Government-Enterprise Collusion and the Effectiveness of Environmental Regulations: Implications for Public Health

Effective environmental management will create a win-win situation for building an ecological civilization with the potential to control the COVID-19 pandemic. From the perspective of government-enterprise collusion (GEC), this study analyzes the moderating effects of the officials' promotion incentives and turnover on the effectiveness of environmental regulations utilizing a panel dataset on 276 cities in China from 2003 to 2019. The study reveals the following empirical results: First, promotion incentives positively moderate the relationship between environmental regulations and environmental pollution, mainly air pollution; results for water pollution are not significant. Compared with general cities, the positive moderating effect of promotion incentives in high-level cities is weaker and the negative moderating effect is more potent. Additionally, the moderating effect of promotion incentives is predominantly positive in the new developmental stage from 2013 to 2019. Second, the negative moderating effect of officials' turnover on the effectiveness of environmental regulations is mainly observed for water pollution but not evident for air pollution. Compared with high-level cities, officials' turnover in general cities is more conducive to the effectiveness of environmental regulations. These findings provide beneficial insights for promoting green growth by improving official governance and destroying GEC.


Introduction
COVID-19 poses a severe threat to public health around the world. Governments have taken measures such as travel restrictions and city blockades to deal with the spread of the pandemic. In addition to spatial strategies such as isolation and interdiction, some studies have found that environmental pollution, especially air pollution, signifcantly impacts the spread of COVID-19 [1][2][3]. Air pollutants (such as PM 2.5 , PM 10 , NO 2 , and SO 2 ) act as carriers to spread the COVID-19 virus and also afect lung health, making those infected more vulnerable to the virus. Terefore, strengthening pollution governance will contribute to alleviating environmental degradation and controlling the COVID-19 pandemic. China has made tremendous economic achievements since it began to open up and reform its economy. However, the ensuing extensive development also brought severe environmental pollution. According to China's 2020 Eco-Environment Bulletin, 43.3% of China's cities (337) exceeded the standard of ambient air quality and 43.6% of the 10,242 groundwater quality monitoring points registered Class V water. Governments at all levels have strengthened the enforcement of environmental regulations; nevertheless, the actual results are not promising [4,5]. Te report "Confessions of the Director of the Environmental Protection Bureau: I am the Director of Public Relations Showmanship" sparked public thinking. According to the report, although the concept of green development continues to gain popularity, some regions still maintain the former development model, in spite of high pollution and high emissions, under the wraps of promotional tournaments and dependence on traditional development models.
Enterprises contribute to tax revenue, while local governments provide shelter for nonclean production. Environmental damage is still tacitly permitted and tolerated, thus creating high growth and heavy pollution concurrently. Western mainstream economic theories cannot efectively explain China's development model of growth and pollution at high levels simultaneously. Some studies advance the theory of government-enterprise collusion (GEC) to integrate growth and pollution into the same analytical framework [6,7].
Te asymmetrical information between the central and local governments provides favorable conditions for collusion, and the ofcials' promotion incentives to pursue shortterm economic growth are undoubtedly an essential driving factor for GEC [7,8]. Local ofcials are desperate to demonstrate their performance and competence during their short-term service under the tenure system, which requires mandatory retirement at a fxed age; conversely, the high turnover rate of local ofcials afects the stability of the collusion network. Evidence shows that local ofcials frequently change, and their terms are generally short in China. Between 2007 and 2017, nearly 80% of municipal party secretaries were reassigned in less than fve years and more than 20% were in ofce for less than a year [9]. Frequent turnover of principal local ofcials undermines the collusion network and creates a deterrent-sensitive period. What are the efects of promotion incentives and ofcials' turnover rates on environmental regulations in the context of collusion enhancement and deterrence perspectives? Are these efects heterogeneous according to pollutant and city class? Te answers to these questions are important to achieve a win-win situation for constructing an ecological civilization and controlling the COVID-19 pandemic.
Te marginal contribution of this study is threefold. First, the study probes the moderating efects of promotion incentives and ofcials' turnover on the relationship between environmental regulations and environmental pollution. To the best of our knowledge, previous research studies focused on the direct efects of promotion incentives and ofcials' turnover on the environmental system, and few studies examined the moderating efects of both. Our analysis incorporates promotion incentives and ofcials' turnover into the efectiveness of environmental regulations from the perspective of GEC and reveals the moderating efects of promotion incentives and ofcials' turnover. Second, the study introduces GEC into the research of environmental regulatory efectiveness, which will expand the application scenario of the theory of GEC and the research perspective of environmental regulatory efectiveness. Previous studies introduced GEC to issues such as high housing prices, food safety, and overcapacity, but few combined GEC and environmental regulatory efectiveness. Based on this research gap, the mechanisms for collusive reinforcement and breaks caused by promotion incentives and ofcials' turnover are discussed and their moderating efects on the efectiveness of environmental regulations are examined. Finally, our research focuses on pollutant heterogeneity. Unlike the existing literature that focused on a single pollutant or pollution composite index, our study includes multiple pollutants to investigate the performance of strategic emission reduction due to pollutant heterogeneity.
Te rest of the study is organized as follows: Section 2 is the literature review. Section 3 presents the theoretical logic and research hypotheses. Section 4 includes the introduction to model construction and variable data description, and Section 5 shows the empirical results and analyses. Te fnal section contains the conclusions, policy implications, and limitations of the study.

