Pollutants Reduction Effect of Vertical Environmental Reform in China

doi:10.21203/rs.3.rs-3228241/v1

Download PDF

Research Article

Pollutants Reduction Effect of Vertical Environmental Reform in China

https://doi.org/10.21203/rs.3.rs-3228241/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 07 Feb, 2024

Read the published version in Environmental Science and Pollution Research →

You are reading this latest preprint version

This study uses data of 263 prefecture-level cities from 2013 to 2020 to examine the effects of China's vertical environmental reform on pollutant emissions in urban areas. The findings demonstrate that vertical environmental reform improves local governments' ability to enforce environmental laws, resulting in a reduction in regional pollutant emissions with long-lasting effects. The reason is that the vertical environmental reform has achieved a division between county-level governments and county-level environmental protection bureaus and has granted formal authority and real authority to the environmental protection bureaus in provincial-level and prefecture-level governments, thereby enhancing the independence of environmental monitoring and enforcement and ensuring the strict enforcement at the grassroots level. According to heterogeneity analysis, local governments' financial pressure and official associations do not have a heterogeneous impact on the vertical environmental reform's ability to reduce pollution, and the reform treats border pollution problems more effectively.

Vertical Management

Environmental Governance

Formal Authority

Real Authority

Pollutant Emissions

Environmental Enforcement

The environment exhibits market failures as a typical public good. Government involvement in the preservation of the environment is important. The key to addressing the problem in the reform of China's environmental management system is how to play the function of the government in the system of environmental governance efficiently. The territorial management model had served as the foundation for China's environmental management system for a very long period. Although local environmental protection bureaus (EPBs) were operationally supervised by higher-level environmental protection departments, local governments, which displayed typical decentralization features, were in charge of hiring staff and allocating funds for such agencies. Following China's reform and openness, local GDP development and local tax revenues have become the most crucial administration objectives for local governments as a result of fiscal decentralization (Qian and Weingast, 1997) and political promotion incentives (Li and Zhou, 2007). In an effort to boost economic growth and increase the number of taxpayers, local government authorities frequently lowered environmental restrictions on polluting businesses and even disturbed environmental monitoring and enforcement.

The national vertical environmental reform (VER) in EPBs was formally launched by the central government in September 2016 in an effort to address the aforementioned environmental governance challenges. The central government expects the change as a crucial step in ending local protectionism and properly implementing the duties of grassroots environmental enforcement. According to statistics on pollutant emissions (Figure 1) published by China's MEE (Ministry of Ecology and Environment), SO₂ emissions have significantly decreased since 2016, which is also the year when VER began. Did China's pollution emissions decrease as a result of the VER?

There has been a broad debate about whether the territorial management model or vertical management model is more efficient for government in the academic. Tiebout (1956) provides a decentralization theory in which local governments have more local information, are more aware of local people's needs, and can respond more precisely than the central government. The disadvantage of the vertical management model is that the central government is unable to consistently account for the variations in preferences, technical advancements, and cultural elements of inhabitants in various regions (Besley and Coate, 2003; Millimet, 2003). In the case of free mobility of local citizens, local governments can deliver public goods in a more targeted and effective manner under a territorial management model (Stigler, 1957). There is also literature that expresses the other viewpoint. For instance, territorial management models may lead to the free-riding behavior of local governments (Lipscomb and Mobarak, 2017). Additionally, it is asserted in the literature that neither territorial nor vertical administration has an absolute advantage, and that the merits of both depend on the marginal advantages of differentiated government provision of public goods versus the external marginal costs (Oates, 2003).

In the field of Chinese environmental governance specifically, the majority of the research contends that the country's long-standing territorial management model of environmental regulation plays a significant role in the environmental governance's failure (Wang et al., 2003; Golding, 2011). Local GDP growth goals and environmental governance goals frequently conflict in China as it industrializes, which makes local governments less inclined to regulate the environment (Kostka, 2013). There are even local protectionist behaviors under the territorial management model, such as local governments tolerating illegal emissions by businesses and lowering environmental standards to entice high-energy and high-polluting businesses to invest and build factories in their jurisdictions (Jia and Nie, 2017).

This paper uses the difference-in-differences (DID) method to investigate the changes in SO₂ emissions of local industries after the vertical environment reform in China from 2012 to 2020 with prefecture-level data, and responds to the dispute between vertical management mode and territorial management mode in the field of government environmental protection. A small number of scholars (Chao et al, 2021; Chao and Han, 2022; ) conducted some empirical studies for the 2002 sub-municipal vertical environment reform pilot in Shanxi Province[1]. However, since the sub-municipal vertical environment reform pilot in Shanxi Province was implemented by local governments in a special context[2], there could be endogeneity between the reason for reform and the result of vertical reform. The nationwide vertical environment reform in 2016 was deployed by the central government, it is more mandatory and coordinated, and has better exogeneity, providing researchers with a better opportunity to empirically test the differences in governance effects between territorial and vertical management. However, due to the short time of the reform and the delay in its actual implementation in the provinces, the empirical literature on this round of reform is still relatively small. Based on the theory of control, this study gives an organizational explanation for the governance effect of vertical reform by describing the shift in the governmental governance model produced by the withdrawal of environmental inspection authority. Additionally, this study tests the mechanism using data from municipal environmental monitoring and enforcement at a more granular level and is able to demonstrate that the impact of vertical reform is primarily due to the regulation of stricter enforcement by the local environmental protection departments.

2.1. Background: rights of control and choice of governance objectives among government organizations before the VER

First, we analyze the rights between local governments and EPBs before the VER (Fig. 2). The provincial EPBs assigned annual emission reduction targets to the municipal EPBs based on the central government's environmental emission reduction targets for the province, and they examined and verified the municipalities' achievement of the targets on a regular schedule. The provincial EPBs were the principal of environmental emission reduction work and had the right to set targets and examined the achievement of the municipal EPBs. The municipal EPBs received the environmental emission reduction targets from the provincial EPBs and then decomposed and assigned the targets to the county-level EPBs. The county-level EPBs, acting as agents, were ultimately the ones who carried out environmental monitoring and enforcement. Municipal EPBs, on the other hand, acted as managers in distributing incentives to the county EPBs. Both municipal and county-level EPBs faced administrative subordination from the local party government and operational guidance from the higher-level EPBs. In this dual management system, the county-level EPBs' incentive allocations were both approved by the municipal EPBs and controlled by the county-level governments' personnel and financial controls. This required a further subdivision of incentive allocation rights into performance approval rights and personnel and financial authority rights.

