Incentives for Corporate Environmental Information Disclosure in China: Public Media Pressure, Local Government Supervision and Interactive Effects

: Disclosing environmental information is essential for listed ﬁrms to demonstrate social conscience. To fulﬁll government and public media supervision, Chinese listed ﬁrms are increasing the quality and quantity of disclosed environmental information. This elicits a new topic of interest: the correlation between media/government supervision and corporate environmental information disclosure (EID). The paper addresses this issue through data analysis and factor correlation study in data from some high-pollution ﬁrms in China during 2017–2019. The study ﬁrst introduces a standardized deﬁnition for the quantiﬁcation of media and government supervision, as well as the degree to which the corporation discloses the environmental information. Then, the correlations between the factors are isolated and reﬁned to three sub-topics: (1) how public media and local government supervision affect the quality of corporate EID; (2) how is the interactive effect of public media and government supervision related to corporate EID; (3) how is the distinct ownership of corporate affect the government supervision on corporate EID. The concluding result from the above factor analysis could provide guidance for authorities to adjust certain laws and regulations so that the media and government supervision better motivates the corporate EID, and furthermore, better sustainable development of ecological environment.


Introduction
In the era of open information, the government's awareness of effective public media dissemination increases [1]. With the development of online media platforms for information disclosure [2], corporate environmental performance has been interactively regulated and supervised by local government and public media in China. In 2015, the Chinese Environmental Protection Agency promulgated the Environmental Information Disclosure Act, which further specifies the timeliness and accuracy of corporate environmental information disclosure [3,4]. The authority then proceeded to guide companies to voluntarily disclose environmental information by emphasizing corporate EID as an indicator of corporate social responsibility [5,6]. Chinese enterprises have great autonomy in the manner, content and quantity of environmental information disclosure [7]. Despite all the efforts and progresses made in regulation of corporate environmental information, there is still a lack of a complete environmental information supervision system in China.
To address the issue in EID and its regulation, many studies have been done to locate or raise highly correlated factors of EID quality. Existing literature mainly studies the impact of the internal characteristics and external pressures on the disclosure of corporate environmental information [8][9][10]. Francis's study shows that the internal characteristics of a company such as industry type, company size and financial status have an significant positive impact on the its environmental information disclosure [11]. The external pushing pressure for corporate environmental information disclosure mainly stems from government supervision and media reports [12,13]. Although the literature has been shown that government supervision obviously improves the corporate EID quality, some scholars have pointed out that there is obvious competition between corporate environmental information disclosure and government supervision [14]. Based on the theory of information asymmetry, increased government supervision has encouraged companies with good environmental behaviors to be more willing to disclose more authentic information [15]. In reality, government supervision may also induce companies with poor environmental performance to forge false environmental information [3][4][5].
The previous research shows a stronger implementation of government supervision, obviously improve corporate environmental performance [2][3][4][5][6][7][8][9][10][11]. Based on legitimacy theory, listed firms will strengthen their own legality management and disclose environmental information in strict accordance with the policies and regulations to obtain the recognition from government [15,16]. Empirical studies have pointed out that the reinforced government supervisions significantly improve the EID quality and reduce litigation risks and administrative penalties [17]. In addition, some studies show a distinct U-shaped relationship between government environmental supervision intensity and the EID quality of heavily polluted listed firms.
Other studies provide the insight of positive correlation between public media supervision and corporate EID [18][19][20]. Research has shown that the rapid evolution of public media environment and the improvement of online technologies both promote the Internet penetration ratio and the quality of public information [3,8]. Based on the theory of corporate legitimacy, media statistics are adopted to forecast the behavioral features of original netizen, and the results significantly affect government management and business operation [4,10]. Empirical studies only analyze the quantity and tendency of media coverage alone, which can significantly affect the quality of environmental information disclosure in enterprises [19]. However, the above literature has not examined how the interaction between media amounts and media coverage tendency affect corporate EID, while the interactive effect could improve EID quality more than the factors alone.
Current research mainly focuses on the impacts of government regulation and public media supervision on corporate EID [10][11][12][13][14][15], suggesting that strengthening government and public media supervision can significantly urge listed enterprises to improve the level of environmental information disclosure [20]. In addition, the relationship between government supervision and public media coverage has drawn scholars' attention [21]. Some papers indicate that Chinese local regulatory authority can promptly evaluate public environmental opinions through the rapid media coverage [22,23]. Other researchers show that public media platforms can assist government to implement environmental protection and disclosure the environmental emergency or accidents [24,25]. However, it appears that few studies evaluated how the interaction between government supervision and public media coverage affects the corporate EID.
When discussing Chinese corporate EID and government supervision, the factor of corporate ownership should always be considered [26,27]. Researchers have realized that diverse property ownership firms may bring a distinct influence on Chinese government supervision and corporate social responsibility [26]. More specifically, the current research shows that state-owned companies not only pay more attention to the business performance, but also undertake higher social responsibility, including environmental protection [27]. Meanwhile, some researchers have argued that the negative media coverage of state-owned firms significantly promotes government supervision [26,27], but it does not further examine whether government and media interaction differently affect the EID quality of state-owned and non-state-owned listed firms. Inspired by these research progress and limitations, this study will reach out further in evaluating the interactive effect of government supervision and public media coverage on corporate EID. This paper aims to: (1) provide the empirical evidence on how media pressure, government supervision pressure and their interactivity effect can improve the quality of EID; (2) investigate the correlation coefficients of media amount, media trend and government supervision pressure to corporate EID; (3) find out correlation between EID quality and corporate ownership. This paper will demonstrate the supplementary but positive role of public media and government supervision in regulating corporate EID.
The remainder of this study is constructed as follows. Section 2 provides the literature review and research hypotheses. Section 3 explains the methodology of this research, consisting of data collection, the description of independent variables, dependent variables and control variables and the regression model. Section 4 presents the results of empirical tests and residuals analysis. Section 5 carries out the discussions and implications. The conclusions and limitations of this study are listed in Section 6.