Literature Review
Scholars have conducted some research studies on the effectiveness of environmental regulations, which can be divided into two categories according to internal and external research perspectives. Te frst category concerns the impact of the heterogeneity of environmental regulations on technological innovation and environmental improvement. Environmental regulations can be divided into marketincentive environmental regulations (MERs) and command-controlled environmental regulations (CERs). Te former occurs mainly through the environmental protection tax system and emission trading to incentivize each subject to reduce emissions; the latter sets emissions standards and even mandatory shutdowns. Te efectiveness of regulations is also quite diferent due to mechanism and application conditions. Sun et al. [10] found that the incentive efect of MER on innovation in Chinese high-tech enterprises was more signifcant and that fexible regulatory tools and implementation methods were more conducive to encouraging enterprises to carry out innovative activities. Similarly, Fang and Shao [11] stated that MER promoted regional green technology innovation, while CER inhibited it based on local data in China. However, MER is not always optimal but depends on the heterogeneity of regulated objects. For example, Zhu et al. [12] argued that CER and voluntary regulations positively impacted the green transformation of Chinese steel companies, while MER did not. Te second category of studies focuses on the external conditions of environmental policy implementation. Huang and Lei [13] found that the marketization level positively modifed the relationship between environmental regulations and enterprises' green investment, which was more signifcant among nonstate-listed enterprises. Other studies analyzed the fscal decentralization level, industry diferences, and industrial agglomeration [14][15][16].
Local government ofcials are an important external factor infuencing the efectiveness of environmental regulations in China [17]. Seeking political promotion can be considered the most critical career incentive for Chinese ofcials [18,19]. In the context of absolute power held by local governments, the promotion incentives and position turnover of government leaders signifcantly impact regional economic, social, and environmental systems. Scholars have also conducted research studies on the impact of promotion incentives and ofcials' turnover on environmental pollution [20][21][22]. On the whole, there is a consensus that competing for promotions with a one-sided emphasis on economic growth exacerbates environmental pollution.
Furthermore, relevant research studies have explored the mechanisms behind the impact of promotion incentives and ofcials' turnover on environmental pollution, including the perspectives of GEC and anticipatory efects. Yu et al. [23] stated that under China's current ofcial evaluation system, local governments collude with enterprises to achieve economic growth, which may lead to more severe environmental pollution. Song et al. [24] analyzed the efects of the frequency and predictability of ofcials leaving ofce in Chinese cities on air quality in their jurisdictions and found that air pollution spiked in the year prior to the leadership turnover when the incumbent anticipated a change of term. Local leaders are keener on short-term policies to stimulate economic activities. Tey demonstrate their competence to superiors at the expense of air quality when ofcials' turnover is frequent and predictable. From the perspective of GEC, promotion incentives and ofcials' turnover can be organically linked, which provides a valuable reference and inspiration for our research.
Te existing literature has explored the impact of promotion incentives and ofcials' turnover on environmental pollution, but there is still room for discussion about the impact of both on environmental systems. Previous research studies focused on the direct efects of promotion incentives and ofcials' turnover on environmental pollution; however, the moderating efects are not sufciently explored, and research from the perspective of GEC is even scarcer. Environmental regulations are currently the primary means for governments to strengthen environmental governance. Systematically, exploring the efects of promotion incentives and ofcials' turnover on the efectiveness of environmental regulations will facilitate the improvement of environmental governance. Meanwhile, it can also provide a reference for clarifying the mechanisms related to ofcial governance and green development. Our study constructs a theoretical logic that promotion incentives and ofcials' turnover afect the efciency of environmental regulations from the perspective of GEC and investigates the moderating efects involved. Te main work of this study includes the following points: First, based on the perspective of GEC, we analyze the reinforcing and deterring efects of promotion incentives and ofcials' turnover on GEC to establish the moderating mechanisms of promotion incentives and ofcials' turnover on the efcacy of environmental regulations. Second, we examine the impact of environmental regulations on environmental improvement by employing city panel data and using it as a benchmark to explore the moderating efects, focusing on the strategic manifestation of collusion arising from pollutant and city-rank heterogeneity. Ultimately, from the empirical research, we draw policy implications to curb GEC and improve the efciency of environmental governance.