Performance approval rights refer to the authority of the manager (municipal EPBs) to give performance approval to the agent (county-level EPBs) for policy implementation results. Personnel and financial rights refer to the power of personnel appointment and removal, and financial fund support, owned by the administrative subordinate superior. According to the relevant fieldwork literature (Zhou and Lian, 2020), prior to the vertical environmental reform, the performance approval of county-level EPBs was decided by the municipal EPBs, while the personnel and financial arrangements of local EPBs depended on the local governments. Although both performance approval rights and personnel and financial rights have incentive allocation functions, they are two different types of incentive functions. Performance approval rights are from the business-directed relationship, and whether county-level EPBs' performance approval is recognized by the county-level government and whether county-level EPBs can be rewarded or punished accordingly still depends on the county-level government, so the performance approval rights of the municipal EPB are symbolical (Aghion and Tirole, 1997) and belong to weak incentives. In contrast, personnel and financial rights are strong incentives based on administrative affiliation. Because the policy directives of the county-level governments are political in nature, the degree of political loyalty is reflected by whether the county EPBs implement them or not. Therefore, personnel financial rights have both formal authority based on organizational status and real authority that determines real rewards and punishments. Personnel financial rights’ incentive intensity is higher than that of performance approval rights. It is easy to explain why the actual effectiveness and durability of governance on the environment is not good. Because the formal authority and real authority of the personnel and financial rights belong to the local governments, the implementation of environmental monitoring and enforcement always need to get the county-level governments’ personnel and financial support, which makes it easy for county-level EPBs to be interfered with by county-level governments when they prioritize economic goals in the governance targets. At the same time, the municipal EPBs need to form an interdependent and long-term coexisting political alliance with county-level governments in order to get the personnel and financial support in the environmental protection work, enabling county-level governments to influence the performance approval of municipal EPBs in their own jurisdictions, and ultimately the performance approval rights tend to be symbolic. Only when higher-level political forces implement campaign-style governance in the environmental field, and environmental goals are cascaded down through the bureaucratic system in the form of political tasks, county-level governments briefly reorient environmental goals to priority goals and achieve transient environmental governance effectiveness.

2.2. Theoretical analysis: the governance effects of vertical environmental reform

The core of the VER is the change in personnel and financial rights. Following the environmental vertical reform, the provincial EPBs appointed and dismissed the municipal EPBs’ director, as well as the party secretary. The appointment and removal of personnel and funding of municipal environmental monitoring agencies are also the responsibility of the provincial EPBs. As shown in Fig. 3, most of the personnel and financial authority of municipal EPBs have been transferred to the provincial EPBs after the the VER.

The environmental monitoring and enforcement in environmental governance are carried out by county-level EPBs, and the ability of county-level enforcement parties to effectively implement environmental monitoring and enforcement and resist local protectionist intervention is the key to environmental governance and the focus of institutional reform in the VER. The county-level EPBs adjusted to the municipal EPBs of the dispatched agencies after the VER; the county-level EPBs are now directly managed by the municipal EPBs, the leadership team is chosen and fired by the municipal EPBs, and the municipal EPBs are in charge of funding. To achieve a separation between county-level party governments and county-level EPBs, county-level EPBs are even administratively subject to the municipal EPBs. It is more challenging for county-level governments to intervene county-level EPBs' environmental enforcement because they lack formal authority over county-level EPBs based on organizational status or real authority over personnel appointments, dismissals, or financial support. A high degree of connection between municipal EPBs and county-level EPBs is also achieved through the centralized transfer of power from county-level EPBs to municipal EPBs. Municipal EPBs no longer need to rely on personnel and financial support from county-level party governments to complete grassroots environmental monitoring and enforcement, effectively breaking the political alliance between municipal EPBs and county-level party governments, guaranteeing the independence of environmental monitoring and enforcement, and facilitating strict enforcement by grassroots environmental departments, thereby reducing regional pollutant emissions. Based on this, the research hypothesis of this paper is proposed:

Hypothesis 1

The VER has reduced the regional pollutant emissions.

Hypothesis 2

The VER reduces regional pollutant emissions by improving environmental enforcement at the municipal level and below.

3.1. Model

Since the implementation time of the VER in each province varies, this paper adopts a time-varying difference-in-difference (DID) model, in which the areas that have implemented the vertical environmental reform in different years of the sample period are used as the treatment group and the areas that have not implemented the vertical environmental reform are used as the control group. This paper uses a time-varying DID model to compare the changes in regional pollutant emissions before and after the implementation of the policy in the two sample groups to estimate the governance effects of the vertical environmental reform. Drawing on Beck et al. (2010), the following econometric model was designed to test hypothesis 1.

$${\text{p}\text{u}\text{l}\text{l}\text{u}\text{t}\text{i}\text{o}\text{n}}_{\text{i}\text{t}}={\alpha }+{\beta }{\text{V}\text{E}\text{R}}_{\text{i}\text{t}}+{\delta }{\text{c}\text{o}\text{n}\text{t}\text{r}\text{o}\text{l}}_{\text{i}\text{t}}+{{\mu }}_{\text{i}}+{{\gamma }}_{\text{t}}+{{\epsilon }}_{\text{i}\text{t}}$$

Where ${\text{p}\text{u}\text{l}\text{l}\text{u}\text{t}\text{i}\text{o}\text{n}}_{\text{i}\text{t}}$ represents the pollutant emission of city i in year t. The dummy variable, ${\text{V}\text{E}\text{R}}_{\text{i}\text{t}}$, indicates if the city is subject to the vertical environment reform or not in year t. Before the implementation of vertical environmental reform in city i, ${\text{V}\text{E}\text{R}}_{\text{i}\text{t}}$=0; After the implementation of vertical environmental reform in city i, ${\text{V}\text{E}\text{R}}_{\text{i}\text{t}}$=1. The ${\text{c}\text{o}\text{n}\text{t}\text{r}\text{o}\text{l}}_{\text{i}\text{t}}$ is a set of control variables. ${{\mu }}_{\text{i}}$represents city fixed effects, controlling for time-invariant city characteristics. ${{\gamma }}_{\text{t}}$represents year-level time fixed effects, controlling for city-invariant year characteristics.${{\epsilon }}_{\text{i}\text{t}}$ is a random disturbance term. The model is estimated using clustered robust standard errors. If hypothesis 1 holds, the coefficient of ${\text{V}\text{E}\text{R}}_{\text{i}\text{t}}$, should be significantly smaller than 0.

