, and Xinyu. In contrast, the bottom three cities are Jingdezhen, Pingxiang and Ganzhou. In terms of pressure performance, the top five cities are Nanchang, Fuzhou, Yingtan, Ji’an, and Ganzhou, while the bottom three cities are Yichun, Xinyu, and Jiujiang. In addition, in terms of state performance ranking, Ji’an, Nanchang, Yingtan, Fuzhou, and Jiujiang rank in the top five, while Pingxiang, Ganzhou, and Xinyu rank in the bottom three with unsatisfactory performance. Finally, the top five cities in response performance are Yingtan, Xinyu, Yichun, Jiujiang, and Nanchang, while the bottom three cities are Fuzhou, Jingdezhen, and Ji’an.

Regarding the sub-analysis, the cities with higher performance rankings in the stress category emit fewer pollutants, with NOx emissions being their limiting indicator. The cities with the top performance ranking in the status category have a higher degree of improvement in air environment quality, and PM10 emission reduction is their limiting indicator. The regions with the top reaction performance rank have higher air pollution investment, among which the operation cost of waste gas treatment facilities is its limiting indicator.

In terms of individual cities, among the cities with the top comprehensive performance ranking, Ji’an ranks higher in both pressure and state performance but lower in response performance. This indicates that although this city is more active in air pollution pressure reduction and the final result of air quality improvement is good, the importance of investment in air pollution governance needs to be improved. Jiujiang ranks high in both the state performance and response performance but ranks low in the pressure performance. This indicates that Jiujiang’s air quality improvement effect is good, and the investment in air pollution governance is more focused but lacks some importance in air pollution pressure reduction. Although Xinyu pays enough attention to investment in air pollution governance, reducing air pollution pressure is not active enough, and improving air quality is not apparent. Nanchang and Yingtan, two cities, not only have an excellent comprehensive performance but also rank higher in pressure performance, state performance, and response performance, indicating that these two cities have reached a better balance in air pollution governance. Of the three cities with lower comprehensive performance rankings, Ganzhou ranks higher in stress performance but lower in the state and response performance. Therefore, it pulls down the overall ranking of Ganzhou. This indicates that Ganzhou is more active in air pollution pressure reduction, but not enough attention is paid to air pollution control investment, and the air quality is not well improved. The other two cities with lower rankings generally have lower rankings in pressure performance, state performance, and response performance, reflecting that the air quality improvement in these two cities is not obvious. These two cities should increase air pollution reduction and air pollution treatment investment.

4.5. Annual Change in Air Pollution Governance Performance

In

Table 5. Change in performance by region 2017 relative to 2014.

Region	Comparison of 2017 and 2014 Performance by Type
	Pressure Performance	State Performance	Reaction Performance	Comprehensive Performance
Nanchang City	Increase	Decrease	Increase	Decrease
Jingdezhen City	Increase	Increase	Increase	Increase
Pingxiang City	Decrease	Increase	Increase	Increase
Jiujiang City	Increase	Increase	Increase	Increase
Xinyu City	Increase	Increase	Decrease	Increase
Yingtan City	Increase	Decrease	Decrease	Increase
Ganzhou City	Increase	Increase	Increase	Increase
Ji’an City	Increase	Decrease	Increase	Increase
Yichun City	Increase	Increase	Increase	Increase
Fuzhou City	Increase	Decrease	Increase	Decrease
Shangrao City	Increase	Decrease	Increase	Decrease

Data source: Compiled by the author.

, the 2017 scores of each city are compared with the 2014 scores, and the rise and fall of each city’s performance in the pressure performance, state performance, response performance, and comprehensive performance are detailed.

As a whole, the air pollution governance state is good. 8 of the 11 cities whose comprehensive performance is up, and only three cities (Nanchang, Fuzhou, and Shangrao) are down. Most cities with rising comprehensive performance are those with higher average values of comprehensive performance in four years (Ji’an, Yingtan, Jiujiang, and Xinyu). In addition to comparing the changes in comprehensive performance, it is also possible to compare the changes in pressure performance, state performance, and reaction performance. In terms of pressure performance, among the 11 cities, only Pingxiang City had a lower pressure performance in 2017 than in 2014, which reflects that the majority of cities have made some achievements in this part of air pollution pressure reduction. Regarding response performance, all cities have increased except Yingtan and Xinyu, whose performance in 2017 was lower than in 2014. This indicates that most cities have increased their attention to air pollution governance. As for the state performance, the cities that went up and down accounted for about half each. This reflects that compared with 2014, half of the cities in Jiangxi have seen a decrease in air quality improvement. A large part of this is due to the high historical debt of these cities in terms of air environment [12].