GEC to Distort Environmental Regulations.
Te distortion of environmental regulations caused by GEC can be divided into explicit and implicit manifestations. On the one hand, pollution control and green technology research require large amounts of capital investments, which will increase operating costs, reduce corporate profts, and afect the performance of local economic development. To achieve prosperous economic performance in the short term and especially to boost the economy after the COVID-19 pandemic, local governments lower environmental enforcement standards, relax environmental supervision eforts, and even shelter polluting enterprises during inspections by higher level governments. Tis explicit distortion of environmental regulations by GEC is easy to observe. On the other hand, local governments introduce and develop heavy-polluting industries by lowering environmental requirements, further exacerbating local environmental pressures. Te explicit and implicit distortions of environmental regulations by GEC are intertwined, with local ofcials and enterprises deriving economic performance and operating profts from collusion and regulatory distortions, respectively. Enterprises lose the motivation and incentive to control pollution gradually, and the efectiveness of environmental regulations diminishes signifcantly.

Promotion Incentives to Establish GEC.
Te widespread principal-agent relationships between central and local governments inevitably lead to information asymmetry. Te degree of information asymmetry is likely to deepen, especially in the context of economic decentralization, which has become one of the critical challenges in the modernization of national governance. Although the central government collects local information through various means, such as the establishment of six regional inspection centers for environmental protection in 2006 to alleviate the information asymmetry dilemma, the efect is still limited compared to the massive amount of information in economic and social systems. GEC is easy to establish in a situation where the central or higher level government does not have complete access to local information.
Asymmetric information prevents the central government from comprehensively understanding all local government behaviors. To avoid moral hazards, the central government has developed diferent assessment systems for local ofcials in various stages of development. Te bias of assessment systems on the gross domestic product (GDP) and fscal tax revenue indicators profoundly impacts the degree of GEC. Te ofcial appraisal system since 1949 can be roughly divided into three stages (1949-1978, 1978-2002, and 2002-2019). Before China's economic reform and opening up, to consolidate the political order and the needs of a highly centralized planned economy, the main criterion for assessing ofcials was political loyalty. From the reform era to the beginning of the 21st century, however, economic development became the focus of work, and economic assessment was the principal criterion at this stage. With the 16th Congress of the Communist Party of China (CPC) in 2002 as the node, the ofcials' appraisal criteria gradually turned to a multidimensional appraisal with comprehensive economic-environmental-social coordination [18]. Although the ofcial evaluation criterion is gradually reversing the drawback of focusing purely on the GDP, local ofcials still regard economic performance as a bargaining chip for Journal of Environmental and Public Health political promotion due to the high visibility of economic indicators and the lagging efect of environmental governance [25,26]. In particular, unlike the promotion of highlevel ofcials, which is infuenced by noneconomic factors such as "nepotism," "political afliation," and "loyalty" [27], the promotion of low-level ofcials is highly linked to economic performance.
Tere is a strict retirement policy in China's administrative system: 60 years for prefecture-level municipal ofcials and 65 for provincial ministerial-level ofcials. In 1986, the CPC Central Committee put forward the reform of rejuvenating the leadership team. Terefore, the age of offcials has become an important consideration for political promotion. Further analysis reveals that the probability of promotion is inversely related to the length of tenure [25] and promotion incentives motivate local ofcials to attain economic achievements under limited terms. Terefore, this study proposes Hypothesis 1: H1: In the context of perverse incentive structures and asymmetric information, promotion incentives will positively moderate the relationship between environmental regulations and environmental pollution, which is detrimental to environmental improvement.