Literature Review and Hypothesis
Strengthening government supervision and expanding public media coverage have a positive impact on corporate environmental performance [3][4][5] and environmental information disclosure [10,11]. Based on the institutional theory, listed firms would improve their EID quality to pursue corporate legitimacy and recognition from government and public media [15][16][17][18]. Under the standardized government regulation, listed firms should disclose environmental information of high quality and trustworthiness by improving corporate management and environmental performance [23,24].
The external pressure on corporate EID comes not only from government supervision, but also from media coverage [25][26][27]. Social mechanisms such as media, stakeholders and environmental accidents directly affect corporate EID [28]; these external pressures play an important role through monitoring mechanism of public opinion [28][29][30]. Positive media coverage of corporate environmental performance inspires listed firms to perform environmentally in annual reports [31]. The quality of corporate EID has thus become the foundation of corporate legitimacy [32].
Based on legitimacy theory, listed firms should comply with the valuable social contracts and operate in environmentally responsible ways [33]. In the condition of sustainability, public opinion pressure motivates listed firms to calculate the losses and gains between sustainable corporate development and business operation [34]. The current research shows that high-level disclosure of corporate environmental performance is the result of government and public media supervision pressure [35,36], and prompts corporate legitimacy as a significant cause of the decreased carbon emissions [37,38]. The constant response to government and public media pressure on corporate environmental issues has gradually become an integral part of corporate legitimacy [36][37][38].
Government supervision can be categorized into three types of institutional pressure: mandatory, normative and imitative [33][34][35]. Mandatory institution pressure refers monitoring corporate performance through the implementation of the environmental policies, regulations and laws using standardized regulations [36]. Previous studies have shown that forced regulatory pressure exerted by the government and its stakeholders can significantly improve corporate EID quality [30,31]. Meanwhile, an increasing number of environmental laws and regulations also significantly strengthen the impact of government supervision on corporate EID [39].
Enterprise legitimacy has been regarded as the metrics for social contract [35][36][37][38], so media coverage and public opinion plays an important role in judging and defining the legitimacy of corporate environmental performance [36][37][38]. The bias in news media coverage can be decisive in guiding public recognition [37,38]. Since then, more than 300 studies from the United States, Germany, Spain and Japan have confirmed a high degree of consistency between public recognition and news media coverage, especially regarding corporate environmental issues [39,40]. With the media platform expansion and the internet media reformation, the binding force of government supervision on corporate EID is constantly transforming [39,40]. This literature indicates that the supervision pressure of local government in different regions differs significantly.

Public Media Pressure and EID
Compared with comprehensive attention on corporate environmental performance, the quantity and quality of environmental information disclosed in the annual report are relatively poor [35][36][37][38][39][40]. Regarding corporate environmental performance, many studies have examined how public media supervision affects the quality of corporate EID [33][34][35], and there has been no consistent conclusion research has not shown a close association between corporate EID and public media supervision [33][34][35]. Other studies, such as those by Du and Chang (2016) and Huang (2019), indicated that public media coverage significantly affected the EID of Chinese listed firms [9,19]. Kim and Lyon (2011) pointed out that the amount and tendency of media coverage both significantly urge listed companies to restrict their environmental performance. In the subsequent empirical research [41], the regression model results show that the amount of media coverage of corporate environmental issues is positively correlated with public opinion, and more corporate news significantly motivates corporations to improve information disclosure quality [42]. This suggests that corporate environmental awareness is stimulated by updated media reporting [43]. However, recent studies have discovered that the numerous media coverage has not always positively influenced corporate information disclosure [28,40], thus this issue offers a new perspective on the relationship between corporate EID and public media supervision.
Other literature has presented the idea that the emotional tendency of media coverage significantly reveals corporate environmental behaviors [32,38]. Indeed,  has found that the Janis-Fadner coefficient can be used for investigating the emotional tendency of public media coverage [27], which emphasizes that public opinion can benefit the supervision and amendment of corporate EID by: (1) gaining a high level of public recognition, (2) providing a high-quality corporate annual report and (3) seeking out more support from stakeholders. Meanwhile, government should be aware that the positive media coverage of corporate environmental performance is related to the high level of corporate information disclosure [3,24]. Some research found that there is a close relationship between public media supervision and corporate environmental performance [11,27], yet hardly conclude the interactive effect of the amount of media coverage and public opinion tendency toward corporate EID. This leads to the following first hypothesis, which is based on linear regression analysis of the relationship among media coverage amount, public media tendency and corporate EID quality.
Hypothesis 1a (H1a). The amount of public media coverage about corporate environmental performance has a significant effect on corporate EID. Hypothesis 1b (H1b). The emotional tendency of public media coverage can significantly encourage listed firms to improve EID quality.
Hypothesis 1c (H1c). The interaction between media amount and media tendency has a more significant influence on corporate EID.
These hypotheses focus on the relationship between public media supervision and corporate environmental information disclosure. The other factor, government regulation, is discussed in the next section.

