Next Article in Journal
Tourism Investment Gaps in Poland
Next Article in Special Issue
Coupling of Rural Energy Structure and Straw Utilization: Based on Cases in Hebei, China
Previous Article in Journal
Environmental Assessment of Two Use Cycles of Recycled Aggregate Concrete
Previous Article in Special Issue
Determinants of Residents’ Willingness to Accept and Their Levels for Ecological Conservation in Ganjiang River Basin, China: An Empirical Analysis of Survey Data for 677 Households
Open AccessArticle

Does China’s Pollution Levy Standards Reform Promote Green Growth?

by Zhengge Tu 1, Tao Zhou 1 and Ning Zhang 2,3,*
1
School of Economics and Business Management, Central China Normal University, Wuhan 430079, China
2
Institute of Blue and Green Development, Shandong University, Weihai 264200, China
3
Department of Economics, Jinan University, Guangzhou 510000, China
*
Author to whom correspondence should be addressed.
Sustainability 2019, 11(21), 6186; https://doi.org/10.3390/su11216186
Received: 9 October 2019 / Revised: 26 October 2019 / Accepted: 29 October 2019 / Published: 5 November 2019

Abstract

Estimating the impact of environmental taxes on economic output is of great theoretical value for promoting green growth in China. Using a dataset of 232 cities from 2004 to 2014, this paper investigates the effect of pollution levy standards reform (PSR) on green total factor productivity (GTFP). We employ directional distance functions (DDF) computed by data envelopment analysis (DEA) to derive GTFP based on the Malmquist–Luenberger (ML) productivity index. Then, we investigate the impacts of PSR on China’s GTFP using Difference-in-Differences (DID) estimation. The results reveal that PSR has an inhibitory effect on GTFP, via the mechanism of technological change. Furthermore, PSR has heterogeneous impacts on different city types. The results indicate that PSR statistically significantly reduces GTFP in key environmental protection cities (KEPCs), large cities, and eastern cities, but that it has less impact on non-KEPCs, small/medium cities, megacities, and cities in central areas.
Keywords: pollution levy standards reform; green total factor productivity; DEA; Difference-in-Differences pollution levy standards reform; green total factor productivity; DEA; Difference-in-Differences

1. Introduction

Over the past three decades, China’s economy has developed rapidly and has had remarkable achievements in many fields. It is now the world’s second-largest economy. However, the long-term economic growth created by an increase in factor inputs and the expansion of scales of production has brought about serious environmental problems, notably air pollution. In the 2008 Global Environmental Performance Index jointly published by Yale University and the World Economic Forum, China ranked 177th out of 180 countries and regions in air quality. In 2017, 239 of 338, or 70.7%, of Chinese cities exceeded air quality standards. A total of 36.1% of 463 Chinese cities with precipitation monitoring experienced acid rain [1]. This terrible environmental pollution has seriously weakened residents’ health, decreased regional economic operational efficiency, and threatened the quality of the nation’s economic development.
Faced with this serious environmental pollution, the Chinese government has undertaken a series of environmental protection policies. Looking back at Chinese environmental protection policies since the reforms and opening up in 1978, the Chinese government has primarily focused on implementing command-and-control environmental regulations. In the 21st century, market-based environmental protection policies have gradually emerged. The implementation of a pilot SO2 emission trading policy in 2002 showed that China had begun to use market-based environmental protection policies to solve environmental problems. Environmental tax reforms are an important class of market-based environmental protection policies that have played an important role in promoting the coordinated development of the environment and the economy in recent years. However, many people are concerned that the government’s strict environmental regulations may slow down China’s economic growth. In theory, environmental regulations may impose additional emissions costs on companies, thereby reducing their productivity and market competitiveness. Some studies have found that the U.S. Clean Air Act, enacted in 1970, caused high structural unemployment in pollution-intensive industrial enterprises and a decline in capital stocks, economic growth rates, and total factor productivity (TFP) [2,3]. Gray and Shadbegian [4] analyzed the relationship between productivity, pollution abatement expenditures, and other measures of environmental regulation in plants across three industries, and they found that more-regulated plants had significantly lower productivity levels, and slower productivity growth rates than less-regulated plants. Compared with command-control environmental regulation, market-based measures that increased coal prices could effectively reduce coal usage and air pollution in India, but also hindered the entry of new enterprises and forced them to withdraw from the market [5]. However, the Porter Hypothesis argues that appropriate and strict environmental regulations can spur innovation, which may in turn increase firm productivity and market competitiveness [6]. Flexible environmental regulations could weaken the mediating effects of technological innovation on the relationship between environmental regulation and business performance. They could also mitigate the negative impact of environmental regulation on both technological innovation and business performance [7]. In recent years, there have been many supporters of the Porter Hypothesis [8,9,10,11].
However, the results of the research on different environmental policies may not be consistent [12]. At present, most of the literature focuses on the impact of command-and-control environmental policies on economic growth. The acid rain and SO2 pollution control zone policies (also known as the “two control zones”), which were implemented in 1998, are the most powerful command-and-control environmental regulations in China at present. Studies have shown that the “two control zones” improved the profits and product conversion rate of export enterprises. They also promoted the TFP of pollution-intensive industrial enterprises by optimizing their industrial structure, upgrading clean technology, and eliminating high-polluting and inefficient enterprises [10,13,14,15]. In addition to the “two control zones,” Li and Chen [16] found that the Revision of Air Pollution Prevention and Control Law (APPCL2000) significantly improved the TFP of industrial sectors that created intensive air pollution. Long and Wan [17] found that the implementation of clean production standards significantly increased enterprise profitability, but it did not promote corporate innovation or subsidies.
There is also some literature on the economic effects of market-based environmental protection policies. Some studies have found that the EU’s carbon trading system has not significantly affected the income or employment of German enterprises [18]. However, some studies claim that the environmental policy represented by the emissions trading mechanism could produce huge economic dividends [19,20,21,22]. In China, Li and Shen [23] found that the emission trading system implemented in 2002 not only failed to reduce pollution, but that it caused even more pollution in the pilot areas. Tu and Shen [24] found that China’s SO2 emissions trading pilot in 2002 did not increase total industrial output in the short or long term, and it also failed to reduce pollution abatement costs [25]. Tang et al. [26] pointed out that the impact of carbon emission trading policies on economic output depended on the carbon emission authorities’ allocation mechanism. At the enterprise level, Ren et al. [27] and Qi et al. [28] found that the SO2 emissions trading pilot in 2002 significantly improved corporate TFP and green innovation. While environmental tax reform is an important approach towards market-based environmental regulations, it has not been extensively studied in the current literature. Some studies have found that pollution discharge fees cannot fundamentally solve the problem of pollution [29]. In the long term, enterprises will increase their investment in technological innovation in order to improve enterprise productivity levers. Generally, after a one-time investment in environmental protection, enterprises are exempt from paying fees or economic penalties for excessive discharge. This is because their pollution emissions meet national and local environmental standards, which reduces the economic burden on the enterprises. Zhang et al. [30] found that the expected effect of the Regulations on the Administration of the Collection and Use of Pollution Discharge Fees was not satisfactory, because it fundamentally failed to reduce pollution in China. Guo et al. [31] found that the SO2 pollution levy standards reform (PSR) significantly reduced industrial SO2 emissions. Lu et al. [32] and Li et al. [33] found that the PSR constrained economic growth.
Although there is a growing amount of literature on environmental regulation, there is little research on environmental tax reform. There have been few investigations in the existing literature on the impact of PSR on green total factor productivity (GTFP), especially in China. To fill this gap, this study sets out to examine whether PSR results in positive changes in regional GTFP. First, unlike Lu et al. and Li et al., who studied the impact of PSR on environmental efficiency [32,33], this paper uses directional distance function (DDF) to calculate Malmquist–Luenberger index (ML index), and study the impact of PSR on GTFP. It also discusses the direct mechanism through which PSR affects GTFP, to discern if the mechanism works through changes in efficiency or technology. Furthermore, this paper uses PSR as a quasi-natural experiment to re-examine the Porter Hypothesis and explore its mechanisms. This is the first paper to study the direct link between PSR and GTFP. Making use of Difference-in-Differences (DID) analysis, we identify the causal effects that PSR has on GTFP. By comparing the treatment and the control groups, we can better control for the effects of observable and unobservable factors, and thus identify the impact of PSR on GTFP. Finally, this paper is also significant as a reference for the recently implemented environmental protection tax.
This study is structured as follows. The second section is a brief description of the SO2 levy standards reform in China. The third section focuses on the data description, variables selection and empirical strategy, which includes the DEA model to measure GTFP, and the DID strategy to analyze how PSR affects GTFP. The fourth section presents the empirical results. The last section represents conclusions derived from the presented research, and some policy implications can also be proposed from the empirical results.