Ofcials' Turnover to Deter GEC.
Information asymmetry of multiple dimensions often exists in transition economies. In addition to information asymmetry between the central and local governments, there is also information asymmetry between enterprises and local governments. China's traditional government-enterprise relationship has been characterized by dependency for thousands of years. Under such circumstances, enterprises are always incentivized to establish stable relationships with governments to obtain more resources efciently. According to the survey of more than 12,000 enterprises in mainland China conducted by the World Bank in 2005, each enterprise spent an average of nearly 60 days a year, dealing with government core departments, and the top 5% of enterprises spent up to 170 days. Although diferent levels of government have implemented a series of initiatives to streamline administration and delegate powers since 2012, there are still some areas where "time tax" remains high for enterprises. If an ofcial serves in a city long term, collusion is more likely to occur because it is easier to manipulate local regulators [28]. Te exchange of ofcials has been incorporated as a formal system in the ofcials' governance since 1990. Under the current system, the transfer of ofcials does not require strict adherence to the tenure of the local leadership team but shows the characteristics of "fexible terms of ofce and transfers at any time." Frequent ofcials' turnover can break the stable collusive relationship [29] and create a politically sensitive period. Te deterrent efect and uncertainty will be more substantial, especially when ofcials fall for corruption. Such collusive shocks and uncertainty can profoundly afect enterprises' behaviors [30,31]. When corruption is significant in the environmental feld, political instability caused by the ofcials' turnover can positively impact the stringency of environmental regulations. Corruption weakens the seriousness of environmental regulations, but this efect disappears as political instability increases [32]. Terefore, the aforementioned discussion leads to the development of Hypothesis 2 which is as follows.
H2: Ofcials' turnover has a positive efect on breaking the collusion network, thus forming a politically sensitive period. Pollution will be curbed, and environmental regulations will be enforced. Consequently, ofcials' turnover will negatively moderate the relationship between environmental regulations and pollution, thereby contributing to environmental improvement.
Utilizing the two aforementioned hypotheses, the logical framework of this study is illustrated in Figure 1

Independent Variable. Environmental regulation (ER)
is the independent variable of this study. Tere is no unifed standard for measuring environmental regulations, and related research studies use the pollutant emission compliance rate or pollution control input costs [33,34]. Tese approaches can partly refect the intensity of regional environmental regulations, but they also sufer from endogeneity. Chen et al. [35] proposed a method to use the frequency of environment-related terms (i.e., PM 2.5 , PM 10 , CO 2 , SO 2 , air, pollution, ecology, green, emissions, energy consumption, low carbon, environmental protection, emissions reduction, and chemical oxygen demand) in government work reports to measure environmental regulations. Tis method has two advantages: First, environment-related terms cover a wider range that more comprehensively measures regional environmental management eforts. Second, government work reports are usually released at the beginning of the year, and the following economic activities do not have a reverse impact, which can mitigate endogenous problems to a certain extent. Due to these advantages, our study draws on this method.

Moderator Variables.
Ofcials' promotion incentives (Age) are one of the moderator variables. Te level of government in China is positively correlated with the division of labor between the party committee and the government. Party secretaries have become the de-facto core of decision-making. An ofcial's age is an essential factor for measuring promotion incentives. According to the retirement age limit of 60 years for the leading positions of prefecture-level municipalities, 55 years is generally a turning point for the strength of promotion incentives. Party secretaries are high-incentive ofcials if they are under 55 when they take ofce and low-incentive ofcials if they are 55 and older [36]. Following this logic, promotion incentives are characterized based on the party secretary's age when taking ofce, denoted as Age. Age is recorded as 1 when the party secretary takes ofce at an age less than 55 and 0 when it is greater than or equal to 55.
Ofcials' turnover (Turnover) is another moderator variable. According to the existing method, turnover is equal to 1 if replacement happens and 0 otherwise. If replacement happens before June 30, we set that year as turnover and the following year as otherwise.

Control
Variables. In addition to environmental regulations, other factors may also afect environmental improvement. We adopt the following control variables to eliminate the efects of noncritical factors [37]: (i) City scale (CityScale): the reasonable city scale can reduce pollution emissions and improve environmental governance by enhancing the division of labor [38]. We use the total population of a municipal district as a proxy for the city scale. (ii) Technology level (Technology): the level of technology is an essential factor for determining pollutant emissions. It is expressed by the number of green invention patent applications in each city. (iii) Industrial structure (Industry): the secondary sector, specifcally industry, is the principal source of pollutant emissions. Te industrial structure is measured by the share of industrial value added in the city's GDP. (iv) Opening-up level (Open): the higher the level of openness in a region, the more deeply it can participate in the division of labor and absorb advanced external technology efciently. We employ the actual amount of foreign investment used as a proxy for the opening-up level.
(v) Economic development (Economy): economic development and pollution are closely linked. We use the per capita GDP to measure a region's economic development level.