Government Regulatory Pressure and EID
Iatridigse (2013) and Liu (2017) pointed out that corporate environmental performance is significantly influenced by government supervision [20,29], but the diverse environmental policies and differentiated execution may make a difference. Cai (2017) and Fornes (2019) discovered significant differences in the degree of Chinese local government supervision because of the different legal, economic and environmental environment [3,14]. Other authors classified Chinese government supervision into two parts, namely the macro-governance of the central regulatory authority and the micro-governance of local governments [15][16][17][18][19][20]. Lee (2017) found that high macro-governance significantly improves the environmental awareness of heavy-polluting firms. In the micro-governance perspective of local government [25], Rivera (2013) and Wu (2021) indicated that environmental supervision and corporate environmental performance are relatively poor [35,44]. More particularly, empirical research data showed a negative correlation between the degree of government supervision and the quality of EID [42][43][44] and Monfort (2021) further discovered that the serious corruption problem in the Chinese government resulted in a negative impact on the implementation of corporate environmental policies [33]. Those researchers observed the serious negative impact of government supervision on heavy-polluting listed firms [45,46], while the Chinese government has also realized the injustice in local authority administration [43,45].
In recent literature, the pollution information transparency index (PITI) is used as the scoring standard of local government supervision in China [22,31], and is officially announced by the Chinese Nature Resource Defense Council (NRDC). Indeed, the current empirical research adopted this index to examine how the pollution source regulation [23,35], pollution governance and information disclosure management affect the corporate environmental performance [17]. It is concluded from the empirical results that the actual PITI index motivates the Chinese local government to enhance environmental supervision and protection, thus poor corporate environmental performance and EID quality are expected to be improved [33][34][35][36][37][38].
Moreover, Weber (2014) and Fornes (2019) both believed the unique corporate ownership of Chinese listed firms is another major factor [14,42]. They found that the supervision effect is significant on state-owned firms. Because the Chinese government has more than 50% stock shares in state-owned firms [42], high public recognition and social responsibility brought these listed firms more governmental supervision pressure [38][39][40][41][42][43]. Consequently, Noradiva (2016) and Lee (2017) found that intensive government supervision significantly influenced the EID of state-owned companies [25,34], but this research has not made the comparative analysis of corporate environmental issues between state-owned firms and non-state-owned firms [46,47].
In summary, the above studies examined how government supervision affects corporate environmental performance, but distinct property ownership should also be considered in further research. Therefore, this research proposes the following hypotheses: Hypothesis 2a (H2a). There are significantly different influences of inconsistent government supervision on corporate EID.
Hypothesis 2b (H2b). The government supervision for listed firms with different property ownership has a different influence on corporate EID.

The Interactive Impact of Media and Government Supervision on EID
To ensure the corporate legitimacy and social responsibility, listed firms are not only supervised by the governmental institution, but also concentrated by the public media [22,27]. Numerous studies have demonstrated that both government regulation and public media supervision significantly affect corporate environmental performance [35][36][37][38].
Some authors pointed out that external pressure of enterprises comes from the regulatory pressure, normative pressure and cognitive pressure given by public media and local government [42][43][44][45]. First, regulatory pressure refers to the uniform and mandatory restrictions on environmental performance of listed firms through the standardized policies and regulations promulgated by the government [8,23]. Second, under the correct propaganda and guidance of the society and public media, the pressure of social norms can prompt enterprises to manage their environmental performance [11,17]. Third, in order to obtain legal recognition from social media and the government, companies will strictly ensure the quality of disclosed information [15][16][17][18][19]. However, these studies only make a single-variable analysis of the government regulation or media supervision on corporate environmental performance [22][23][24][25][26], and they have not studied the interrelationship between government and media supervision.
Haque and Ntim (2018) mentioned the radical promotion of digital media technologies and the sustained expansion of network media platforms [16], and they realized that the government supervision effect can be expanded through the dissemination of public media [16,17]. Furthermore, the other empirical research examines the concept that corporate environmental performance would be strictly regulated under the dual supervision of the government and the media [22][23][24][25][26]. Therefore, whether the interaction between public media coverage and government supervision has a more positive and significant impact on corporate EID need to be further discussed. Accordingly, the following hypothesis is proposed:

Hypothesis 3 (H3).
Chinese local government supervision, under the dissemination of public media, has a more significant impact on the quality of corporate EID.
The limited literature discussed the impact of public media supervision and government regulation on corporate EID quality [3][4][5]; it is unclear whether their interactive effect has a positive impact on the EID of heavy-polluting firms. In addition, how internal corporate characteristics such as distinct property and the auditing system affect the government and public media supervision also need to be justified. In any case, further research should consider those factors and carry out further exploratory analysis.