The mediating effect model is designed to test whether the VER reduces regional pollutant emissions by improving environmental enforcement at the municipal level and below. We use the two-step approach to achieve the mediating effect model. The first step is to test the impact of VER on the grassroots environmental enforcement:

$${\text{e}\text{n}\text{f}\text{o}\text{r}\text{c}\text{e}\text{m}\text{e}\text{n}\text{t}}_{\text{i}\text{t}}={\alpha }+{\beta }{\text{V}\text{E}\text{R}}_{\text{i}\text{t}}+{\delta }{\text{c}\text{o}\text{n}\text{t}\text{r}\text{o}\text{l}}_{\text{i}\text{t}}+{{\mu }}_{\text{i}}+{{\gamma }}_{\text{t}}+{{\epsilon }}_{\text{i}\text{t}}$$

which${\text{e}\text{n}\text{f}\text{o}\text{r}\text{c}\text{e}\text{m}\text{e}\text{n}\text{t}}_{\text{i}\text{t}}$ represents the grassroots environmental enforcement’s strength. Other variables and model settings are the same as in (1). The second step is to regress the ${\text{V}\text{E}\text{R}}_{\text{i}\text{t}}$ and ${\text{e}\text{n}\text{f}\text{o}\text{r}\text{c}\text{e}\text{m}\text{e}\text{n}\text{t}}_{\text{i}\text{t}}$with ${\text{p}\text{u}\text{l}\text{l}\text{u}\text{t}\text{i}\text{o}\text{n}}_{\text{i}\text{t}}$. The other variables and model settings remain the same as in (1).

$${\text{p}\text{u}\text{l}\text{l}\text{u}\text{t}\text{i}\text{o}\text{n}}_{\text{i}\text{t}}={\alpha }+{\beta }{\text{V}\text{E}\text{R}}_{\text{i}\text{t}}+{\phi }{\text{e}\text{n}\text{f}\text{o}\text{r}\text{c}\text{e}\text{m}\text{e}\text{n}\text{t}}_{\text{i}\text{t}}+{\delta }{\text{c}\text{o}\text{n}\text{t}\text{r}\text{o}\text{l}}_{\text{i}\text{t}}+{{\mu }}_{\text{i}}+{{\gamma }}_{\text{t}}+{{\epsilon }}_{\text{i}\text{t}}$$

If the mediating effect of environmental enforcement at the grassroots level exists, firstly, the implementation of the vertical environment reform should significantly increase the environmental enforcement at the grassroots level, so that ${\beta }$ in Eq. (2) should be significantly greater than 0. Besides, the increase of grassroots environmental enforcement should reduce the regional pollutant emissions, ${\phi }$ in Eq. (3) is is significantly less than 0.

3.2. Variables

3.2.1. Pollutant Emissions

In this paper, industrial sulfur dioxide emissions (SO₂) are selected to construct the regional pollution emission variable. The reason is that China's coal-based energy structure determines that the main form of environmental pollution in China is air pollution. SO₂ from coal combustion is a serious health hazard to the population and has been one of the main targets of pollution reduction by the China’s central government since 1998. Using industrial SO₂ emissions as the explanatory variable can examine the achievement of environmental emission reduction targets by each local government. Also SO₂ is a common indicator choice for scholars to explore environmental pollution problems (Antweiler et al., 2001; Sultani and Shengbin, 2021). In addition, the other pollutants, industrial wastewater emissions (effluents) and industrial smoke emissions (fumes), are also selected for robustness testing based on data availability. These three pollutants are the main targets of China’s central government's emission reduction efforts from the Eleventh Five-Year Plan to the Fourteenth Five-Year Plan.

3.2.2. Grassroots environmental enforcement efforts

This paper measures regional grassroots environmental enforcement efforts as ${\text{e}\text{n}\text{f}\text{o}\text{r}\text{c}\text{e}\text{m}\text{e}\text{n}\text{t}}_{\text{i}\text{t}}$ using the number of environmental protection administrative penalty decisions made by EPBs at the municipal and county levels in a single year. Compared to the existing literature (Chao et al., 2021; Chao and Tian, 2022; Jin et al, 2023) which uses the frequency of environmental inspections and the amount of environmental enforcement fines at the provincial level in the province where the prefecture-level city is located to measure the strength of environmental enforcement proxy municipal environmental enforcement, this paper uses the number of formal administrative penalty decisions of municipal and county-level EPBs, which can better reflect the enforcement strength of grassroots EPBs strength.

3.2.3. Control variables

The level of economic development (Grossman and Krueger, 1995), industrial structure, population density, fiscal pressure, and foreign investment may affect regional pollutant emissions and environmental enforcement. In this paper, we control for regional GDP (gdp), GDP growth rate (gdpg), industrial structure (ind), population density (dens), fiscal pressure (fisc), and foreign real investment (fdi). Meanwhile, considering the possible lagged effects of social factors such as sudden environmental pollution accidents and public environmental demands on regional environmental enforcement and pollutant emissions, this paper adds the number of sudden environmental incidents in the previous year (acc) as a control variable. Due to the lack of data on environmental emergencies at the prefecture-level city level, this paper draws on existing literature (Chao et al., 2021; Chao and Tian, 2022) using provincial data from the province where the prefecture-level city is located for matching.

3.3. Data sources and descriptive statistics

The data of SO₂, effluents and fumes is from the the “Statistical Yearbook of China’s City”. Data on provincial environmental monitoring and enforcement, and environmental emergencies is obtained from the “China Environmental Yearbook” in previous years. Data on regional economy, population, finance, etc. are from the China Stock Market & Accounting Research Database (CSMAR). Missing data on economy, population and environment of some cities are obtained by collecting local government’s statistical yearbooks of each city to make up for them. Missing data on provincial environmental emergencies are obtained by searching the provincial EPBs. The data on administrative environmental penalty decisions of prefecture-level cities were obtained by searching the PKULAW database for prefecture-level and county-level administrative environmental penalty documents. After excluding sub-province-level cities, ethnic autonomous cities, leagues, cities without districts and counties, and samples with missing data, the unbalanced panel data of 263 prefecture-level cities from 2013 to 2020 were obtained as the samples for this paper. The reasons for using 2013 as the starting year and 2020 as the end year of the sample are: (1) the DID method requires retaining the sample for a period before the implementation of the policy for differential estimation with the period after the implementation of the policy, which should ensure that the explanatory variables are not affected by other policies. Since the Chinese government and society have paid significantly more attention to environmental issues since the 18th CPC Congress in 2012, this paper uses 2013 as the starting year of the sample to avoid the influence of the above factors on the estimation results to a certain extent; (2) since the regional pollutant emission data in this paper are obtained from the China Urban Statistical Yearbook, the latest data currently counted are for the year 2020. Restricted by data acquisition, the year 2020 is the latest point of time when relevant data can be obtained. The sample of 263 prefecture-level cities in this paper covers 27 provinces, which can better represent the overall situation of the country. In order to avoid the influence of extreme values, all continuous variables are conventionally winsorized at 1% and 99%. The descriptive statistics of the main variables in this paper are shown in Table 1.