Model Specifcation.
To capture the moderating efect of promotion incentives on the efectiveness of environmental regulations, we constructed the following benchmark model with the example of wastewater emissions as where i and t denote the city and year, respectively. β 0 is the constant term, and the coefcient of β 1 indicates the impact of environmental regulations on environmental improvement. β j denotes the regression coefcient for each control variable. Finally, η i and μ it represent the city efect and the random error term, respectively. Te moderator variables are centralized; subsequently, the interaction terms of environmental regulations with promotion incentives and ofcials' turnover are introduced based on the benchmark model, as shown in Equations (2) and (3). Te coefcient of α 1 in Equations (2)  Journal of Environmental and Public Health the interaction term is signifcant, it suggests that moderating efects exist. Table 1 reports the results of descriptive statistics for variables. Te emissions of various pollutants and the level of environmental regulations vary greatly due to diferent stages of development. Figure 2 depicts the age of party secretaries taking ofce and the corresponding number of secretaries. Findings show that the large number of party secretaries takes ofce at age 53, a total of 148, accounting for 11.41% of the total sample. Figure 3 depicts the number of cities with party secretary turnovers, and a total of 1,296 turnovers occurred within the sample.  Table 3 Te underidentifcation test and the weak IV test indicate that the IV is reasonable. Comparing the results in Table 2, the estimated coefcients of environmental regulations are all correspondingly higher (β 1 = −0.281, p < 0.01; β 1 = −0.477, p < 0.01; and β 1 = −0.442, p < 0.01, respectively) after controlling endogeneity, indicating that the benchmark regressions mixed with endogeneity tend to underestimate the improvement efect of environmental regulations. In summary, after considering endogeneity, the direction and signifcance of the impact remain unchanged although the value fuctuates to some extent. Consequently, the benchmark regression results are robust and reliable on the whole.

Summary Statistics.
(2) Substitution of Dependent Variables. By replacing the dependent variables in the benchmark regression from pollutant emission intensity to pollutant emissions, the corresponding regression results are reported in Table 4.