Data Collection
This research takes into evaluation Chinese critical pollutant enterprises of A-shares during the period of 2017-2019 in the Yangtze River Delta region of China. In November 2019, the Chinese National Development and Reform Commission announced the promotion of the integrated development of the green environment in the Yangtze River Delta, which mainly includes Shanghai, Jiangsu and Zhejiang. Compared with other regional provinces, the implementation of government supervision in the Yangtze River Delta region is more complete and the development of the new media platform environment is more mature. The strengthening supervision of public media and government shows a more significant effect on corporate performance. Thus, this study selects listed firms registered in the Yangtze River Delta as the research samples. Furthermore, according to the Environmental and Ecological Guidelines issued by the China Environmental and Ecological Protection Department [29,30], high-polluting enterprises mainly include metal manufacturing, power and energy, machinery manufacturing, pharmaceutical and biological manufacturing, petrochemical industries, etc. Since 2017, the Chinese Public Environment Center and other environmental organizations have jointly released the "Evaluation Index for the Disclosure of China's Urban Pollution Source Regulatory Information (PITI Index)". Based on the latest survey data statistics, the organization has publicly released data only until 2019. Thus, this study chose the sample data during the period from 2017 to 2019.
The data selected in this study mainly are from the following databases, namely the not only provides the media coverage information of listed companies, but also examines the degree of local government supervision. This scoring standard measures government performance, such as the quantity of clean production, the quality of audit information disclosure, sewage charge disclosure and annual emission data disclosure. Second, RKS is the authoritative institution that discloses the quality of corporate social responsibility. More specifically, the advantage of this database is to provide the accurate, comprehensive and sustainable rating data of Chinese listed companies. Third, CSMAR in China is a financial and economic research database with accurate and large-scale information and comprehensive data of public firms. This database refers to CRSP, COMPUSTAT, TAQ, THOMSON and other database standards, and combines Chinese actual national conditions to develop a research-oriented precision database and provide up-to-date fundamentals and financial information.
Owing to descriptions of certain corporate information in the above relevant database, this study further screened all research samples using the following steps: (1) Select samples of Chinese listed firms in A-share market during the period from 2017 to 2019. (2) Classify public firms registered in Yangtze River Delta region, including the Jiangsu, Zhejiang and Shanghai provinces. (3) Exclude the financial listed firms and divide the remaining samples into heavy-pollution and non-heavy-pollution firms. (4) Eliminate the samples with a lack of substantial information disclosure. Thus, the total number of corporate samples used for analysis is 360. Table 1 presents the description of the main variables. Table 1. The classification of the contents disclosed in the annual report.

Content Description
Project Environmental protection policy and annual environmental protection expectation achievements.

Consumption
Total annual resource consumption of an enterprise.
Hardware and software support Enterprise environmental protection investment and technology development.

Sustainability
Waste treatment, recycling and comprehensive utilization. Expenditure Environmental protection expenses.

Others
Additional disclosures such as industry comparisons and rankings.

Environmental Information Disclosure (EID)
This study used the most representative content analysis method to evaluate the EID quality of the sample company. Based on the existing research, the evaluation approach includes two different types. One is a content analysis index of the Global Reporting Initiative (GRI) standard created by Clarkson [6]. It can effectively assess the level of corporate EID in environmental social responsibility reports and public media websites. This evaluation method divides the enterprise environmental information into six areas: project, consumption, hardware and software support, sustainability, expenditure and others, and then calculates the above six aspects. Table 1 describes the six aspects of this corporate environmental information. Another way to evaluate EID is to classify environmental information into three levels: general disclosure, professional disclosure and quantitative disclosure [39]. Table 2 represents the detailed standard of the corporate environmental performance. Based on the comprehensive consideration of the above two research methods, this research also refers to the corporate environmental information score disclosed in the RKS database. The corporate environmental indicator provided by RKS database not only considers the content of environmental information disclosure, but also conducts a standardized evaluation of corporate EID. Therefore, in order to eliminate the subjective error of self-constructed EDI indicators for standard selection and scoring, this study uses the environmental indicator scores provided by the RKS database as a substitute variable for corporate EID.

EID Scoring Standard
Quantity Number of lines related to EID in the annual report.

Sig
One score: disclosure only in the non-financial part. Two scores: disclosure only in the financial part. Three scores: disclosure both in the financial and non-financial part.

Line
One score: text description only. Two scores: quantitative description only.
Tree scores: monetary description.

Time
One score: only current information.
Two scores: only future information. Three scores: current and future information.

Interaction Effect Variables
In the mechanism analysis, media amount, media trend, government supervision pressure and their interactive effects are investigated. To test whether or not public media pressure and government supervision pressure will affect the quality of corporate EID through the paths mentioned above, the following variables are used to investigate the impact of public media, government supervision and their interaction terms on corporate EID.