2. A Brief Description of SO2 Levy Standards Reform in China

Most research agrees that environmental pollution derives mainly from the externalities of economic behavior [34]; therefore, internalizing the cost of environmental pollution is the best way to solve it. With this theory in mind, countries around the world have responded by levying taxes on polluters. China’s pollution levy system was first mentioned in the Report on Environmental Protection Work in 1978. Furthermore, in 1979, the Environmental Protection Law (Enforcement) clearly stipulated that pollutant discharges exceeding national standards would be fined according to their concentration and quantity. Since then, a pollution charge system has gradually been developed in various provinces. The Administration for Levy and Use of Pollution Discharge Fees promulgated by the State Council in 2003 has made major adjustments to many aspects, including levy objects, levy standards, management and use of pollution discharge fees, and the total charge system clarification [35]. Subsequently, Management Measures for Levy [36] and Use of Pollutant Discharge Fees [37] were promulgated, and the system of Pollutant Discharge Fees was comprehensively and systematically established in China [38]. Although the Administration for Levy and Use of Pollution Discharge Fees raised the SO2 levy standard from 0.2 RMB/kg to 0.63 RMB/kg, and the levy range was extended from the two control zones to the entire country, China’s current pollution levy standards are still very low. Therefore, polluters would rather pay the pollution fees than tackle pollution itself [39]. In response to this problem, the State Council issued the Comprehensive Work Plan for Energy Conservation and Emission Reduction [40] in May 2007, requiring all provinces to raise their SO2 levy standards, doubling the SO2 discharge fee from 0.63 RMB/kg to 1.26 RMB/kg. After the issuance of this regulation, provinces actively adjusted their SO2 levy standards. Moreover, the Notice on Adjusting the Levy Standard of Pollution Discharge Fee [41] issued in September 2014 required all provinces to adjust their SO2 levy standards to no less than 1.2 RMB/kg by the end of June 2015. Therefore, other provinces that had not yet changed their SO2 emission fee adjusted their SO2 levy standards before 2015.
This study collected detailed information on the reform of SO2 levy standards for each province from 2007 to 2014, as shown in Table 1 (The provinces of Shanxi and Heilongjiang stipulated that only enterprises that have not completed the construction of flue gas desulfurization facilities, or whose SO2 emissions exceeded standards, should adjust their SO2 levy standards. Therefore, this study did not include these two provinces in the treatment group). It considers the adjustment of SO2 levy standards as a quasi-natural experiment, placing the provinces shown in Table 1 into the treatment group by the date on which they adjusted their SO2 discharge fees. This study used the DID method to evaluate the impact of improvements in SO2 levy standards on regional GTFP.

3. Data and Methods

3.1. Green Total Factor Productivity

This study used the DDF and ML productivity index to measure the GTFP of cities, estimating the effect of PSR on the GTFP of cities using the DID method. The purpose of using the DDF is to reduce the pollution emissions (in this study, industrial SO2 emissions) while meeting the need for output growth [42,43]. The expression is as follows:
D 0 t y t , x t , b t ; g y ,   g b   =   s u p [ β : y t + β g y , b t β g b p t x t ] ,
where g = (gy,gb) is a set direction vector; x is an input vector; y is a “desirable” output vector generally referring to economic growth; t represents the year; b is an “undesirable” output (industrial SO2 emissions), and P(x) is the feasible output set (for both “desirable” output y and “undesirable” output b) for the given input vector x. β represents the maximum possible quantity of “desirable” output increase and “undesirable ” output decrease.
Färe et al. [44], Tu [42], and Wang et al. [43] set the direction vector as gt = (yt,−bt), with the mathematical programming expression to compute the DDF given by:
D 0 t y k t , x k t , b k t ; y k t , b k t   =   m a x β ,
s . t . k   =   1 K z k y k , m t 1 + β y k , m t , m   =   1 , , M k   =   1 K z k b k , j t   =   1 β b k , j t , j   =   1 , , J k   =   1 K z k x k , n t x k , n t , n   =   1 , , N z k 0 , k   =   1 , , K .
The non-parametric linear programming technique is used to compute the DDFs. D 0 t y k t , x k t , b k t ; y k t ,   b k t is the distance between the specific region and the ‘meta’ best-practice frontier (the regions with the largest output and the fewest pollution emission under a specific technical structure and factor input) in a certain period. If the value of the DDF is zero, the city’s production is technically efficient; otherwise, it is inefficient. We can construct a TFP index on that basis. Chung et al. [45] define the ML index of productivity between period t and t + 1 as:
M L t t + 1   =   1 + D 0 t x t , y t , b t ; g t 1 + D 0 t x t + 1 , y t + 1 , b t + 1 ; g t + 1 × 1 + D 0 t + 1 x t , y t , b t ; g t 1 + D 0 t + 1 x t + 1 , y t + 1 , b t + 1 ; g t + 1 1 2
The ML index can be decomposed into efficiency change (EFFCH) and technological progress (TECH).
E F F C H t t + 1   =   1 + D 0 t x t , y t , b t ; g t 1 + D 0 t + 1 x t + 1 , y t + 1 , b t + 1 ; g t + 1
T E C H t t + 1   =   1 + D 0 t + 1 x t , y t , b t ; g t 1 + D 0 t x t , y t , b t ; g t × 1 + D 0 t + 1 x t + 1 , y t + 1 , b t + 1 ; g t + 1 1 + D 0 t x t + 1 , y t + 1 , b t + 1 ; g t + 1 1 2
If there were no changes to the inputs and outputs for two time periods, then ML = 1. If there has been an increase in productivity, then ML > 1, while a decrease in productivity means ML < 1. EFFCH in Equation (5) indicates the change in output caused by a change in production efficiency. EFFCH > 1 indicates that efficiency improved from t to t + 1, otherwise efficiency declined. TECH in Equation (6) indicates the change in output due to technological progress. If TECH > 1, technical change enabled the production of more good outputs and fewer bad outputs, otherwise the frontier shifted towards fewer good outputs and more bad outputs. Equation (1) needs to solve four directional distance functions, including the current directional distance function D 0 t x t , y t , b t ; g t , D 0 t + 1 x t + 1 , y t + 1 , b t + 1 ; g t + 1 ; and two mixed directional distance functions D 0 t + 1 x t , y t , b t ; g t ,   D 0 t x t + 1 , y t + 1 , b t + 1 ; g t + 1 .