Te Moderating Efect of Promotion Incentives.
Teoretically, promotion incentives may have a moderating efect on the efectiveness of environmental regulations through GEC. To prove Hypothesis 1, we obtain the relevant regression results through Equation (2). To verify the robustness of the regression results, tenure is included in delineating the strength of promotion incentives. Te study examined two scenarios that included 3-year tenure and 5year tenure and one that did not include the year-by-year promotion incentive delineation criteria. Specifcally, columns (2) and (3), columns (5) and (6), and columns (8) and (9) of Table 5 show that the coefcients of interaction terms in both soot and SO 2 are positive and signifcant (α 1 � 0.053, p < 0.05; α 1 � 0.044, p < 0.05; α 1 � 0.036, p < 0.01; α 1 � 0.019, p < 0.05; α 1 � 0.044, p < 0.01; and α 1 � 0.038, p < 0.01, respectively). Promotion incentives have a positive moderating efect on the relationship between environmental regulations and air pollution. Since pollutant emission intensity is an inverse indicator, high promotion incentives weaken the efectiveness of environmental regulations for both categories of air pollutants in all three classifcation cases. In contrast, the positive moderating efect of promotion incentives is insignifcant in wastewater and even shows a negative moderating efect (α 1 � −0.012, p < 0.10) in the 5-year tenure criterion. Ulteriorly, it reveals that the distorting efect of promotion incentives on environmental regulations is more likely to occur in the area of air pollutants instead of water pollution from the perspective of GEC. Tus, Hypothesis 1 is partially verifed because it does not consider pollutant heterogeneity. Te heterogeneity in the moderating efect of promotion incentives on diferent pollutants may be associated with the variation in pollutant spillover. Compared with wastewater, air pollutants have a substantial spillover. Especially in the   Note. Te numbers in brackets stand for t-statistics; * , * * , and * * * indicate signifcance at the level of 10%, 5%, and 1%, respectively, which are the same in the following tables except Table 3. Note. Te numbers in brackets stand for z-statistics; the numbers in the square bracket stand for p value; * , * * , and * * * indicate signifcance at the level of 10%, 5%, and 1%, respectively.     Journal of Environmental and Public Health research sample where prefecture-level cities dominate, the spillover is more signifcant due to the small area of jurisdiction. Moreover, air pollutants are more likely diluted and difused by the wind. Terefore, the benefts and costs of air pollution control will be uncertain if relevant systems and collaborative governance mechanisms are not sound. Air pollutant emissions are prone to resulting in beggar-thyneighbor situations and "race to the bottom" competitions. Te spillover of wastewater is weak, and the boundary is clearly defned. With the implementation of the "river chief system," the four-level river chief system (i.e., province, city, county, and township) in China has been established; the river chief system is also maintained by the key leaders of local governments. Te river chief system adds political factors to water pollution control and promotes "yardstick competition" among regions. In the context of gradually increasing environmental performance appraisal, strengthening water pollution control can also give local ofcials a green label and increase promotion leverage to some extent.
To mitigate side efects caused by focusing on economic growth alone, the evaluation of ofcials has moved in a comprehensive and coordinated multidimensional direction. Te central government has placed considerable value on green development since the 18th CPC National Congress in 2012, and the proportion of environmental indicators used in evaluating ofcials has also gradually increased. Terefore, taking 2012 as the node, we examine whether the moderating efect of promotion incentives changes by dividing the whole sample into two periods: 2003-2012 and 2013-2019. To save space, we only report the coefcients of interaction terms in the moderating efect model in Table 6 (we do the same in Tables 7 and 8). Te comparison between the two diferent periods shows that the positive moderating efect of promotion incentives on air pollutants has increased numerically and signifcantly in varying degrees from 2013 to 2019, which implies that the ofcial governance has not efectively reversed the trend of GEC in air pollutants. Te moderating efect of promotion incentives on water pollution does not change signifcantly in the two periods, and neither period reaches statistical signifcance.
Although the environmental assessment indicators for ofcials are gradually increasing, it is awkward that the binding force is still limited. Environmental governance requires long-term investment and has a signifcant lag in efectiveness. It may be difcult to see the results of environmental governance within limited tenure, but it increases the performance of the next ofcial. Terefore, local ofcials do not have sufcient incentives to invest in environmental governance. Furthermore, compared with the direct positive incentive of economic growth on ofcials' promotion, the current environmental appraisal is mainly in the form of "punishment," which means that local ofcials will be held accountable only when major environmental accidents occur in their jurisdictions. We also notice that anticorruption eforts have increased unprecedentedly [41] and corruption behaviors have been efectively deterred since 2012. However, the empirical results do not show that promotion incentives have a weakened positive moderating efect on the efectiveness of environmental regulations. Tis is primarily because the concept of GEC used in this study contains not only corrupt collusion but also the complicity of local governments in allowing enterprises to adopt an unclean production mode for economic growth [42]. Te latter, which accounts for the majority, is mainly caused by an extensive development concept without corruption and beneft exchange.
Signifcant diferences exist between the city level and developmental stage among cities (municipalities, provincial capital cities, independent planning cities, and ordinary prefecture-level cities). Terefore, there should be heterogeneity in the moderating efect of promotion incentives among diferent city levels. Te sample of 276 cities is divided into 243 ordinary prefecture-level cities (abbreviated as general cities) and 33 municipalities, provincial capital cities, and independent planning cities (abbreviated as highlevel cities) to explore heterogeneity. Te regression results for subsamples are presented in Table 7. Specifcally, in the case of water pollution, columns (1), (4), and (7) show that the coefcients of the interaction terms are small and do not reach statistical signifcance in 243 general cities. In highlevel cities, the coefcients are signifcantly negative (α 1 � −0.044, p < 0.01 and α 1 � −0.023, p < 0.10) in columns (4) and (7), except at the 3-year tenure criterion. In the case of soot, columns (2), (5), and (8) show that the coefcients of the interaction terms are signifcantly positive (α 1 � 0.046, p < 0.05; α 1 � 0.044, p < 0.01; and α 1 � 0.052, p < 0.01, respectively) in general cities; in high-level cities, it is only signifcant at the 10% level (α 1 � 0.105, p < 0.10) with the 3year term criterion in column (2). In terms of SO 2 , the moderating efect of promotion incentives is not signifcantly diferent between general and high-level cities.
Overall, promotion incentives in high-level cities have a more substantial negative moderating efect and a weaker positive moderating efect on the relationship between environmental regulations and environmental pollution than in general cities. Several reasons can account for this phenomenon. On the one hand, high-level cities have more stringent environmental standards and higher public demand for the environment. Moreover, compared with general cities, high-level cities are already at a more advanced economic developmental stage and have a weaker motivation to develop the economy by destroying the environment. On the other hand, high-level cities experience a more robust legal system and vigorous public scrutiny, which increase the cost of collusion and the chances of exposure accordingly. As a result, high-level cities tend to be more cautious about collusion.