Table 1

Descriptive statistics of main variables
Variable	Definition	No. of obs.	Mean	S.D.	Min	Max
SO₂	Regional industrial SO₂ emissions (million tons)	1749	2.804	3.188	0.027	15.840
effluents	Regional industrial wastewater emissions (10 million tons)	1749	4.651	4.798	0.053	26.360
fumes	Regional industrial smoke emissions (10 million tons)	1749	2.179	2.730	0.031	15.400
enforcement	Number of environmental administrative penalty decisions (hundred)	1749	1.201	5.381	0.000	2.591
VER	Vertical environment reform, = 1 if a province has been implemented, = 0 otherwise	1749	0.286	0.452	0.000	1.000
gdp	Regional GDP (hundred billion yuan)	1749	2.036	1.846	0.231	10.210
gdp2	Squared items of regional GDP	1749	7.553	16.090	0.054	104.200
dens	Density, =population/area (10,000 people/km2)	1749	0.041	0.026	0.001	0.185
ind	Industry structure, =GDP of secondary industry / total GDP	1749	0.452	0.098	0.186	0.697
fisc	Fiscal Pressure, =Fiscal Expenditure / Fiscal Revenue	1749	3.241	1.963	0.876	11.800
fdi	Foreign investment (USD billion)	1749	5.438	8.886	0.000	50.400
acc	Number of environmental emergencies in the previous year	1749	22.900	36.942	0.000	125.000

4.1. Pollutant reduction effect

In the regression of Eq. (1), columns (1) (2) and (3) of Table 2 report the effects of the vertical environment reform on the regional industrial emissions of SO₂, effluents, and fumes respectively. The coefficients of the $\text{V}\text{E}\text{R}$ in columns (1) and (3) are all negative at the 1% significance level, while the coefficients of the $\text{V}\text{E}\text{R}$ in column (2) are negative at the 5% significance level, indicating that the implementation of vertical environment reform has significantly reducing the emissions of SO₂, effluents, and fumes in the region. The hypothesis 1 of this paper is verified.

Table 2

The impact of VER on regional pollutant emissions
	(1)	(2)	(3)
Variables	SO₂	effluents	fumes
VER	-0.519***	-0.543**	-1.303***
	(-2.992)	(-2.318)	(-6.161)
gdp	0.446	-0.559	0.906***
	(0.960)	(-1.299)	(2.739)
gdp2	-0.057*	0.019	-0.088***
	(-1.833)	(0.570)	(-4.082)
dens	25.251***	23.940**	12.410**
	(3.343)	(2.440)	(2.348)
ind	7.296***	5.062***	1.813
	(3.764)	(3.078)	(1.405)
fisc	0.119	-0.028	0.078
	(0.977)	(-0.413)	(1.023)
fdi	-0.009	-0.016	-0.001
	(-0.809)	(-1.033)	(-0.057)
acc	-0.149***	-0.023	-0.055***
	(-4.361)	(-0.828)	(-2.840)
Constant	0.902	3.852***	0.759
	(0.721)	(3.214)	(1.034)
Year fixed effects	Yes	Yes	Yes
City fixed effects	Yes	Yes	Yes
Observations	1749	1749	1749
R-squared	0.579	0.378	0.452
Notes: T-values in parentheses, *, , * represent p less than 0.01, 0.05, 0.1, respectively. same below.

4.2. Mechanism analysis

Column (1) of Table 3 reports the regression results for Eq. (2), omitting the control variables and the constant term. The coefficient of VER are positive at the 1% level of significance, indicating that the VER has significantly improved environmental enforcement at the grassroots level. Columns (2) (3) and (4) of Table 3 report the regressions using Eq. (3) on regional industrial SO₂ emissions, industrial smoke emissions, and industrial wastewater emissions, respectively, and the results show that the coefficients of the grassroots enforcement efforts${\phi }$ are all less than 0 at the 1% significance level, indicating the existence of a mediating effect. The results show that the coefficients of grassroots enforcement are less than 0 at the 1% significance level, indicating that there is a mediating effect. This further verifies hypothesis 2 of this paper: the VER reduces regional pollutant emissions by improving environmental enforcement at the municipal level and below.

Table 3

Mechanism analysis
	(1)	(2)	(3)	(4)
Variables	enforcement	SO₂	effluents	fumes
VER	0.219***	-0.382**	-0.431**	-1.151**
	(5.289)	(-2.123)	(-1.992)	(-2.340)
enforcement		-0.626***	-0.511***	-0.693***
		(-3.304)	(-3.598)	(3.482)
Control variables	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes
City fixed effects	Yes	Yes	Yes	Yes
Observations	1749	1749	1749	1749
R-squared	0.231	0.581	0.387	0.463

4.3. Parallel trend test and dynamic effect analysis

The parallel trend assumption of the DID method requires that the treatment group and the control group have the same trend before the policy occurs. In this paper, the difference in pollutant emissions between the treatment and control groups before and after the implementation of the VER was calculated to test the parallel trends and analyze the dynamic effects of the VER. Two dummy variables, ${\text{b}\text{e}\text{f}\text{o}\text{r}\text{e}}_{\text{i}\text{j}}$ and${\text{a}\text{f}\text{t}\text{e}\text{r}}_{\text{i}\text{k}}$, are set. When the observation time of the sample is the year j before the implementation of the VER in city i, the dummy variable ${\text{b}\text{e}\text{f}\text{o}\text{r}\text{e}}_{\text{i}\text{j}}$ is 1, otherwise it is 0. When the sample is observed in the year k after the implementation of the VER in city i, the dummy variable ${\text{a}\text{f}\text{t}\text{e}\text{r}}_{\text{i}\text{k}}$ is 1, otherwise it is 0. The test model is designed as follows:

$${\text{p}\text{u}\text{l}\text{l}\text{u}\text{t}\text{i}\text{o}\text{n}}_{\text{i}\text{t}}={\alpha }+\sum _{\text{j}=1}^{6}{{\beta }}_{\text{j}}{\text{V}\text{E}\text{R}}_{\text{i}\text{t}}\times {\text{b}\text{e}\text{f}\text{o}\text{r}\text{e}}_{\text{i}\text{j}}+\sum _{\text{k}=1}^{4}{{\beta }}_{\text{k}}{\text{V}\text{E}\text{R}}_{\text{i}\text{t}}\times {\text{a}\text{f}\text{t}\text{e}\text{r}}_{\text{i}\text{k}}+{\delta }{\text{c}\text{o}\text{n}\text{t}\text{r}\text{o}\text{l}}_{\text{i}\text{t}}+{{\mu }}_{\text{i}}+{{\gamma }}_{\text{t}}+{{\epsilon }}_{\text{i}\text{t}}$$

Equation (4) is the regression equation for the parallel trend test and dynamic effects analysis of pollutant emissions. The equation does not include the VER dummy variable for the year of implementation of the VER at to avoid full multicollinearity. Other variables and model settings are the same as those in (1). All coefficients of the regression results are plotted as dynamic effects. Figure 4, Fig. 5, and Fig. 6 show the results of parallel trend tests and dynamic effects analysis for industrial sulfur dioxide emissions, industrial wastewater emissions, and fumes, respectively. The results show that estimated values and 95% confidence intervals of the coefficients before the VER were around 0, indicating that the hypothesis of parallel trend is satisfied before the implementation of the VER in each region. The coefficients and 95% confidence intervals of the coefficients in the years after the VER are around 0. The estimated values and 95% confidence intervals of the coefficients are below the 0-axis, indicating that the implementation of the VER has a negative impact on SO₂, effluents, and fumes. The coefficients and 95% confidence intervals of the coefficients are below the 0 axis. The estimated values and 95% confidence intervals of the coefficients have been shifted downward, indicating that the emission reduction effect of the VER has been gradually enhanced over time, and the governance effect of the VER has a good sustainability.