Combined with results from Table 3, it is found that not all cities with rising stress performance in 2017 compared to 2014 are rising year by year. A few cities show a wave-like upward trend (e.g., Ji’an and Fuzhou). In contrast, among the cities with rising reaction performance over 2014, most of them show a wave-like rise. Few cities show a year-on-year increase in reaction performance (e.g., Ganzhou). Similarly, among the cities with rising performance in the comprehensive and state categories, all of them show a wave-like upward trend. Thus, it can be seen that although some cities’ 2017 performance in various categories increased from 2014, most of them showed a wave-like increase. Therefore, the air pollution governance performance of most cities in Jiangxi is not a continuous improvement but an intermittent one.

5. Conclusions and Prospect

In recent years, Jiangxi has effectively raised its political stance and is fully engaged in the battle of air pollution prevention and control. This contributes to the construction of beautiful China’s “Jiangxi model” and promotes Jiangxi and even the country to build a moderately prosperous society. The core objective of this paper is to improve the performance of air pollution governance and enhance air quality. The research aims to investigate the air pollution governance performance of 11 prefecture-level cities in Jiangxi. The main findings of this paper are as follows.

First, regarding the types of indicators, the order of weights obtained by the entropy value method are state indicators, pressure indicators, and response indicators. The pressure indicator reflects the impact and damage of human activities on the environment, the state indicator describes the state and change of the air environment in a specific period, and the response indicator reflects the prevention, remediation, and recovery measures taken to deal with the problem of air pollution. Therefore, to improve the environment and obtain higher air pollution governance performance, more emphasis must be placed on controlling pollution sources and reducing pollution emissions rather than post-pollution governance. This is consistent with the conclusions of scholars like Li [12], that is, controlling pollution sources and reducing pollution emissions are the internal driving force to promote the improvement of regional air environment quality.

Second, regarding regional comparison, only a few regions show a good balance between “pressure-state-response”, while most regions show the phenomenon of ignoring one and losing the other. When the four-year averages of various performance categories are ranked among the prefecture-level cities in Jiangxi Province, only Nanchang and Yingtan are among the top two cities in all categories. They not only ranked top in comprehensive performance but also performed well in pressure performance, state performance, response performance, and other sub-categories, showing a good balance between “P-S-R”. Most of these cities belong to areas with good economic development or ecological environment resources, so they have strong comprehensive management ability or rich natural resource endowment. For example, the number and scale of ecological funds invested, and remediation projects carried out in Nanchang City are increasing. It has increased investment in air pollution control through many fiscal transfer payments, guaranteeing air pollution management. At the same time, it has continued to optimize the industrial layout, shut down a large number of highly polluting enterprises, and strictly monitored the discharge of pollutants from the remaining enterprises, thus greatly reducing air pollution in the region.

Thirdly, in terms of annual variation, the comparison between 2017 and 2014 shows that the performance of most regions has increased in several categories. It shows that, on the whole, the air pollution governance of all cities has achieved good results, and the air quality of cities divided into districts in the province continues to consolidate and improve. This is closely related to Jiangxi Province’s continuous efforts in recent years. In the past five years, several important policy documents have been issued successively, including the Implementation Rules of Jiangxi Province’s Air Pollution Prevention and Control Action Plan, Regulations of Jiangxi Province on the Prevention and Control of Exhaust Pollution from Motor Vehicles, and Measures of Jiangxi Province for the Prevention and Control of Dust Pollution. The provincial government and municipalities have signed targets for air pollution prevention and control, making solid progress to ensure effective results in pollution governance. In addition, in terms of the characteristics of change, these categories’ performance has not increased yearly but in waves. For example, the trend of pressure performance in Ji’an and Fuzhou, response performance in most cities, and comprehensive performance and state performance in all cities are all rising in waves. Thus, the air pollution governance performance of most cities in Jiangxi is not a continuous improvement but an intermittent one. The possible reason is the difficulty and complexity of the task of regional pollution governance. With the rapid development of industry, energy consumption, and the rapid growth of motor vehicle ownership in Jiangxi Province, the secondary pollution in all urban areas shows an aggravating trend, and the compound and regional air pollution phenomenon is increasingly prominent. At the same time, adverse meteorological conditions such as regional temperature inversion occur from time to time, resulting in poor air diffusion capacity and pollutant accumulation, making it difficult to promote haze governance work.