3.2. Econometric Strategy

3.2.1. Benchmark Difference-in-Differences

The DID method was utilized to study the effect of PSR on GTFP. This is described by:
Y i t   =   α + β P S R it + δ x i t + μ i + γ t + ε i t
where the outcome variable Yit measures the growth of GTFP, as calculated using Equation (4), in city i and time t. PSRit is an indicator variable; if city i adjusted its SO2 levy standard in year T, and t > T (Since some provinces implemented PSR in the middle of the year, this paper stipulated that, if PSR began in July of the current year, the implementation of the province’s policy was set as the following year), PSRit is 1, otherwise PSRit is 0. β reflects the effect of the SO2 levy standards reform on GTFP in the treatment group relative to the control group. μi is the city fixed effect, γ t is the time fixed effect, and ε i t is a stochastic disturbance term. Xit represents other control variables that also affect GTFP. These variables include:
Economic development, as measured by regional GDP per capita (lngdpp) (This study used the logarithm of GDP per capita, and computed the regional GDP per capita, based on constant 2003 prices, the year before which the analysis starts): may have a serious effect on GTFP due to scale effects and pollutant emissions [46,47]. Moreover, according to the Environment Kuznets Curve theory proposed by Grossman and Krueger [48], there is an inverted U-shaped relationship between economic growth and environmental pollution.
The ratio of foreign direct investment to GDP (fdi_gdp): by introducing advanced foreign technologies and management models, foreign direct investment forced China to strengthen its environmental regulations, thereby increasing GTFP [49].
Technology innovation (inno): Technological innovation is the key driving force in economic growth and GTFP. Here, the city innovation index was utilized to measure technology innovation [50,51].
Industrial structure (ind): Enterprises of different sizes and industries consume different amounts of resources in the production process, thus their contributions to regional GTFP also differ. Optimizing industrial structure improves the allocation efficiency of production factors, which affects GTFP by improving technological efficiency [52,53,54].
Capital–labor ratio (lncap_l): Yuan et al. [55] found that, compared with other industries, the high-tech industry had better energy efficiency. This study uses the logarithm of the ratio of regional fixed-asset investment to employment in order to evaluate the capital–labor ratio [56]. Table 2 presents details on each variable.

3.2.2. Parallel Trend Assumption and Time Trend Analysis

The parallel trend assumption is the basic premise of DID analysis. Therefore, this study conducted dynamic effect analysis to test whether the benchmark regression met the parallel trend assumption, as well as to identify the time effect of policy. An event study approach was employed to study the dynamic effect of PSR on GTFP. The model is described as follows:
Y it   =   α 0 + β τ τ P S R i τ + β L P S R i L + β R P S R i R + δ x i t + μ i + γ t + ε i t .
According to Table 1, 2004–2014 was the period during which SO2 levy standard reform in the provinces occurred. τ , in this model, identifies the time when the reform began. τ = 0 is when a province implemented PSR, τ = −1 refers to the year before implementing PSR, and τ = −2 is two years before implementing PSR. Therefore, τ = [−9, 7], other variables are the same as the benchmark model. This study followed Greenstone and Hanna [57] and finally unified τ = [−5, 6] by excluding τ = −1, the year before publishing PSR, to serve as a comparison group. The aim is to obtain adequate sample size for parallel trend assumption and time effect analysis. The model also introduces L and R to replace other periods in order to estimate the average annual effect in the rest years.

3.3. Data

Our study acquired a panel dataset covering 232 cities over the 2004 to 2014 period in China. The sample contains more than two-thirds of all the prefectural-level cities in China. Based on statistics from 2004 and 2014, these 232 prefecture-level cities account for 93.7% of China’s total GDP and 97.3% of population of all prefectural-level cities in China. Therefore, the selected sample is representative. The data for the previously-mentioned variables were collected from official sources, e.g., the China Urban Statistical Yearbook and China Yearbook for Regional Economy, etc. Table 3 provides some descriptive statistical results for the variables.

4. Empirical Results

4.1. Benchmark Regression Results

Table 4 presents the results from Equation (7) utilizing the panel ordinary least squares method. The effect of PSR on GTFP (DEA-based measure) without control variables is shown in column (1), while columns (2)–(6) list the results as the control variables were gradually introduced in order to re-examine the effects. Table 4 shows that the effect of PSR was statistically significant and negative, regardless of the control variables. This indicated that implementing PSR exerted a negative influence on green growth. Furthermore, PSR decreased GTFP statistically significantly, by approximately 1.86% without control variables and nearly 1.58% with all the control variables. Therefore, this research did not support the Porter Hypothesis, which states that enterprises obtain compensation for innovation by improving their technology when environmental regulations become more stringent. The results coincide with the literature that found that promoting pollution levy standards would increase the cost of environmental governance, crowding out enterprise investment and innovation and ultimately hindering productivity improvements [58,59]. Regarding the five control variables, the coefficients for lngdp and fdi_gdp were positive but statistically insignificant. The coefficients for technological innovation, proportion of secondary industry, and capital–labor ratio were statistically significant and positive, which means that GTFP increased as technology improved and industrial structures were optimized. One possible explanation is that the efficiency of production factors allocation improves with technological innovation and better industrial structures.

4.2. Parallel Trend Assumption and Time Trend Analysis

This study tested the parallel trend hypothesis and analyzed the time trend. Figure 1 illustrates the results of Equation (8), reflecting the dynamic effect analysis results of how PSR affects GTFP. β was not significant in the previous policy implementation period, indicating that there was no significant difference between the treatment and control groups before the implementation of the policy, satisfying the parallel trend assumption. After the implementation of the policy, PSR had a negative and statistically significant effect on GTFP, with the negative effect fluctuating over time.