4.4. Robustness

4.4.1. Sample selectivity bias mitigation (PSM-DID)

If the actual implementation time of VER in each province is related to each province's own characteristics, it will lead to systematic differences between the treatment and control groups at different time points before the occurrence of policy shocks, thus generating Sample Selection Bias (SSB). This paper employs Propensity Score Matching (PSM) to select the samples of the two groups in order to mitigate any potential Sample Selection Bias, and then bases its usage of the DID method on this. Specifically, we use the provincial GDP, the quadratic term of provincial GDP, population density, industrial share, fiscal pressure, actual foreign capital utilization, and the number of environmental emergencies in the previous year as covariates to calculate the propensity score matching for each province for each year of VER using the logit model, and then use the nearest neighbor matching method to find the matching score for the experimental group. The results are presented in Table 4, and the results of the control variables are omitted. The results show that the estimated coefficients ${\beta }$ of the matched VER implementation variables on regional industrial sulfur dioxide emissions, industrial wastewater emissions, and industrial smoke emissions are greater than 0 at the 1% significance level, indicating that the VER still has a significant effect on emission reduction after mitigating the selective bias of the sample, which also proves that the estimation results of the baseline regression are robust.

Table 4

PSM-DID estimation results
	(1)	(2)	(3)
Variables	SO₂	effluents	fumes
VER	-0.673***	-0.503***	-1.607***
	(-4.398)	(-3.372)	(-5.359)
Year fixed effects	Yes	Yes	Yes
City fixed effects	Yes	Yes	Yes
Observations	1002	1002	1002
R-squared	0.591	0.393	0.480

4.4.2. Placebo test

In order to test the effect of VER without randomness, a placebo test was conducted by randomly selecting the implementation point of VER in each prefecture-level city. For each prefecture-level city, the implementation time of the VER was randomly selected between 2013 and 2020, and then the pollutant emissions were regressed using the model in (1). The above process was randomly simulated 1000 times. The coefficients’ T-values from the regression results of 1000 random simulations were generated as kernel density plots, as shown in Fig. 7, Fig. 8, and Fig. 9, representing the placebo test results for SO₂, effluents, and fumes, respectively. The vertical solid lines in the plots represent the mean T-values of β for 1000 random simulations, the three vertical dashed lines from left to right in Figs. 7 represent the T-value levels of 1.64, 1.96, and 2.68, respectively. The three vertical dashed lines from right to left in Figs. 8 and 9 represent the T-value levels of -1.64, -1.96, and − 2.68, respectively. The results show that the T-values in Figs. 7, Fig. 8, and Fig. 9 are roughly normally distributed, and the mean T-values are all very close to 0. The absolute values of most of the T-values are less than 1.64, indicating that most of the random simulation results cannot pass the 10% significance test. Therefore, the random selection of the implementation point of VER has no significant effect on industrial sulfur dioxide emissions, industrial wastewater emissions, and industrial smoke emissions, the placebo is ineffective. This proves that the estimation results of the baseline regression in this paper are robust.

4.4.3. Replacement of the explanatory variable measure

The absolute indicators of the three types of pollutant emissions are used as the explanatory variables in the baseline regression, and two relative indicators of regional per capita pollutant emissions and pollutant emissions per unit of GDP are used here as the explanatory variables for robustness testing. The absolute indicators in the control variables are also converted into relative indicators per capita or relative indicators per unit of GDP. When using pollutant emissions per unit of GDP, gdp and gdp2 are excluded from the control variables. the specific model design is as follows:

$${\text{p}\text{e}\text{r}\text{p}\text{u}\text{l}\text{l}\text{u}\text{t}\text{i}\text{o}\text{n}}_{\text{i}\text{t}}={\alpha }+{\beta }{\text{V}\text{E}\text{R}}_{\text{i}\text{t}}+{{\delta }}_{1}{\text{p}\text{e}\text{r}\text{g}\text{d}\text{p}}_{\text{i}\text{t}}+{{\delta }}_{2}{{\text{p}\text{e}\text{r}\text{g}\text{d}\text{p}}_{\text{i}\text{t}}}^{2}+{{\delta }}_{3}{\text{d}\text{e}\text{n}\text{s}}_{\text{i}\text{t}}+{{\delta }}_{5}{\text{i}\text{n}\text{d}}_{\text{i}\text{t}}+{{\delta }}_{6}{\text{f}\text{i}\text{s}\text{c}}_{\text{i}\text{t}}+{{\delta }}_{7}{\text{p}\text{e}\text{r}\text{f}\text{d}\text{i}}_{\text{i}\text{t}}+{{\delta }}_{8}{\text{p}\text{e}\text{r}\text{a}\text{c}\text{c}}_{\text{i}\text{t}}+{{\mu }}_{\text{i}}+{{\gamma }}_{\text{t}}+{{\epsilon }}_{\text{i}\text{t}}$$

$$\frac{{\text{p}\text{u}\text{l}\text{l}\text{u}\text{t}\text{i}\text{o}\text{n}}_{\text{i}\text{t}}}{{\text{g}\text{d}\text{p}}_{\text{i}\text{t}}}={\alpha }+{\beta }{\text{V}\text{E}\text{R}}_{\text{i}\text{t}}+{{\delta }}_{3}{\text{d}\text{e}\text{n}\text{s}}_{\text{i}\text{t}}+{{\delta }}_{5}{\text{i}\text{n}\text{d}}_{\text{i}\text{t}}+{{\delta }}_{6}{\text{f}\text{i}\text{s}\text{c}}_{\text{i}\text{t}}+{{\delta }}_{7}\frac{{\text{f}\text{d}\text{i}}_{\text{i}\text{t}}}{{\text{g}\text{d}\text{p}}_{\text{i}\text{t}}}+{{\delta }}_{8}\frac{{\text{a}\text{c}\text{c}}_{\text{i}\text{t}}}{{\text{g}\text{d}\text{p}}_{\text{i}\text{t}}}+{{\mu }}_{\text{i}}+{{\gamma }}_{\text{t}}+{{\epsilon }}_{\text{i}\text{t}}$$