1.
The amount of media coverage (Media_amount). The media amount of the listed firms is measured by the natural logarithms of media coverage, including the positive, negative and neutral reports. The amount of media coverage is derived from the Institute of Public and Environmental Affairs database (IPE).

2.
The tendency of media coverage (Media_trend). The Janis-Fadner coefficient is used to measure the tendency of media coverage in this research. The Janis-Fadner coefficient is an index for content analysis method proposed by Janis and Fadner (1965). It was first introduced into the study of corporate legitimacy by Deephouse (1996). Since then, Clarkson (2011) used the Janis-Fadner coefficient as a measure of the pressure on corporate legitimacy caused by public opinion supervision [6]. The formula is calculated as follows: where e is the number of positive media coverage, c is the number of negative media coverage and t is the total amount of the positive and negative media coverage. The value of the J-F coefficient is between −1.0 and +1.0. The more positive media coverage of listed firms, the closer the J-F coefficient value is to +1. On the contrary, the more negative media coverage, the closer the J-F coefficient value is to −1.

3.
Government supervision degree (Gov). In this study, the Pollution Information Transparency Index (PITI index) is used as a supervised index to measure the local government's supervision degree for corporate environmental performance. PITI Index is created by two independent research institutions, namely the Institute of Public and Environmental Affairs (IPE) and the National Research and Development Council (NRDC). This index collects EID information of 113 cities in China and evaluates the environmental regulation of local governments. Additionally, the referenced factors of PITI index mainly include the following aspects, respectively, the enterprise exceeds the standard violation, the letter and petition complaint handling situation, the EIA acceptance and environmental acceptance publicity. Furthermore, this indicator is quantitatively and qualitatively analyzed from the four aspects of government supervision: systemic, timeliness, integrity and user-friendliness. In China, the PITI index is the most comprehensive and objective evaluation for local government to reflect the implementation of the environmental supervision policies. Thus, this study defined the PITI index as the degree of local government supervision for corporate environmental performance.

Control Variables
Considering the quality of corporate EID is also affected by other factors, including basic and financial features [20,22]. First, we must consider the basic characteristics of listed firms such as property ownership (State) and auditing system (Big4). Property ownership is an exclusive feature of Chinese public firms, therefore typical property ownership impacts corporate environmental behaviors differently [20]. This study considers whether the listed company is audited by PWC, DTT, KPMG and EY institutions. If the company is audited by these above four audit institutions, the value of the dummy variable Big_4 is 1 and otherwise is 0. Additionally, the financial characteristics of enterprises include the return on assets (ROA) that is measured as the year-end net profit of net assets [35]. The leverage of listed firms (LEV) is defined as the total debt over total assets. TobinQ is served as the market value of the listed firms and is equal to the market share of total assets [25].

Regression Model
This study aims to evaluate the interactive effect of public medium pressure and government supervision on corporate EID and the function of property ownership. The following models are developed. In order to better test the interaction between public opinion pressure and government supervision, this study devises a three-step empirical research. The first step of regression model design is to select the media amount, media trend and government supervision (PITI index) as the independent variables, to evaluate the effect of media pressure and government supervision on corporate EID. The first model is presented as follows.

EID(EID_quantity, EID_sig, EID_line, EID_time
The second step is to investigate the interactive effect of media amount and media trend, thus the variable of Media_amount * Media_trend is supplemented in the following second model.
The third step is to make the comprehensive analysis of the interactive effect of media pressure and government supervision on corporate EID; therefore, this research adds the cooperative indicator of Media_trend * Gov in the next, third model.
The framework of this study is as follows in Figure 1.
pressure and government supervision on corporate EID; therefore, this research adds the cooperative indicator of Media_trend * Gov in the next, third model.
The framework of this study is as follows in Figure 1.

Descriptive Statistics and Correlations
As shown by the descriptive analysis of main variables in Tables 3 and 4, first of all, the maximum value of the corporate EID index is 89.0, while the mean value is 44.27 and the median value is 43.62, disclosing that corporate EID quality is low in general. Moreover, the results show that there is a significant difference of corporate EID quality among different firms in China. Second, the mean natural logarithms of media_amount is 6.3947 and the median is 6.3784, revealing that the amount of media reporting of high-polluting listed firms is normally a lot. Third, the minimum value of media_trend is only 0.2771 while the maximum value is 0.8638, reflecting that there are significant differences of the tendency of public opinion in selected sample enterprises. Fourth, this table shows that the highest government regulatory score was 78.80 among the sampled firms, while the lowest score was only 56.30, the average score was 66.63 and the median value was only 67.50. It also can be seen that the most of the selected Chinese listed firms are subject to the lower government supervision degree; only a few regional governments have a better regulatory effect.
Apart from the statistical analysis of dependent and independent variables, the relevant control variables are also descriptively counted in this study. First, the property ownership is used as an important control variable and the average value of State is 0.4286. It means that 42.86% of sample observations are state-owned firms. Second, this study counted the audit situation of listed firms, while the statistics found that only a particularly small number of Chinese listed firms chose the four major audit institutions. Third, this study considers the financial information and their characteristics of sample firms, including Return on Assets (ROA), Leverage (LEV) and TobinQ. Because this study refers to the Akaike Information Criterion rule (AIC), when testing the regression model with R programming, it is found that the above financial indicators are not obvious. Thus, the system of R programming automatically excludes the above financial information indicators and retains the control variables of State and Big4.