4.3. Robust Check

Table 4 and Figure 1 both illustrate that, overall, PSR had a statistically significant and negative impact on the growth of regional GTFP. However, in order to ensure the regression results were robust, this study performed four robustness checks (see Table 5).
First, to test the time randomness of PSR, we assumed that the policy began two years earlier and then re-examine PSR’s effect on GTFP. We found that the negative effect of PSR on regional GTFP became not statistically significant, and the coefficient was greatly reduced. This regression result is inconsistent with the benchmark regression results, supporting the assumption that the policy implementation times were random.
Second, according to Chambers et al. [60], the Luenberger productivity index involves both a reduction in inputs and an increase in good outputs without choosing a measurement angle, and it is more popular than the Malmquist productivity and ML productivity indexes [59]. Therefore, this study used the Luenberger productivity index to re-examine PSR’s effect on GTFP. The results from column (3) suggest that PSR’s negative effect on regional total factor productivity was still statistically significant.
Third, the provinces of Shanxi and Heilongjiang stipulated that only enterprises that had not completed the construction of flue gas desulfurization facilities, or whose SO2 emissions exceeded the standard, should adjust their SO2 levy standards. Therefore, in the robustness analysis, this study excluded the Heilongjiang and Shanxi provinces from the sample. Column (4) shows that the regression result was consistent with the benchmark regression result, indicating that the regression results were not affected by the data from those two provinces.
Finally, the sample in this study consisted of 232 cities, including Beijing, Shanghai, Tianjin, and Chongqing. Due to the special economic and environmental conditions of these municipalities, we excluded their data to eliminate their city-level impact and further verify the reliability of the estimates. Column 5 lists those results, which reveal that PSR still had a statistically significant inhibitory effect on GTFP, demonstrating that the regression results were robust.

4.4. Mechanism Analysis

Based on the analysis above, PSR led to a statistically significant reduction in the growth of GTFP. A lot of research has examined the factors that affect GTFP. Most research simply divides the ML index into EFFCH and TECH, and then concluded that technological progress is the main source of total factor productivity growth by doing a numerical comparison [43,61]. Following with the existing literature, this section discusses the two direct mechanisms of EFFCH and TECH, aiming to discover whether PSR affects GTFP through EFFCH or TECH. Therefore, we replaced the dependent variable in Equation (7) with EFFCH and TECH to analyze whether PSR affects EFFCH and TECH, with the regression results shown in Table 6; Table 7.
The results on Table 6 and Table 7 show that PSR statistically significantly reduced regional production efficiency and technological progress, without control variables, at the 10% and 5% levels of significance, respectively. However, Table 6 reports that PSR had no significant impact on EFFCH with all control variables were included. Table 7 shows that PSR was statistically significant and negative, whether or not the control variables were included. This indicates that PSR exerted a negative influence on the level of TECH. More specifically, the PSR has significantly decreased TECH by approximately 0.76%, with or without the control variables. Therefore, we concluded that PSR primarily affected GTFP by influencing TECH. One possible explanation is that PSR not only increased enterprise operating costs, but also had a crowding-out effect on the enterprise’s productive investment and technological innovation [58]. The increased cost of environmental governance due to stringent environmental regulations crowded out enterprise investment and innovation in other areas, and ultimately hindered productivity improvement.

4.5. Heterogeneity Analysis

This section estimates the heterogeneity effects of PSR on GTFP by dividing the samples into different groups. Because China’s regional economic development is unbalanced, and because there a great difference in its industrial structures, there are reasons to suspect that the effects of PSR vary between regions. According to the central government’s classification [62], the counties in our sample can be divided into the eastern, central, and western regions. To evaluate whether the effects of PSR vary among regions, we divided the sample into the three regions, and then re-estimate the Equation (7). In Table 8, panel A, B, and C display the regression results of the eastern, central and western samples, respectively. The results show that PSR affected the eastern and western regions, but it had no significant effect on the central region. Specifically, PSR statistically significantly inhibits technological advances in the eastern region, which reduces GTFP. However, PSR statistically significantly promoted regional production efficiency in the western region, which increased the GTFP of western China. One possible explanation is that stricter environmental regulations in the eastern region increased the cost of environmental protection, which ultimately leads to a decrease in GTFP. In addition, the “Pollution Haven Hypothesis” suggests that industries will transfer production to regions with weaker environmental pollution regulations [63]. The new industries then promote efficiency and technological innovation in those regions, so the “Pollution Haven Hypothesis” maybe the chief reason that the PSR statistically significantly improved GTFP in the western regions, since that region had relatively lenient environment regulations.
According to the Delimitation Scheme of Key Cities for Air Pollution Prevention and Control [64] promulgated in 2002, 113 cities were designated as key environmental protected cities (KEPCs), including municipalities, provincial capitals, open coastal cities, and tourist cities. We then split the sample into KEPCs and non-KEPC and reran (7). Table 9 shows that PSR had no statistically significant effect on GTFP in non-KEPCs, but it statistically significantly reduced GTFP in KEPCs. Column (2) and (3) in Table 9 show that TECH was the main reason why PSR reduced GTFP in KEPCs. This may be because, compared with non-KEPCs, KEPCs are the major cities with greater air pollution control and city planning, and their more stringent environmental regulations may cause enterprises to leave, and regional economic to shrank.
Moreover, environmental regulation may affect small and large cities differently, due to their different levels of efficiency and technology. On the one hand, large-scale cities create an economic agglomeration effect, more efficient resource allocation, and more frequent foreign economic exchange, all of which affect urban productivity, which promotes high-quality development. On the other hand, large-scale cities are vulnerable to crowding effects and aggravated urban problems that lower urban productivity. To decipher the impact of PSR on cities of different sizes, we divided the sample into three segments based on the State Council’s city size division standards promulgated in 2014. We categorized cities with populations of less than 1 million as small/medium, large cities as those between 1 and 5 million, and megacities as those larger than 5 million, and re-calculated Equation (7). Table 10 gives the results. The results from panels A and C indicate that PSR had no statistically significant effect on GTFP in small/medium cities or megacities. In panel B, we found that PSR statistically significantly reduced the GTFP of large cities by decreasing TECH and EFFCH, which indicates that the effects of PSR on GTFP vary by city size.

5. Discussion

This study used the ML index to calculate GTFP, examined the impact of PSR on the green growth, and thoroughly analyzed the direct impact mechanism. The empirical results showed that PSR has a statistically significant and negative effect on regional GTFP. A series of robustness analyses validated the results, which are contrary to the conclusions of the Porter Hypothesis. In this sample, the growth rates of GTFP and industrial output between 2004 and 2014 were 4.1% and 17.6%, respectively. Therefore, GTFP growth accounts for 23.8% of total industrial output growth in China from 2004 to 2014 (As per Chen [65], the share was defined as the ratio of productivity growth to total industrial output growth). PSR implementation reduced the growth rate of GTFP by 1.58% when including all control variables, so this study concludes that PSR reduced the growth rate of industrial output by 0.37%. PSR in this study may have affected GTFP by promoting efficiency changes and technological progress, but the mechanism analysis proved that PSR only affected GTFP through technological progress. In addition, PSR reduced the technological growth rate by 0.76% when including all control variables. Further analysis found that technological innovation, industry structure, and the proportion of capital-intensive industry could greatly improve regional GTFP. The heterogeneity analysis revealed that PSR had a greater impact on GTFP in eastern region, KEPCs, and large cities and that all affected GTFP through technology. In large cities, PSR affected GTFP by reducing technology and efficiency levels.
This paper, therefore, proposes the following policy suggestions. First, as mentioned above, the effect of PSR on GTFP varies by region. Therefore, different policies should be formulated according to the conditions in each region, such as reasonable environmental tax rates that do not curb local economic development and gradually reduce pollution. Through a suitable environmental policy system, enterprises could gradually improve their environmental performances while achieving high-quality development.
Second, TECH is the main mechanism through which PSR affects GTFP, thus improvements in technological innovation and industrial structure can significantly improve GTFP. Therefore, the government should endeavor to support enterprise innovation and improve enterprise resource allocation and production efficiency. In addition, the government needs to design a better blueprint for guiding regional industrial transformation and promoting regional GTFP.
Finally, in order to achieve a positive policy effect from its environmental protection tax, China needs to improve the measures supporting it, strengthen the scope and intensity of the environmental protection tax, and vigorously construct an environmental legal system. The higher-level government should increase the weight of environmental performance in evaluations of official performance and avoid problems such as “political shielding” and “economics always takes priority” at the expense of the environment.