Columns (1)(2)(3) of Table 5 report the regression results on industrial sulfur dioxide emissions per capita (perso2), industrial wastewater emissions per capita (pereffluents), and industrial smoke emissions per capita (perfumes), respectively. Columns (4)(5)(6) report the regression results on industrial sulfur dioxide emissions per unit GDP (so2/gdp), industrial wastewater emissions per unit GDP (effluents/gdp), and industrial smoke emissions per unit GDP (fumes/gdp), respectively. The results show that the coefficients ${\beta }$ of the core explanatory variables in columns (1)-(6) are all negative at the 1% or 5% significance level, indicating that the implementation of the VER also has a more significant effect on the reduction of industrial sulfur dioxide, industrial smoke, industrial wastewater emissions per capita and emissions per unit of GDP. Therefore, the conclusions of this paper are still robust after changing the variable measures.

Table 5

Estimated results of pollutant emissions per capita and per unit of GDP
	(1)	(2)	(3)	(4)	(5)	(6)
Variables	perSO₂	pereffluents	perfumes	SO₂/gdp	effluents/gdp	fumes/gdp
VER	-0.002***	-0.002**	-0.004***	-0.208**	-0.239**	-0.594***
	(-2.859)	(-2.103)	(-4.461)	(-2.573)	(-2.385)	(-3.172)
Year fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
City fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Observations	1749	1749	1749	1749	1749	1749
R-squared	0.397	0.588	0.353	0.372	0.358	0.350

4.5. Heterogeneity analysis

4.5.1. Financial pressure

It has been shown that the higher the financial pressure, the lower the environmental governance efficiency of local governments, the higher the pollutant emissions (Zhang et al, 2018)^[19]. The VER blocks the control of local governments over grassroots environmental protection departments through the redistribution of personnel and financial power, which theoretically prevents local governments from relaxing environmental protection supervision and enforcement due to financial pressure. Accordingly, the following model is designed to estimate the moderating effect of vertical environmental reform on pollutant emissions in areas affected by financial pressure.

$${\text{p}\text{u}\text{l}\text{l}\text{u}\text{t}\text{i}\text{o}\text{n}}_{\text{i}\text{t}}={\alpha }+{\beta }{\text{V}\text{E}\text{R}}_{\text{i}\text{t}}+{\theta }{\text{V}\text{E}\text{R}}_{\text{i}\text{t}}\times {\text{f}\text{i}\text{s}\text{c}}_{\text{i}\text{t}}+{\delta }{\text{c}\text{o}\text{n}\text{t}\text{r}\text{o}\text{l}}_{\text{i}\text{t}}+{{\mu }}_{\text{i}}+{{\gamma }}_{\text{t}}+{{\epsilon }}_{\text{i}\text{t}}$$

Since the control variable${\text{c}\text{o}\text{n}\text{t}\text{r}\text{o}\text{l}}_{\text{i}\text{t}}$ (fisc) variable has been included in the control variables, Eq. (5) only adds the interaction term of financial pressure (fisc) and VER on the basis of Eq. (1), the coefficient${\theta }$ represents the regulating effect of VER on pollutant emissions in areas affected by fiscal pressure, and the regression results are shown in Table 6. The coefficients of the interaction terms in columns (1), (2) and (3) are all significantly negative at the 1% level, indicating that VER has a significant negative moderating effect on pollutant emissions in areas affected by financial pressure. The transfer of personnel financial power from the local government on the "block" to the environmental protection department within the "article" can avoid the relaxation of environmental protection supervision and enforcement by the local government due to financial pressure.

Table 6

Impact of VER to crack financial pressure on regional pollutant emissions
	(1)	(2)	(3)
variant	so2	effluents	fumes
VMR	-0.248**	-0.179**	-0.938***
	(-2.294)	(-2.318)	(-3.259)
fisc	0.246	0.102**	0.195*
	(1.293)	(2.083)	(1.733)
VMR$\times$ fisc	-1.320***	-1.196***	-1.605***
	(-3.028)	(-2.632)	(-2.953)
control variable	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes
City fixed effect	Yes	Yes	Yes
Observations	1749	1749	1749
R-squared	0.581	0.391	0.456

4.5.2. Government official affiliation

The network of relationships among government officials may influence local environmental protection enforcement as well as local firms' emissions. In this paper, we determine whether there is an official connection between the local party governments and the EPBs by whether the director of the eEPBs of prefecture-level cities had working experiences in the party government of that prefecture-level city. The official association variable connection is designed to take the value of 1 if there is an association, otherwise it is 0. The specific model is as follows:

$${\text{p}\text{u}\text{l}\text{l}\text{u}\text{t}\text{i}\text{o}\text{n}}_{\text{i}\text{t}}={\alpha }+{\tau }{\text{V}\text{E}\text{R}}_{\text{i}\text{t}}\times {\text{c}\text{o}\text{n}\text{n}\text{e}\text{c}\text{t}\text{i}\text{o}\text{n}}_{\text{i}\text{t}}+{\beta }{\text{V}\text{M}\text{R}}_{\text{i}\text{t}}+{\delta }{\text{c}\text{o}\text{n}\text{t}\text{r}\text{o}\text{l}}_{\text{i}\text{t}}+{{\mu }}_{\text{i}}+{{\gamma }}_{\text{t}}+{{\epsilon }}_{\text{i}\text{t}}$$

In Eq. (6), the coefficient${\tau }$ represents the moderating effect of VER on pollutant emissions in the area affected by officials' connection, and adds officials' connection variables in the control variables, and other settings are the same as in Eq. (1). The regression results are shown in Table 7, and the coefficients of the interaction terms VER in columns (1), (2) and (3) are all at the 1% level. The coefficients of the $\text{V}\text{E}\text{R}\times \text{c}\text{o}\text{n}\text{n}\text{e}\text{c}\text{t}\text{i}\text{o}\text{n}$ in columns (1), (2) and (3) are all significantly negative at the 1% level, indicating that the "environmental protection reform" has a significant negative moderating effect on the pollutant emissions in the areas affected by officials' connection.