Descriptive Statistics and Correlations
As shown by the descriptive analysis of main variables in Tables 3 and 4, first of all, the maximum value of the corporate EID index is 89.0, while the mean value is 44.27 and the median value is 43.62, disclosing that corporate EID quality is low in general. Moreover, the results show that there is a significant difference of corporate EID quality among different firms in China. Second, the mean natural logarithms of media_amount is 6.3947 and the median is 6.3784, revealing that the amount of media reporting of high-polluting listed firms is normally a lot. Third, the minimum value of media_trend is only 0.2771 while the maximum value is 0.8638, reflecting that there are significant differences of the tendency of public opinion in selected sample enterprises. Fourth, this table shows that the highest government regulatory score was 78.80 among the sampled firms, while the lowest score was only 56.30, the average score was 66.63 and the median value was only 67.50. It also can be seen that the most of the selected Chinese listed firms are subject to the lower government supervision degree; only a few regional governments have a better regulatory effect. Table 3. Variable definition.

EID
The quality of corporate environmental information disclosure.

EID_quantity
The scores of the quantity of corporate environmental performance.

EID_sig
The scores of the content of corporate environmental performance.

EID_line
The scores of the description of corporate environmental reporting.

EID_time
The scores of the update of the corporate environmental reporting.

Media_amount
Ln function of the amount of the media reporting. Media_trend Janis-Fadner coefficient of the media tendency.
Gov PITI index to measure the degree of local government supervision.
State A dummy variable that is equal to 1 if the firm is a state-owned firm and 0 otherwise.

Big4
A dummy variable that is equal to q if the audit institution of firm is PWC, DTT, KPMG and EY institutions and 0 otherwise.

ROA
Fiscal year-end net income/year-end total assets. LEV Total debt/year-end total assets.

TobinQ
TobinQ is a measure based market, and is measured by firm's market value/total assets. Apart from the statistical analysis of dependent and independent variables, the relevant control variables are also descriptively counted in this study. First, the property ownership is used as an important control variable and the average value of State is 0.4286. It means that 42.86% of sample observations are state-owned firms. Second, this study counted the audit situation of listed firms, while the statistics found that only a particularly small number of Chinese listed firms chose the four major audit institutions. Third, this study considers the financial information and their characteristics of sample firms, including Return on Assets (ROA), Leverage (LEV) and TobinQ. Because this study refers to the Akaike Information Criterion rule (AIC), when testing the regression model with R programming, it is found that the above financial indicators are not obvious. Thus, the system of R programming automatically excludes the above financial information indicators and retains the control variables of State and Big4.

Regression Results
According to the principle of Akaike Information Criterion (AIC), the interactive effect model 3 is used for regression analysis. Firstly, this study adopts the joint hypotheses test to examine the effect of media amount, media trend and government supervision on corporate EID. It can be seen from the last column in Table 5 that the media amount and media trend have a significant effect on corporate EID, respectively, at the 10% and 5% level. This indicates that higher media pressure, such as a decreasing number of positive media reporting, is more likely to urge listed companies to attach more importance to improve corporate EID level. Thus, Hypothesis 1a and 1b are supported. Moreover, the regression results show that the collaborative effect of media amount and media trend is significantly positive related to corporate EID quality at the 1% level. This points out that the listed firms obtaining higher media attention have more incentives to release more superior environmental information. Thus, Hypothesis 1c is supported. Secondly, the results of regression models show the relationship between government supervision and corporate EID. The coefficient of "Gov" is significant at the 10% level. The degree of local government supervision also has a negatively significant impact on the quality of corporate EID. It means that the government has stronger supervision over enterprises with poor environmental performance. Therefore, Hypothesis H2(a) is supported. Additionally, the interactive coefficient of "Gov * State" is also significantly positive at the 5% level, which indicates that property ownership positively moderates the relationship between the government supervision degree and corporate environmental disclosure quality. Therefore, as the hypothesis indicates, the connection between government supervision and corporate EID is more intense for state-owned firms, which suggests that state-owned firms and their environmental performance are more sensitive to public opinion and the current medium environment, thus Hypothesis H2(b) is supported.
Thirdly, this study examines the interactive effect of media pressure and government supervision on corporate EID. The empirical results indicate that the quality of corporate EID is positively consistent with the interaction between the media and government supervision. The coefficient of "Media_trend * Gov" is positively significant at the 10% level. The independent variable of "Media_trend * Gov" reflects the external pressure of listed firms. With higher external pressure, the firms' future environmental information quality will be more improved. Therefore, it can be seen that the joint supervision of the media and the government have a more significantly interactive impact on the disclosure of corporate environmental information, and Hypothesis H3 is supported.
Finally, this study examines the impact of property ownership and audit institutions of listed firms on the relationship between government supervision and corporate environmental performance. As shown in Tables 6 and 7, the control variables of "State" and "Big4" both have a significant effect on corporate environmental information quality. Moreover, the parameter estimates of the interaction term of "State * Big4" is also significant at the 5% level, which suggests that the state-owned listed firms with the auditing system of PWC, DTT, KPMG and EY institutions can provide high-quality environmental information; therefore, the supervised pressure has a weaker binding effect. Table 6. The comparison of the regression results of three models.