Author Contributions

These authors contributed equally to this work. All authors wrote, revised and approved the final manuscript.

Funding

This work is supported by the National Key R&D Program of China (2018YFC0213600), National Social Science Fund of China (18ZDA051) and the National Natural Science Foundation of China (71822402).

Conflicts of Interest

The authors declare no conflict of interest.

Nomenclature

DDFdirectional distance functions
DEAdata envelopment analysis
DIDDifference-in-Differences estimation
MLMalmquist–Luenberger index
GTFPgreen total factor productivity
PSRpollution levy standards reform
KEPCskey environmental protection cities
xtxt is an input vector in period t
ytyt is a “desirable output” vector in period t
btb is an “undesirable output” in period t
gg is the vector of directions in which outputs can be scaled
pt(xt)pt(xt) is the feasible output set for the given input vector x in period t
EFFCHefficiency change
TECHtechnological progress
YitYit is the growth of green total factor productivity measured by ML index
gdppregional GDP per capita
fdi_gdpforeign direct investment
innotechnology innovation
indindustrial structure
cap_lcapital–labor ratio
ygross value of industrial output
SO2industrial SO2 emissions
capitalindustrial fixed asset investment
laboremployees in industrial units
energyindustrial electricity consumption

References

  1. 2017 China environmental situation bulletin published by the Ministry of Ecology and Environment of China. Available online: http://www.mee.gov.cn/hjzl/zghjzkgb/lnzghjzkgb/ (accessed on 5 November 2019). (In Chinese)
  2. Greenstone, M. The impacts of environmental regulations on industrial activity: Evidence from the 1970 and 1977 clean air act amendments and the census of manufactures. J. Political Econ. 2002, 110, 1175–1219. [Google Scholar] [CrossRef]
  3. Greenstone, M.; List, J.A.; Chad, S. The Effects of Environmental Regulation on the Competitiveness of US Manufacturing (No. w18392); National Bureau of Economic Research: Cambridge, MA, USA, 2012. [Google Scholar]
  4. Brolund, J.; Lundmark, R. Effect of Environmental Regulation Stringency on the Pulp and Paper Industry. Sustainability 2017, 9, 2323. [Google Scholar] [CrossRef]
  5. Harrison, A.; Hyman, B.; Martin, L.; Nataraj, S. When Do Firms Go Green? Comparing Price Incentives with Command and Control Regulations in India (No. w21763); National Bureau of Economic Research: Cambridge, MA, USA, 2015. [Google Scholar]
  6. Porter, M.E.; van der Linde, C. Toward a New Conception of the Environment-Competitiveness. J. Econ. Perspect. 1995, 4, 97–118. [Google Scholar] [CrossRef]
  7. Zhou, G.; Liu, W.; Zhang, L.; She, K. Can Environmental Regulation Flexibility Explain the Porter Hypothesis? An Empirical Study Based on the Data of China’s Listed Enterprises. Sustainability 2019, 11, 2214. [Google Scholar] [CrossRef]
  8. Lanoie, P.; Patry, M.; Lajeunesse, R. Environmental regulation and productivity: Testing the porter hypothesis. J. Prod. Anal. 2008, 30, 121–128. [Google Scholar] [CrossRef]
  9. Chen, C.; Lan, Q.; Gao, M.; Sun, Y. Green Total Factor Productivity Growth and Its Determinants in China’s Industrial Economy. Sustainability 2018, 10, 1052. [Google Scholar] [CrossRef]
  10. Jefferson, G.H.; Tanaka, S.; Yin, W. Environmental regulation and industrial performance: Evidence from unexpected externalities in China. Available SSRN 2013, 2216220. [Google Scholar] [CrossRef]
  11. Söderholm, K.; Bergquist, A.K. Growing green and competitive—A case study of a Swedish pulp mill. Sustainability 2013, 5, 1789–1805. [Google Scholar] [CrossRef]
  12. Feng, Z.; Chen, W. Environmental regulation, green innovation, and industrial green development: An empirical analysis based on the Spatial Durbin model. Sustainability 2018, 10, 223. [Google Scholar] [CrossRef]
  13. Sheng, D.; Zhang, H. Environmental Regulations and Upgrades of Exports Quality: Evidence form Two Control Zones Policy in China. Financ. Trade Econ. 2017, 38, 80–97. [Google Scholar]
  14. Han, C.; Sang, R. Enterprises’ Product Switching and Product Quality Improvement under Environmental Regulation. China Ind. Econ. 2018, 2, 43–62. [Google Scholar]
  15. Sheng, D.; Zhang, G. Environmental Regulations of Two Control Zones and Enterprise Total Factor Productivity Growth. Manag. World 2019, 35, 24–42. [Google Scholar]
  16. Li, S.; Chen, G. Environmental Regulation and the Growth of Productivity in China—Evidence from the Revision of Air Pollution Prevention and Control Law in 2000. Econ. Res. J. 2013, 1, 17–31. [Google Scholar]
  17. Long, X.; Wang, W. Environmental Regulation, Corporate Profit Margins and Compliance Cost Heterogeneity of Different Scale Enterprises. China Ind. Econ. 2017, 6, 157–176. [Google Scholar]
  18. Anger, N.; Oberndorfer, U. Firm performance and employment in the EU emissions trading scheme: An empirical assessment for Germany. Energy Policy 2008, 36, 12–22. [Google Scholar] [CrossRef]
  19. Brännlund, R.; Chung, Y.; Färe, R.; Grosskopf, S. Emissions trading and profitability: The Swedish pulp and paper industry. Environ. Resour. Econ. 1998, 12, 345–356. [Google Scholar] [CrossRef]
  20. Anderson, B.J.; Convery, F.; Di Maria, C. Technological Change and the EU ETS: The Case of Ireland. SSRN Electron. J. 2010, 216, 233–238. [Google Scholar] [CrossRef]
  21. Färe, R.; Grosskopf, S.; Pasurka, C.A., Jr. Tradable permits and unrealized gains from trade. Energy Econ. 2013, 40, 416–424. [Google Scholar] [CrossRef]
  22. Färe, R.; Grosskopf, S.; Pasurka, C.A., Jr. Potential gains from trading bad outputs: The case of US electric power plants. Resour. Energy Econ. 2014, 36, 99–112. [Google Scholar] [CrossRef]