Te Moderating Efect of Ofcials' Turnover.
Frequent turnover of principal local ofcials will disrupt the existing collusion network and create a sensitive deterrent period, thus restraining the behavior of all parties. As a result, ofcials' turnover may negatively moderate the relationship between environmental regulations and pollution, thereby contributing to environmental improvement. Te     Table 9 are obtained by Equation (3) to test Hypothesis 2. Considering the lagged efect of ofcials' turnover, we take a one-period lag of ofcials' turnover. Moreover, according to Figure 3, the years (2003, 2008, 2013, and 2017) are the peak years for turnover, and the sample of these 4 years is excluded to reduce the impact of predictable turnover; corresponding results are shown in columns (4)-(6) of Table 9. From the full-sample regressions in columns (1)-(3), the fndings show that similar to the moderating heterogeneity of promotion incentives, the moderating efect of ofcials' turnover is also signifcantly heterogeneous across pollutants. From the results presented in column (1), the coefcient of the interaction term in wastewater is negative and signifcant at the 5% level (α 1 � −0.018,p < 0.05), indicating that ofcials' turnover has a negative moderating efect on the relationship between environmental regulations and water pollution. Consequently, it further suggests that ofcials' turnover has a deterrent efect on water pollution collusion, which will facilitate the efectiveness of environmental regulations. In the area of air pollutants, soot, and SO 2 , the moderating efect of ofcials' turnover is insignifcant. Although the coefcient of the interaction term in SO 2 is negative (α 1 � −0.014), it does not reach statistical signifcance. After excluding the peak years of turnover, the signifcance of the estimated coefcients does not change compared to that of the whole sample, with only minor fuctuations in values.
Compared to wastewater, air pollutants have a high spillover and boundaries are more blurred. Even though ofcials' turnover can create shock damage to collusion, the deterrent efect on air pollutants is weak. Hypothesis 2 is also partially verifed.
Following the idea that the moderating efect of promotion incentives is heterogeneous under diferent city levels, is there also city-rank heterogeneity in the moderating efect of ofcials' turnover? Terefore, based on the classifcation of the city rank mentioned above, this study examines diferences in the moderating efects of ofcials' turnover under city-rank heterogeneity, and the results are presented in Table 8. Columns (1) and (3) show that the coefcients of the interaction terms are signifcantly negative (α 1 � −0.025, p < 0.01 and α 1 � −0.019, p < 0.10), which means that ofcials' turnover negatively moderates the intensity of wastewater and SO 2 emissions in general cities. By contrast, only one coefcient of the interaction term is negative (α 1 � −0.007) in the intensity of wastewater emissions in high-level cities in column (1), but it does not reach statistical signifcance. Moreover, from columns (4)-(6), after excluding the peak years of turnover, the direction of the moderating efect does not change, with only minor fuctuations in values and signifcance.
Compared with high-level cities, the collusion deterrent efect caused by ofcials' turnover is more signifcant in improving the efectiveness of environmental regulations in general cities. General cities still lag in terms of the development stage, legal environment, and governmententerprise relationship, and enterprises depend more on local governments. Terefore, the efect of fractured collusion networks caused by ofcial mobility can be more profound in general cities.

Conclusions and Policy Recommendations
6.1. Conclusions. Environmental pollution limits the sustainability of economic development and signifcantly impacts the transmission capacity and pathogenicity of COVID-19. Terefore, efective environmental management will create a win-win situation for the ecological civilization and COVID-19 pandemic control. Based on the novel perspective of GEC, this study constructs a theoretical logic framework to explore the moderating efects of ofcials' promotion incentives and turnover on the efectiveness of    Second, the positive moderating efect of promotion incentives on the efectiveness of environmental regulations is refected in air pollutants but is not conspicuous in water pollution. In the new developmental stage (2013-2019), the binding force of the ofcials' environmental appraisal is still limited, so the moderating efect of promotion incentives remains predominantly positive. In terms of city-level heterogeneity, compared with general cities, promotion incentives in high-level cities have a weaker positive but a more potent negative moderating efect. Tird, the collusive deterrent efect of ofcials' turnover negatively moderates the efectiveness of environmental regulations mainly in water pollution rather than air pollution. Compared with high-level cities, ofcials' turnover in general cities is more conducive to the efectiveness of environmental regulations.