Residual Analysis
To make sure the reliability of the empirical results obtained, this study conducts a chain residual analysis. Figures 2-4, respectively, present the residual plots and the normal Q-Q plots for model 1, model 2 and model 3.
As we can see in Figure 4, there seem no significant patterns in the residual plot, which means that the errors are independent and the variances do not have significant differences. Moreover, the above normal Q-Q plots shows that the residuals are generally on the line, indicating that the errors are normally distributed.     As we can see in Figure 4, there seem no significant patterns in the residual plot, which means that the errors are independent and the variances do not have significant differences. Moreover, the above normal Q-Q plots shows that the residuals are generally on the line, indicating that the errors are normally distributed.

Discussion and Implications
Currently, major research merely focuses on the direct impact of public media and government supervision on corporate environmental performance [2][3][4][5]. It is shown that corporate EID would be improved under the supervision of government and public media [3][4][5][6][7][8], because regulatory pressure forces companies to maintain corporate legitimacy. However, it is necessary to deeply study whether government and public media supervision have an interactive effect on corporate EID, especially because the media's exposure of negative environmental events would quickly draw more attention from the government and society [9][10][11]. Continuous reporting of online media is more likely to promote government supervision, but too much media exposure may hinder the supervision of government [12][13][14][15]. Therefore, this study supports previous research and presents new conclusions for the interactive effect of government and public media supervision on corporate EID.
In this study, main conclusions of empirical research suggest that the interaction between public media and local government supervision significantly improves the quality of corporate EID in China. PITI index is served as the degree of local government supervision, and the Janis-Fadner coefficient is defined as the tendency of public media [16,17]. Additionally, this research directly examines the cooperative impacts of media amount

Discussion and Implications
Currently, major research merely focuses on the direct impact of public media and government supervision on corporate environmental performance [2][3][4][5]. It is shown that corporate EID would be improved under the supervision of government and public media [3][4][5][6][7][8], because regulatory pressure forces companies to maintain corporate legitimacy. However, it is necessary to deeply study whether government and public media supervision have an interactive effect on corporate EID, especially because the media's exposure of negative environmental events would quickly draw more attention from the government and society [9][10][11]. Continuous reporting of online media is more likely to promote government supervision, but too much media exposure may hinder the supervision of government [12][13][14][15]. Therefore, this study supports previous research and presents new conclusions for the interactive effect of government and public media supervision on corporate EID.
In this study, main conclusions of empirical research suggest that the interaction between public media and local government supervision significantly improves the quality of corporate EID in China. PITI index is served as the degree of local government supervision, and the Janis-Fadner coefficient is defined as the tendency of public media [16,17]. Additionally, this research directly examines the cooperative impacts of media amount and media trend on EID. Media amount and media tendency can precisely reflect the public attention and firms' response. Increasing public attention and negative public opinions bring more pressure to listed firms and improve the corporate EID [18,21]. The other conclusion is that property ownership and auditing institutions of listed firms also significantly influence the relationship between the degree of government supervision and corporate environmental information quality.
The empirical results strengthen the previous research and offer new insight into the interactive effect of public media and government supervision. One innovation of this research is to classify the listed firms by industry and by distinct property for regression analysis. This comparative analysis has also been done by several studies as well [22][23][24][25][26]; therefore, the empirical results shed light on the research of corporate environmental information disclosure from the perspective of the public media and government supervision from the following aspects.
Firstly, with regard to how the government regulation affect the corporate EID, previous literature discussed different aspects of this issue: Haque and Ntim (2018) indicated that the mandatory environmental protection policy from the government significantly improves the quality of corporate environmental information disclosure [16]. In addition, the more stringent the government execution, the more restraint the corporate environmental behavior [27][28][29]. Other authors have mentioned that the environmental performance of Chinese state-owned firms is more intensively supervised by the government [32,33]. Our study supports the conclusion of the previous study and further comparatively analyze the impact of government supervision on non-state-owned firms.
Secondly, the regression results show that public media coverage has significant impact on corporate EID. More specifically, both the number of media reports and the media coverage tendency lead to the obvious changes in social recognition. This conclusion proves the importance of public opinion as one of the supervision pressures, which has significant effect on corporate environmental performance [17,35]. Surprisingly, our study finds that increasing negative media coverage has more positive and significant correlation with the corporate EID. This result differs from other studies indicating that the positive public opinion of corporate coverage significantly improves the quality of EID [37][38][39]. In addition, existing research realized the fact that media coverage amount and public opinion tendency have different influence on corporate environmental performance [40,41], but does not look into the interactive effect. However, our research found that an increasing number of negative media reports have a more significant effect on improving the corporate EID than positive reports.
Finally, our research brings new insight into the interactive effect of public media supervision and government regulation. Empirical results of regression models reflect that negative media reports better urge the government to supervise the disclosure of corporate environmental information. This discovery is based on previous research and made a more in-depth interactive analysis [12][13][14][15]. It also expands to focus on the empirical analysis of Chinese heavy-polluted firms, and makes a comparison between state-owned firms and non-state-owned firms [19,40]. Furthermore, the conclusion supports the previous research by discovering that distinct property and the auditing system also have significant impacts on corporate EID quality [28,39].
Consequently, the research discussed above has made comprehensive analysis and studied the impact of public media and local government supervision on corporate EID, as well as the interactive effect. The implication of this research is advanced in practice and theory.
In terms of practical implications, this study proposes that the high quality of corporate EID is not merely a conformation with government environmental regulation, but also an ideal solution for listed firms to achieve high satisfaction and recognition from the public and society. More significantly, Chinese local government combines the public opinion of corporate performance and focuses more on the negative media coverage; listed firms realize that disclosing false information and poor environmental performance brings more risks to business operations [33][34][35]. Thus, the positive practical management of environmental information disclosure is adopted by listed firms to achieve more social recognition. Additionally, due to the positive relationship between government supervision and public media coverage in China, the government considers public media coverage as a non-mandatory regulatory tool for supervising corporate environmental performance. Meanwhile, listed firms are willing to provide more comprehensive environmental information to public media and deliver positive news to investors and government. Furthermore, this study discovers that the effect of government supervision on Chinese listed firms with different property ownership is significantly different, thus the regulatory authority will carefully modify and supplement corporate environmental information regulations.
In terms of theoretical implications, this research draws on political economy theory to analyze the interaction between the media and the government on corporate EID. First, by using the Janis-Fadner coefficient as a proxy for the tendency of public media, this study discovers that the cooperation between media amount and media trend contributes to the improvement of corporate EID quality. Second, based on legitimacy theory, the standard for corporate legitimacy is that the enterprise behaviors are supposed to be highly recognized by public media and local government. When public media discloses the legitimacy issues of listed firms, the Chinese local government will immediately notice and adopt a series of regulatory principles to improve the corporate environmental performance. Third, the core of the agenda setting theory is that issues reported by the media significantly affect the views and concerns of the public and government [45,46]. Therefore, this study concluded that interactive supervision between public media and local government has a significantly positive influence on improving corporate EID.