Table 7

Impact of "environmental protection reform" to break the association of officials on regional pollutant emissions
	(1)	(2)	(3)
variant	so2	effluents	fumes
VER$\times \text{c}\text{o}\text{n}\text{n}\text{e}\text{c}\text{t}\text{i}\text{o}\text{n}$	-0.429***	-0.691***	-0.583***
	(-3.021)	(-2.975)	(-2.628)
VER	-0.023	-0.007	-0.019
	(-1.528)	(-0.621)	(-0.952)
$\text{c}\text{o}\text{n}\text{n}\text{e}\text{c}\text{t}\text{i}\text{o}\text{n}$	0.126**	0.113**	0.202**
	(2.031)	(2.128)	(2.421)
control variable	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes
City fixed effect	Yes	Yes	Yes
Observations	1749	1749	1749
R-squared	0.579	0.379	0.452

4.5.3. Border contamination

Due to the negative externalities of environmental pollution, under the territorial management system, local governments are prone to opportunistic behavior by placing polluting enterprises at the borders of their administrative jurisdictions, enjoying all the economic benefits brought by the polluting enterprises, but only bearing part of the negative effects of environmental pollution (Agarwal et al., 2020; Silva and Zhu, 2009)^[20,21]. By enhancing the independence of environmental monitoring and enforcement at the grassroots level, the "environmental protection reform" can avoid the intervention of "local protectionism", and theoretically, it also has a governance effect on border pollution. The following econometric model is designed:

$${\text{p}\text{u}\text{l}\text{l}\text{u}\text{t}\text{i}\text{o}\text{n}}_{\text{i}\text{t}}={\alpha }+{\eta }{\text{V}\text{E}\text{R}}_{\text{i}\text{t}}\times {\text{b}\text{o}\text{r}\text{d}\text{e}\text{r}}_{\text{i}\text{t}}+{\lambda }{\text{V}\text{E}\text{R}}_{\text{i}\text{t}}\times (1-{\text{b}\text{o}\text{r}\text{d}\text{e}\text{r}}_{\text{i}\text{t}})+{\delta }{\text{c}\text{o}\text{n}\text{t}\text{r}\text{o}\text{l}}_{\text{i}\text{t}}+{{\mu }}_{\text{i}}+{{\gamma }}_{\text{t}}+{{\epsilon }}_{\text{i}\text{t}}$$

In Eq. (7), the ${\text{b}\text{o}\text{r}\text{d}\text{e}\text{r}}_{\text{i}\text{t}}$ represents whether the prefecture-level city is on the border of a province. If the prefecture-level city borders another province, the${\text{b}\text{o}\text{r}\text{d}\text{e}\text{r}}_{\text{i}\text{t}}$ The coefficient represents the impact of the VER on pollutant emissions in the border cities, and the coefficient is zero.${\eta }$ represents the impact of VER on pollutant emissions in the border cities, and the coefficient${\lambda }$ represents the impact of VER on pollutant emissions in non-border cities, and the coefficient of VER on pollutant emissions in non-border cities is added in the control variables.${\text{b}\text{o}\text{r}\text{d}\text{e}\text{r}}_{\text{i}\text{t}}$ The other settings are the same as Eq. (1). The regression results are shown in Table 8, the coefficients of interaction term VER$\times \text{b}\text{o}\text{r}\text{d}\text{e}\text{r}$ and the interaction term VER$\times (1-\text{b}\text{o}\text{r}\text{d}\text{e}\text{r})$ in columns (1) (2) (3) are less than 0 at different levels of significance, indicating that the VER has a certain inhibitory effect on pollutant emissions in border cities and non-border cities. The coefficients of the non-border cities are smaller and more significant, indicating that the VER has a more significant effect on pollutant emissions in the border cities and has a more significant governance effect.

Table 8

Effect of the VER on the management of border pollution
	(1)	(2)	(3)
variant	SO₂	effluents	fumes
VER$\times \text{b}\text{o}\text{r}\text{d}\text{e}\text{r}$	-0.724***	-0.802***	-1.384***
	(-5.294)	(-6.531)	(-6.384)
VER$\times (1-\text{b}\text{o}\text{r}\text{d}\text{e}\text{r})$	-0.238**	-0.149*	-0.195**
	(2.053)	(1.702)	(1.998)
control variable	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes
City fixed effect	Yes	Yes	Yes
Observations	1749	1749	1749
R-squared	0.579	0.379	0.452

This study examines the governance effects of the vertical management reform of the sub-provincial environmental protection sector using data from 263 prefectures from 2013 to 2020, explaining the mechanism of the provincial vertical management reform of the environmental protection sector from an organizational perspective based on the theory of control, and comes to the following conclusions: First, the VER considerably lowers emissions of industrial smoke (dust), industrial wastewater, and sulfur dioxide in the region. From the 11th Five-Year Plan to the 14th Five-Year Plan, these three pollutants were the main focus of the central government's efforts to reduce emissions. The local environmental protection departments' ability to carry out the central government's attempts to reduce pollution has been greatly enhanced by the VER's implementation; The provincial VER of environmental protection departments has allegedly decreased regional pollution emissions by enhancing environmental enforcement at the municipal level and lower, according to the intermediate effect concept. The following point is that the mediating effect model demonstrates that by stepping up environmental enforcement at the municipal level and below, the provincial VER of the environmental protection department has decreased regional pollution emissions. The political alliance between the municipal environmental protection bureaus and the county-level party governments has been broken by the VER's organizational separation of the county-level party governments from the county-level environmental protection departments, ensuring the independence of environmental monitoring and enforcement and strict enforcement by the grassroots environmental protection departments, which ultimately reduces regional pollution. Analysis of the dynamic effects demonstrates that the VER's governance effects are long-lasting. Comparatively to campaign-based governance, such environmental inspections, this organizational transformation can have a long-lasting impact on governance.

Different management systems produce different governance effects. For a country as vast and populous as China, how to adopt an effective management system and achieve effective governance of grassroots affairs is a real challenge for the Chinese government. The debate between the advantages and disadvantages of territorial and vertical management is both a theoretical focus in academia and a practical dilemma in the real world. The policy implication of this paper is that it provides a systematic quantitative assessment of the pollution control effects of the vertical reform of environmental protection using national prefecture-level municipal data and identifies the "black box" of the mechanism of the VER effect on emission reduction. This study provides empirical evidence on the effective division of labor and collaboration between local governments and environmental protection departments, as well as between higher and lower levels of environmental protection departments, to ensure long-term environmental governance. Given that local government interventions characterized by "local protectionism" have been widely and long-standing in other areas of government governance, the experience of breaking local protectionism found in this paper in the environmental governance area can also be used as a reference for reforming the organizational structure of government in other areas. At the theoretical level, this paper provides a new theoretical perspective on the debate between the advantages and disadvantages of local and vertical management by analyzing the organizational logic of the shift from the territorial management model to the vertical management model in the field of environmental management in China. In addition, this paper introduces a marginal expansion of the theory of residual control at the municipal and county levels based on the theoretical framework of residual rights of control, which is a further development of the residual control theory in the field of government governance.

Author information

Authors and Affiliations

School of Public Policy and Management, Tsinghua University, 100084, Beijing, People’s Republic of China

Qibo Chen

Contributions

QC: conceptualization, methodology, investigation, data curation, software, visualization, writing, reviewing, and editing.