  23. Li, Y.; Shen, K. Emission Reduction Effect of China’s Pollution Control Policy—Empirical Analysis Based on provincial Industrial Pollution Data. Manag. World 2008, 7, 7–17. [Google Scholar]
  24. Tu, Z.; Shen, R. Can emissions trading scheme achieve the porter effect in China? Econ. Res. J. 2015, 7, 160–173. [Google Scholar]
  25. Tu, Z.; Shen, R. Can China’s Industrial SO2 Emissions Trading Pilot Scheme Reduce Pollution Abatement Costs? Sustainability 2014, 6, 7621–7645. [Google Scholar] [CrossRef]
  26. Tang, W.; Wu, L.; Qian, H. From Pollution-heaven to Green-growth—Impact of Carbon-market Relocation of Energy-intensive-sectors. Econ. Res. J. 2016, 51, 58–70. [Google Scholar]
  27. Ren, S.; Zheng, J.; Liu, D. Does Emissions Trading System Improve Firm’s Total Factor Productivity—Evidence from Chinese Listed Companies. China Ind. Econ. 2019, 5, 5–23. [Google Scholar]
  28. Qi, S.; Lin, S.; Cui, J. Do Environmental Rights Trading Schemes Induce Green Innovation? Evidence from Listed Firms in China. Econ. Res. J. 2018, 53, 129–143. [Google Scholar]
  29. Peng, F.; Li, B. Analysis on the Concept of Environmental Protection Investment. Environ. Sci. Technol. 2005, 28, 72–74. [Google Scholar]
  30. Zhang, Y.; Qin, F.; Wu, Y. Sustainable Growth or Growth with Pollution—An Analysis on the Sales Growth Patterns of Chinese Industrial Companies. China Ind. Econ. 2015, 2, 89–101. [Google Scholar]
  31. Guo, J.; Fang, Y.; Yang, Y. Does China’s Pollution Levy Standards Reform Promote Emissions Reduction? J. World Econ. 2019, 42, 121–144. [Google Scholar]
  32. Lu, H.; Liu, Q.; Xu, X.; Yang, N. Can environmental protection tax achieve ‘reducing pollution’ and ‘economic growth’? Based on the change perspective of China’s sewage charges. China Popul. Resour. Environ. 2019, 29, 130–137. [Google Scholar]
  33. Li, W.; Chen, N.; Wang, B. The Impact of pollution fee on Green Development. Urban Probl. 2019, 7, 4–16. [Google Scholar]
  34. Pigou, A.C. The Economics of Welfare; Macmillan: London, UK, 1920. [Google Scholar]
  35. The Administration for Levy and Use of Pollution Discharge Fees. Available online: http://www.mee.gov.cn/gzfw_13107/zcfg/fg/xzfg/201605/t20160522_343331.shtml (accessed on 5 November 2019). (In Chinese)
  36. Management Measures for Levy. Available online: http://fgs.mee.gov.cn/gz/gwybmyggz/201811/t20181129_676604.shtml (accessed on 5 November 2019). (In Chinese)
  37. Wang, J.; Long, F.; Ge, C.; Gao, S. Adjustment of pollution charges standard and reform direction of pollution charge system. Environ. Prot. 2014, 42, 37–39. [Google Scholar]
  38. Li, J. The Impact of Environmental Protection Fee to Tax Reform on Eco-environment and Economic Development in China. Manag. World 2017, 3, 170–171. [Google Scholar]
  39. Use of Pollutant Discharge Fees. Available online: http://zfs.mee.gov.cn/gz/bmgz/qtgz/200701/t20070131_100414.shtml (accessed on 5 November 2019). (In Chinese)
  40. Comprehensive Work Plan for Energy Conservation and Emission Reduction. Available online: http://fgs.mee.gov.cn/fg/gwyfbdgfxwj/201811/t20181129_676427.shtml (accessed on 5 November 2019). (In Chinese)
  41. Notice on Adjusting the Levy Standard of Pollution Discharge Fee. Available online: http://zfs.mee.gov.cn/hjjj/gjfbdjjzc/lssfzc/201603/t20160331_334524.shtml (accessed on 5 November 2019). (In Chinese)
  42. Tu, Z. The Coordination of Industrial Growth with Environment and Resource. Econ. Res. J. 2008, 2, 93–105. [Google Scholar]
  43. Wang, B.; Wu, Y.; Yan, P. Environmental Regulation and Total Factor Productivity Growth: An Empirical Study of the APEC Economies. Econ. Res. J. 2008, 5, 19–32. [Google Scholar]
  44. Färe, R.; Grosskopf, S.; Norris, M. Productivity growth, technical progress, and efficiency change in industrialized countries: Reply. Am. Econ. Rev. 1997, 87, 1040–1044. [Google Scholar]
  45. Chung, Y.; Färe, R.; Grosskopf, S. Productivity and undesirable outputs: A directional distance function approach. J. Environ. Manag. 1997, 51, 229–240. [Google Scholar] [CrossRef]
  46. Sadorsky, P. Do urbanization and industrialization affect energy intensity in developing countries? Energy Econ. 2013, 37, 52–59. [Google Scholar] [CrossRef]
  47. Tu, Z.; Hu, T.; Shen, R. Evaluating public participation impact on environmental protection and ecological efficiency in China: Evidence from PITI disclosure. China Econ. Rev. 2019, 55, 111–123. [Google Scholar] [CrossRef]
  48. Grossman, G.M.; Krueger, A.B. Economic growth and the environment. Q. J. Econ. 1995, 110, 353–377. [Google Scholar] [CrossRef]
  49. Yuan, Y.; Xie, R. FDI, Environmental Regulation and Green Total Factor Productivity Growth of China’s Industry: An Empirical Study Based on Luenberger Index. J. Int. Trade 2015, 8, 84–93. [Google Scholar]
  50. Cheng, Z.; Li, W. Independent R and D, Technology Introduction, and Green Growth in China’s Manufacturing. Sustainability 2018, 10, 311. [Google Scholar] [CrossRef]
  51. Kou, Z.; Liu, X. FIND Report on City and Industrial Innovation in China; Fudan Institute of Industrial Development, School of Economics, Fudan University: Shanghai, China, 2017. [Google Scholar]
  52. Han, N.; Yu, W. Quantitative Analysis of the Impact of Industrial Structure on Environmental Pollution in China. Stat. Decis. 2015, 20, 133–136. [Google Scholar]
  53. Yu, Y.; Liu, R.; Yang, X. Research on the Impact of Industrial Structure Upgrading to TFP in China. Ind. Econ. Rev. 2016, 4, 45–58. [Google Scholar]
  54. Elliott, R.; Sun, P.; Zhu, T. The direct and indirect effect of urbanization on energy intensity: A province-level study for China. Energy 2017, 123, 677–692. [Google Scholar] [CrossRef]