Policy Recommendations.
Strengthening pollution management and implementing green development are sufcient guarantees to coordinate COVID-19 controls and social development. Te economic development model is the internal logic behind the moderating efects of promotion incentives and ofcial turnover on the efciency of environmental regulations. Transition economies represented by China have abundant labor resources but generally weaker social institutions and high transaction costs, so they rely on massive resource inputs to drive economic growth in the early developmental stages. GEC is the primary institutional driver of extensive development, which satisfes the parties' frenzied pursuit of economic growth in the short term. Te extensive development model, characterized by high input, heavy pollution, and low efciency, is unsustainable with the disappearing demographic dividend, tightening resource constraints, and rising transaction costs. Terefore, it is necessary to gradually improve the system construction, cut of the chain of GEC, and release the innovative vitality of enterprises to make the transformation into an intensive development model [43]. Otherwise, it is straightforward to fall into the middle-income trap. To break away from the sloppy development model, the following policy recommendations are also essential for transition economies with national conditions similar to those in China: (1) Scientifc resumption of production: Owing to the severe economic impact of COVID-19, governments are prone to boosting the economy in the short term by relaxing regulatory policies, which may lead to a high incidence of GEC. Terefore, governments at all levels should frmly establish the concept of green development and scientifcally organize the resumption of work and production; otherwise, the stimulation of the economy at the expense of the environment may lay a hidden danger for the spread of the COVID-19 virus.
(2) Rethink the evaluation mechanism of ofcials: Te evaluation mechanism determines the behavior of local ofcials. In addition to increasing the weight of environmental indicators, the central government should innovate and adjust ofcials' evaluation methods, which are invariably in the form of "punishment." Dredging environmental pressure transmission can reverse local ofcials' one-sided pursuit of GDP and shape the ofcial governance model of "growth for harmony." (3) Accelerate the formation of regional joint governance mechanisms: For pollutants with a high spillover, the regional-coordinated governance mechanism can be formed by strengthening the vertical management of central-level or upper-level governments. Te externality of pollution is gradually internalized to reduce the race to the bottom by forming a regional coordination governance mechanism. (4) Clarify the division of responsibilities: Te "lifelong responsibility system for environmental damage" restrains wanton pollution behaviors under promotion incentives by setting policy red lines. However, the system still needs a more precise division of responsibilities and a perfect procedure to achieve realistic and lifelong responsibility. (5) Strengthen the legal system: While the ofcials' turnover may have a deterrent efect on collusion, new collusion networks can quickly emerge. Terefore, relying on the ofcials' turnover to improve the environment is not a long-term solution. A more sustainable plan is to improve the legal system and build a new type of proclear governmententerprise relationship.

Limitations and Future Research Directions.
Te limitations of this study point the way to future research, and three limitations stand out. First, the robustness of the empirical results can be further verifed using additional measures of promotion incentives as well as mayoral data.
In addition to the age factor, external factors such as GDP growth rates, fscal surplus, and unemployment rates also measure promotion incentives in diferent dimensions.
Mayors are an important faction that infuences the regional economy, so the robustness of the results can be further tested by using multiple promotion incentive measures and combining them with mayoral data. Second, the impact of heterogeneity in the types of ofcials' turnover on collusion deterrence needs to be focused on. Ofcials' turnover includes retirement, promotion, reassignment, and corruption, which may cause diferences in the damage degree of GEC. Terefore, future research studies should probe the impact of this variability by grouping or setting up dummy variables. Tird, more pollutants need to be included in future research studies. Integrating data availability and workload, three representative pollutants (wastewater, soot, and SO 2 ) were included in this study. Solid waste with lower spillover and nitrogen oxides in air pollutants can be introduced to further verify the robustness of the relevant results.

Data Availability
Te original contributions presented in the study are included in the article/supplementary material, and further inquiries can be directed to the corresponding authors.

Conflicts of Interest
Te authors declare that they have no conficts of interest..