Corresponding author

Correspondence to Qibo Chen.

Ethics declarations

Ethics approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

The copyright is transferred to Springer effective if and when the article is accepted for publication. The authors sign for and accept responsibility for releasing this material on behalf of any and all co-authors. The copyright transfer covers the exclusive right to reproduce and distribute the article including reprints, translations, photographic reproductions, microform, electronic form (offline, online), or any other reproductions of similar nature.

Conflict of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. All authors have approved the manuscript.

Funding This work was supported by the Major Program of National Fund of Philosophy and Social Science of China (CN) (grant number 18ZDA040).

Data availability Some or all data, models, or codes generated or used during the study are available from the corresponding author upon request.

Competing interests The authors declare no competing interests.

Agarwal S , Foo Sing T , Yang Y .The impact of transboundary haze pollution on household utilities consumption[J].Energy Economics, 2020, 85.
Aghion P, Tirole J. Formal and Real Authority in Organizations[J]. Journal of Political Economy ,1997,105:37-83.
Antweiler W, Copeland B, Taylor M. Is Free Trade Good for the Environment[J]. American Economic Review, 2001, 91(4):877-908.
Beck T, Ross L, Alexey L. Big Bad Banks? The Winners and Losers from Bank Deregulation in the United States[J]. Journal of Finance,2010,65 (5): 1637-67.
Besley T, Coate S. Centralized versus Decentralized Provision of Local Public Goods: A Political Economy Approach[J]. Journal of Public Economics, 2003,87(12): 2611-2637.
Bing Zhang, Xiaolan Chen, Huanxiu Guo. Does central supervision enhance local environmental enforcement? Quasi-experimental evidence from China J. Public Econ., 164 (2018), pp. 70-90
Chao Han, Xian-Liang Tian,Less pollution under a more centralized environmental system: Evidence from vertical environmental reforms in China, Energy Economics, Volume 112, 2022,106-121
Chao Han, Xiaolin Sun and Xian-Liang Tian,Land Economics, November 2021, 97 (4) 723-744
Grossman G, Krueger A. Economic Growth and the Environment[J].The Quarterly Journal of Economics,1995,2(110):353-377.
Hao Jin, Wenhan Liu, Shiyi Chen, Centralization and firm pollution emissions: Evidence from City–county Merger in China, Journal of Cleaner Production, Volume 416,2023,137964.
Hongbin Li, Li-An Zhou,. 2005. Political turnover and economic performance: the incentive role of personnel control in China[J]. Journal of Public Economics, (89), 9–10.
Hua Wang, Nlandu Mamingi, Benoit Laplante, Susmita Dasgupta Incomplete enforcement of pollution regulation: bargaining power of Chinese factories Environ. Resour. Econ., 24 (3) (2003), pp. 245-262
Kostka，G．，2013，“Environmental Protection Bureau Leadership at the Provincial Level in China: Examining Diverging Career Backgrounds and Appointment Patterns”，Journal of Environmental Policy and Planning，15( 1) ，41—63．
Lipscomb M, Mobarak A M. Decentralization and Pollution Spillovers: Evidence from the Re-drawing of County Borders in Brazil[J].Review of Economic Studies, 2017,84(1): 464-502.
Millimet D L. Assessing the Empirical Impact of Environmental Federalism[J]. Journal of Regional Science, 2003,43(4) : 711- 733.
Oates W E, Portney P. Chapter 8-The Political Economy of Environmental Policy[J]. Handbook of Environmental Economics, 2003,1( 2):325-354.
Qian Y, Weingast B R. Federalism as a commitment to reserving market incentives[J]. Journal of Economic Perspectives, 1997,11(4):83-92.
Ruixue Jia, Huihua Nie; Decentralization, Collusion, and Coal Mine Deaths[J].The Review of Economics and Statistics, 2017,99 (1): 105- 118.
Silva E C D , Zhu X .Emissions trading of global and local pollutants, pollution havens and free riding[J].Journal of Environmental Economics & Management, 2009, 58(2):169-182.
Stigler G. The Tenable Range of Functions of Local Government. In U.S. Congress, Joint Economic Committee (ed.) Federal Expenditure Policy for Economic Growth and Stability, Washington D. C.: U.S. Government Printing Office, 1957. Growth and Stability, Washington D. C.: U.S. Government Printing Office, 1957.
Sultani, Sheng, Bin. How the opening up of foreign investment in the service sector affects the environmental performance of firms-an experience from China[J]. China Industrial Economics,2021,399(06):61-79.
Tiebout C M. A Pure Theory of Local Expenditures[J]. Journal Political Economy, 1956, 64(5): 416-424.
W.F. Golding.Incentives for change: China’s cadre system applied to water quality Pac. Rim Law Policy J., 20 (2011), pp. 399-428
Xueguang Zhou, Hong Lian, 2020, “Modes of Governance in the Chinese Bureaucracy.” China Journal, Volume 84pp. 51 – 75.

There has been the sub-city environmental protection administrative management system reform views in Shaanxi Province from 2002.
An important reason for Shaanxi Province's implementation of sub-municipal vertical management reform in 2002 was the Han River water pollution incident in 2001, which sparked great concern in society as well as the central government.

Download PDF

Journal Publication

published 07 Feb, 2024

Read the published version in Environmental Science and Pollution Research →

Reviewers agreed at journal
20 Sep, 2023
Reviewers invited by journal
19 Sep, 2023
Editor invited by journal
01 Sep, 2023
Editor assigned by journal
14 Aug, 2023
First submitted to journal
08 Aug, 2023

You are reading this latest preprint version

Pollutants Reduction Effect of Vertical Environmental Reform in China

Status:

Journal Publication

Version 1

Abstract

Figures

1. Introduction

2. Background and Theoretical Hypothesis

2.1. Background: rights of control and choice of governance objectives among government organizations before the VER

2.2. Theoretical analysis: the governance effects of vertical environmental reform

3. Empirical design

3.1. Model

3.2. Variables

3.2.1. Pollutant Emissions

3.2.2. Grassroots environmental enforcement efforts

3.2.3. Control variables

3.3. Data sources and descriptive statistics

4. Empirical analysis

4.1. Pollutant reduction effect

4.2. Mechanism analysis

4.3. Parallel trend test and dynamic effect analysis

4.4. Robustness

4.4.1. Sample selectivity bias mitigation (PSM-DID)

4.4.2. Placebo test

4.4.3. Replacement of the explanatory variable measure

4.5. Heterogeneity analysis

4.5.1. Financial pressure

4.5.2. Government official affiliation

4.5.3. Border contamination

5. Conclusion and discussion

Declarations

References

Footnotes

Status:

Journal Publication

Version 1