  55. Yuan, Y.; Guo, L.; Sun, J. Structure, technology, management and energy using efficiency: Analysis based on 2000-2010 provincial panel data in China. China Ind. Econ. 2012, 28, 18–30. [Google Scholar]
  56. Qin, X.; Yu, W. Foreign direct investment, economic growth and environmental pollution: An empirical study based on spatial panel data of 259 prefecture-level cities in China. Macroecon. Res. 2016, 36, 99–103. [Google Scholar]
  57. Greenstone, M.; Hanna, R. Environmental regulations, air and water pollution, and infant mortality in India. Am. Econ. Rev. 2014, 104, 3038–3072. [Google Scholar] [CrossRef]
  58. Jaffe, A.B.; Peterson, S.R.; Portney, P.R.; Stavins, R.N. Environmental Regulation and the Competitiveness of U. S. Manufacturing: What Does the Evidence Tell US. J. Econ. Lit. 1995, 33, 132–163. [Google Scholar]
  59. Jiang, F.; Wang, Z.; Bai, J. The Dual Effect of Environmental Regulations’ Impact on Innovation—An Empirical Study Based on Dynamic Panel Data of Jiangsu Manufacturing. China Ind. Econ. 2013, 7, 44–55. [Google Scholar]
  60. Chambers, R.G.; Chung, Y.; Färe, R. Benefit and distance functions. J. Econ. Theory 1996, 70, 407–419. [Google Scholar] [CrossRef]
  61. Li, H.; Zhang, J.; Osei, E.; Yu, M. Sustainable Development of China’s Industrial Economy: An Empirical Study of the Period 2001–2011. Sustainability 2018, 10, 764. [Google Scholar] [CrossRef]
  62. The Central Government’s Classification. Available online: http://www.stats.gov.cn/tjsj/zxfb/201701/t20170120_1455967.html (accessed on 5 November 2019). (In Chinese)
  63. Cole, M.A. Trade, the pollution haven hypothesis and the environmental Kuznets curve: Examining the linkages. Ecol. Econ. 2004, 48, 71–81. [Google Scholar] [CrossRef]
  64. Delimitation Scheme of Key Cities for Air Pollution Prevention and Control. Available online: http://www.mee.gov.cn/gkml/zj/wj/200910/t20091022_172141.htm (accessed on 5 November 2019). (In Chinese)
  65. Chen, S. Green Industrial Revolution in China: A Perspective from the Change of Environmental Total Factor Productivity. Econ. Res. J. 2010, 45, 21–34. [Google Scholar]
Figure 1. Dynamic effect analysis. Notes: Figure 1 displays the estimated coefficient and its 95% confidence interval in the dynamic model. The year before the policy implementation was used as the benchmark period. Therefore, the estimated coefficient at time −1 is zero. The regression includes year-fixed and city-fixed effects, and the control variables were added to the model.
Figure 1. Dynamic effect analysis. Notes: Figure 1 displays the estimated coefficient and its 95% confidence interval in the dynamic model. The year before the policy implementation was used as the benchmark period. Therefore, the estimated coefficient at time −1 is zero. The regression includes year-fixed and city-fixed effects, and the control variables were added to the model.
Sustainability 11 06186 g001
Table 1. The provinces that adjusted Levy Standard of SO2 Discharge Fee from 2007 to 2014.
Table 1. The provinces that adjusted Levy Standard of SO2 Discharge Fee from 2007 to 2014.
ProvinceAdjusted DatePre-Adjusted PriceAdjusted Price
Jiangsu2007.7.10.63 RMB/kg1.26 RMB/kg
Anhui2008.1.11.26 RMB/kg
Shanxi2008.1.11.26 RMB/kg
Hebei2008.7.11.26 RMB/kg
Shandong2008.7.11.26 RMB/kg
Neimenggu2008.7.101.26 RMB/kg
Guangxi2009.1.11.26 RMB/kg
Shanghai2009.1.11.26 RMB/kg
Yunnan2009.1.11.26 RMB/kg
Guangdong2010.4.11.26 RMB/kg
Liaoning2010.8.11.26 RMB/kg
Tianjin2010.12.201.26 RMB/kg
Xinjiang2012.8.11.26 RMB/kg
Heilongjiang2012.8.10.95 RMB/kg
Heilongjiang2013.8.11.26 RMB/kg
Beijing2014.1.110 RMB/kg
Ningxia2014.3.11.26 RMB/kg
Zhejiang2014.4.11.26 RMB/kg
Table 2. Main variables and the associated definitions.
Table 2. Main variables and the associated definitions.
Variable TypeVariableDescriptionDefinition
Dependent variable Total factor productivityMLMalmquist-Luenberger index
INDEPENDENT variablesPSRPolicy Dummy variableDummy variable
Control variableslngdppRegional GDP per capitaTaking the logarithm of regional real GDP per capita
fdi_gdpForeign direct investmentProportion of the GDP that is made up of FDI
innoTechnology innovationCity innovation index
indIndustrial structureSecondary industry output value/regional
GDP
lncap_lCapital–labor ratioFixed-asset investment/payrolls
Other indicators used in evaluating total factor productivityyGross value of industrial output
SO2Industrial SO2 emissions
capitalIndustrial fixed asset investment
laborEmployees in industrial units
energyIndustrial electricity consumption
Table 3. The statistical description of the main variables.
Table 3. The statistical description of the main variables.
VariableUnitObservationsMeanStd.DevMinMax
ML-25521.0410.0950.5071.992
PSR-25520.2510.43401
gdpp10,000 RMB/person25523.5793.9900.26147.49
fdi_gdp-25520.0220.02200.153
inno-25526.14830.1690666.958
ind%255250.39410.6982.66090.970
cap_l10,000 RMB/person255217.99312.0510.88688.048
y100 million RMB25522405.8063702.511.9533,000
SO210,000 tons25526.5846.0950.00668.316
capital100 million RMB2552723.537989.30119.6528449.582
labor10,000 people255218.79024.8740.780260.925
energy10,000 tons of standard coal255270.985105.4480.264990.279
Notes: Table 3 is a statistical description of the standard numerical values (no logarithm) of the main variables in this study.
Table 4. Benchmark regression results.
Table 4. Benchmark regression results.
(1)(2)(3)(4)(5)(6)
PSR−0.0186 ***−0.0160 **−0.0151 **−0.0150 **−0.0125 **−0.0158 **
(0.0065)(0.0064)(0.0065)(0.0065)(0.0061)(0.0062)
lngdpp 0.0633 ***0.0609 ***0.0605 ***0.0398 ***0.0213
(0.0139)(0.0140)(0.0140)(0.0147)(0.0160)
fdi_gdp 0.3349 **0.3313 **0.2803 *0.1763
(0.1606)(0.1612)(0.1584)(0.1596)