Based on the above conclusions, the following recommendations are made.

First, the performance evaluation index system of air pollution governance should be improved. According to the General Office of the State Council issued in 2014, “Measures for assessing the Implementation of the Action Plan for Air Pollution Prevention and Control (Trial),” careful consideration of the state of air pollution governance in Jiangxi through the “pressure-state-response” relationship, to determine the classification and linkage between performance assessment indicators. Not only should the air pollution emission reduction be included in the performance assessment index, but also the air pollution control investment and air environment improvement should be included in the assessment, and the weight of each index should be reasonably assigned.

Second, the government should promote the comprehensive governance of air pollution. Comprehensive governance is reflected in the whole process of air pollution emission reduction, air pollution management input, and air environment improvement. In air pollution governance, some cities in Jiangxi province appear to give priority to one and lose another. That is, pay attention to air pollution reduction and ignore the other, or pay attention to air pollution control investment and ignore the other. For example, in Xinyu, Jiujiang, Yichun, and other cities, the air pollution control investment is increasing, but the effect of air pollution reduction is not satisfactory. Such cities should focus on how to effectively control the sources of pollution and reduce air pollution emissions. Therefore, each region in Jiangxi’s air pollution governance process must consider air pollution reduction, environmental improvement, and air pollution control investment to achieve comprehensive governance.

Thirdly, the government should focus on the weaker areas of air pollution governance. According to the Cask principle, the water capacity of a barrel is closely related to the length of each plank that makes it up. If the boards have no leakage seams with each other and the bottom board is solid, the volume of water in the barrel depends on the shortest piece of wood [44]. The Cask principle can perfectly explain the overall high and low performance of air pollution governance in Jiangxi. Air pollution governance has been carried out for many years in various prefecture-level cities in Jiangxi, and there is no shortage of cities with remarkable governance results every year. However, Jiangxi’s inter-provincial air pollution governance performance does not rise year after year. To a large extent, is the governance performance of the city behind the overall performance of Jiangxi Province. Therefore, Jiangxi should establish a list of cities to focus on each year and include those cities that are behind and in a downward trend (such as Jingdezhen and Pingxiang) in the list. And put forward targeted rectification requirements to break the weak areas of air pollution governance one by one to improve the overall performance of air pollution governance in Jiangxi and thus improve air quality.

Fourth, the government should strengthen the regional collaborative governance of air pollution. The mobility and complexity of air pollution show that the ability of air pollution governance relying solely on the territorial model is very limited. In order to improve the regional air quality, the traditional governance pattern of local governance should be abandoned and transformed into a regional collaborative governance model. The once severe air pollution in the Pearl River Delta (PRD) region has become a national benchmark after many years of collaborative regional governance, with a sustained reduction in air pollution and the completion of the national air quality improvement target ahead of schedule. Thus, based on the successful experience of air pollution governance in the PRD region, and combined with the local situation, to strengthen intra-regional collaborative governance and deepen inter-regional collaborative governance. Through collaborative governance, we can improve the air pollution governance model, accelerate the improvement of air pollution policies and measures, and improve the governance performance of air pollution.

In conclusion, this paper studies air pollution governance from the micro level and puts forward the analysis method of constructing an index system based on the PSR model, which has reference significance for scientific guidance of social and economic activities and formulation of corresponding air environmental protection policies and measures. At the same time, due to the constraints of data availability and uncertainty of pollutant measurement in this paper, the evaluation index and data time series can be further optimized if conditions permit.