Conclusions and Limitations
Nowadays, due to an increasing awareness of environmental and ecological protection in China, public media and government have pay more attention to corporate environmental performance and information disclosure. Based on the legitimacy theory in political science and communication science, this research selects high-polluting public firms in the Yangtze River Delta region as this research object. Then, this study refers to the content analysis method used by RKS Database to measure the quality of environmental information disclosure in the 2017-2019 annual reports of 120 Chinese listed companies. Moreover, this research uses the media amount and the Janis-Fadner coefficient as the measure of public media pressure, and the PITI index provided by the Institute of Public and Environmental Affairs (IPE) and the National Research and Development Council (NRDC) as a measure of the degree of government supervision. This study also carried out an empirical analysis of the interactive effect of public media and local government supervision on corporate EID quality.
The regression results show that: (1) Public media pressure significantly impacts the corporate environmental performance. In the period from 2017 to 2019, the increasing number of public media coverage was dominated by positive news. The rising amount of news and information reflects the idea that public media and local government will pay more attention to cooperate environmental performance. Apart from the media amount, the trend of public media coverage also significantly influences corporate environmental information quality. More specifically, the regression results reveal that the negative media coverage significantly inspires Chinese listed firms to improve EID quality. Moreover, this study set up the interactive variable of media amount and media trend and the value of interactive variable reflects the supervision degree of public media. The empirical results show that the higher public media supervision can significantly improve the corporate EIA quality. (2) This study also finds that a higher degree of local government supervision is crucial to improve corporate EID quality. Furthermore, there is a more significant effect of local government supervision on the EID of state-owned firms. (3) This research further investigates the interactive effect of public media pressure and local government supervision on corporate EID, and reveals that the interaction between public media and local government supervision has a significantly positive impact on the quality of corporate EID. With a strengthened government emphasis on corporate environmental behavior, the supervision pressure of public opinion has also effectively restrained corporate environmental performance. (4) This study concludes that the distinct property ownership and auditing institutions of listed firms has a different impact on corporate environmental performance, which is more helpful for the governments and regulators to approve the appropriate supervision approaches for listed firms with different characteristics.
Although this study inventively puts forward the interactive influence of public media pressure and local government supervision on corporate EID, there are also some limitations. First, this research refers to a manual scoring method and relevant data from RKS database to measure corporate EID quality, which is relatively subjective to some degree. Second, this study makes the partial analysis of high-polluting listed firms in the Yangtze River Delta region of China, which results in a small sample size and is noncomprehensive for empirical analysis. Further research with more wide-ranging samples will be more beneficial to precisely investigate how public media pressure and government supervision affect corporate EID.