inno −0.00000.00000.0001 *
(0.0001)(0.0001)(0.0001)
ind 0.0017 ***0.0013 **
(0.0005)(0.0005)
lncap_l 0.0016 ***
(0.0004)
Constant1.0372 ***1.0286 ***1.0197 ***1.0199 ***0.9428 ***0.9570 ***
(0.0058)(0.0061)(0.0071)(0.0072)(0.0265)(0.0267)
Year effectsYesYesYesYesYesYes
City effectsYesYesYesYesYesYes
Observations255225522552255225522552
R-squared0.04500.05020.05170.05140.05490.0620
Notes: Robust t-statistics in parentheses; *, **, and *** represent 10%, 5%, and 1% significant levels, respectively.
Table 5. Results of the robustness checks.
Table 5. Results of the robustness checks.
(1)(2)(3)(4)(5)
PSR−0.0158 **−0.0160 ***−0.0238 ***−0.0128 **−0.0161 **
(0.0062)(0.0061)(0.0085)(0.0065)(0.0063)
PSR_preceding 0.0004
(0.0055)
Constant0.9570 ***0.9569 ***−0.04100.9464 ***0.9576 ***
(0.0267)(0.0266)(0.0349)(0.0283)0.0273
Control VariablesYesYesYesYesYes
Year effectsYesYesYesYesYes
City effectsYesYesYesYesYes
Observations25522552255223432508
R-squared0.06200.06160.06580.06710.0615
Notes: The dependent variable in columns (1), (2), (4), and (5) was GTFP (as measured by the ML index). Column (1) shows the benchmark regression results. Column (2) shows the counterfactual test with the two-period policy advance. The dependent variable in column (3) was the Luenberger index. Column (4) gives the regression result after eliminating the data from the provinces of Shanxi and Heilongjiang. Column (5) is the regression result after eliminating the data from four municipalities (Beijing, Tianjin, Chongqing, and Shanghai). Robust t-statistics are in parentheses. *, **, and *** represent the 10%, 5%, and 1% levels of significance, respectively.
Table 6. The impact of SO2 levy standards reform on efficiency change.
Table 6. The impact of SO2 levy standards reform on efficiency change.
(1)(2)(3)(4)(5)(6)
EFFCHEFFCHEFFCHEFFCHEFFCHEFFCH
PSR−0.0108 *−0.0092−0.0083−0.0088−0.0068−0.0087
(0.0060)(0.0060)(0.0060)(0.0060)(0.0058)(0.0059)
lngdpp 0.0386 **0.0361 **0.0387 **0.02260.0124
(0.0151)(0.0152)(0.0153)(0.0165)(0.0175)
fdi_gdp 0.3497 **0.3724 **0.3328 **0.2754 *
(0.1574)(0.1588)(0.1558)(0.1600)
inno 0.0001 ***0.0002 ***0.0002 ***
(0.0000)(0.0000)(0.0000)
ind 0.0013 ***0.0011 **
(0.0004)(0.0005)
lncap_l 0.0009 **
(0.0004)
_cons1.0151 ***1.0098 ***1.0006 ***0.9995 ***0.9397 ***0.9475 ***
(0.0051)(0.0055)(0.0069)(0.0070)(0.0223)(0.0228)
Year effectsYesYesYesYesYesYes
City effectsYesYesYesYesYesYes
Observations255225522552255225522552
R-squared0.08590.08750.08900.08940.09120.0929
Notes: Robust t-statistics in parentheses; *, **, and *** represent 10%, 5%, and 1% significant levels, respectively.
Table 7. The impact of SO2 levy standards reform on technological progress.
Table 7. The impact of SO2 levy standards reform on technological progress.
(1)(2)(3)(4)(5)(6)
TECHTECHTECHTECHTECHTECH
PSR−0.0076 **−0.0067 **−0.0068 **−0.0063 **−0.0060 *−0.0076 **
(0.0032)(0.0032)(0.0032)(0.0032)(0.0032)(0.0033)
lngdpp 0.0229 ***0.0233 ***0.0208 ***0.0177 **0.0086
(0.0069)(0.0068)(0.0067)(0.0072)(0.0075)
fdi_gdp −0.0563−0.0780−0.0857−0.1369 *
(0.0762)(0.0761)(0.0754)(0.0768)
inno −0.0001 **−0.0001 **−0.0001
(0.0001)(0.0001)(0.0001)
ind 0.00030.0001
(0.0002)(0.0002)
lncap_l 0.0008 ***
(0.0002)
Year effectsYesYesYesYesYesYes
City effectsYesYesYesYesYesYes
Observations255225522552255225522552
R-squared0.30950.31090.31080.31250.31240.3165
Notes: Robust t-statistics in parentheses; *, **, and *** represent 10%, 5%, and 1% significant levels, respectively.
Table 8. Regression results for eastern, central, and western China.
Table 8. Regression results for eastern, central, and western China.
(1)(2)(3)
MLTECHEFFCH
Panel A Eastern areas
PSR−0.0252 ***−0.0164 **−0.0112
(0.0093)(0.0066)(0.0100)
Observations103410341034
R-squared0.08970.40210.1470
Panel B Central areas
PSR−0.00960.0043−0.0141
(0.0142)(0.0055)(0.0134)
Observations902902902
R-squared0.07370.31020.1001
Panel C Western areas
PSR0.0296 **−0.00240.0298 **
(0.0140)(0.0059)(0.0129)
Observations616616616
R-squared0.03180.18710.0436
Control VariablesYesYesYes
Year fixed effectsYesYesYes
City fixed effectsYesYesYes
Notes: Robust t-statistics in parentheses; *, **, and *** represent 10%, 5%, and 1% significant levels, respectively.
Table 9. Regression results for the key environmental protected city (KEPC) and non-KEPC.
Table 9. Regression results for the key environmental protected city (KEPC) and non-KEPC.
(1)(2)(3)
MLTECHEFFCH
Panel A KEPCs
PSR−0.0185 **−0.0109 **−0.0067
(0.0085)(0.0050)(0.0083)
Observations117711771177
R-squared0.06250.37780.1362
Panel B Non-KEPCs
PSR−0.0141−0.0038−0.0118
(0.0088)(0.0042)(0.0080)
Observations137513751375
R-squared0.06370.27230.0649
Control VariablesYesYesYes
Year fixed effectsYesYesYes
City fixed effectsYesYesYes
Notes: Robust t-statistics in parentheses; *, **, and *** represent 10%, 5%, and 1% significant levels, respectively.
Table 10. Regression results for different city sizes (small/medium, large, and mega).
Table 10. Regression results for different city sizes (small/medium, large, and mega).
(1)(2)(3)
MLTECHEFFCH
Panel A Small/Medium cities (<1 million)
PSR0.0312−0.00430.0360
(0.0369)(0.0083)(0.0416)
Observations777777
R-squared−0.03170.3637−0.0214
Panel B Large cities (1 < population < 5 million)
PSR−0.0217 ***−0.0096 **−0.0217 ***
(0.0080)(0.0040)(0.0080)
Observations165016501650
R-squared0.05750.24570.0619
Panel C Megacities (>5 million)
PSR−0.0030−0.0010−0.0028
(0.0100)(0.0056)(0.0102)
Observations825825825
R-squared0.07940.49060.1904
Control VariablesYesYesYes
Year fixed effectsYesYesYes
City fixed effectsYesYesYes
Notes: Robust t-statistics in parentheses; *, **, and *** represent 10%, 5%, and 1% significant levels, respectively.
Back to TopTop