Corporate Environmental Governance Strategies Under the Dual Supervision of the Government and the Public

Institutional Pressure From the Government and the Public

Policymakers and implementers of national environmental policies are often separate entities. Decision-makers are often obsessed with principal-agent relationships, which makes it difficult to fully understand the implementation of policies (Malesky et al., 2014; Teets, 2018), resulting in “government-firm collusion” and local environmental governance failures. The extensive principal-agent relationships between the central and the local governments may result in two types of problems, namely, adverse selection and moral hazard, thus causing incomplete implementation of environmental policies and hindering the improvement of the local environment. Specifically, local governments deliberately conceal local governance and performance information, which exacerbates the difficulty in obtaining information and the degree of information asymmetry for the central government (H. Jin et al., 2005). Local governments may deliberately bypass central regulations and do nothing to shelter local polluting enterprises for their performance claims (B. Zhang & Cao, 2015). When profitable firms exhibit environmentally damaging behaviors, local governments often avoid measures to achieve environmental compliance at the expense of local employment and tax revenue (Morduch & Sicular, 2002; Shi & Zhang, 2006).

Second, local governments may exploit their information advantages and dominance to take actions that deliberately violate the requirements of the central government and the public, tending to “cut corners” in formalism in environmental governance (Dimitrov, 2015; Lorentzen, 2013). The implementation of environmental policies is often reduced to a tool that benefits local employment, increases fiscal revenue, and promotes the career development of local officials (Qi et al., 2008; Zheng et al., 2015). Therefore, to avoid the self-interested behavior of local governments and realize the effectiveness of environmental governance, the central government must overcome information asymmetry and fully collect relevant information, including the implementation of local government policies and the actual situation of pollution by local corporates, thereby effectively supervising local environmental governance activities (Chen & Xu, 2017). It is generally believed that top-down administrative information collection and bottom-up public opinion collection are considered the two typical channels by which the central government collects information.

CEPI is an important top-down channel for the central government to collect information on local environmental governance, while bottom-up information collection relies mainly on public participation (Li et al., 2020; M. Wang, 2021). In fact, both channels can exert external institutional pressure on polluting firms. According to institutional theory, institutions can restrict and influence organizational behavior. Institutional pressure can be divided into regulatory pressure, normative pressure, and imitation pressure (Scott, 2005), all of which affect corporates’ management decisions and practices (DiMaggio & Powell, 1983). In terms of environmental management, most studies agree that regulatory and normative pressures are critical for firms (Brammer et al., 2012; Colwell & Joshi, 2013; Kassinis & Vafeas, 2006). Regulatory pressure mainly comes from government constraints and regulations (i.e., most formal institutions; Bruton et al., 2010). This pressure usually comes from implementing the standards and regulations rather than the standards and regulations themselves (Berrone et al., 2013). For example, CEPI inspects the implementation and effectiveness of local environmental standards and regulations, thereby exerting regulatory pressure on local governments and firms. Normative pressure comes from the supervision of the public (i.e., informal institutional pressure). Firms need to maintain contact with the public when conducting business, and the public can define the appropriate standards, norms, and behaviors for firms (Roxas & Coetzer, 2012). To seek and maintain legitimacy, firms consciously compare themselves with the demands of society and do their best to operate following common standards, norms, and societal expectations in the institutional realm (Berrone et al., 2013; Hyatt & Berente, 2017).

All in all, these two information collection channels constitute the information base of the central government in supervising and motivating local governments and simultaneously exert formal regulatory pressure and informal normative pressure on corporate environmental governance.

Top-Down CEPI, Bottom-Up Public Participation, and Corporate Environmental Strategies

CEPI and public participation are typical of formal and informal institutions, respectively, and their interaction provides us with a good scenario for combining formal and informal institutions. CEPI is an essential top-down approach for the central government to gather information on local environmental governance (Li et al., 2020; M. Wang, 2021). Comprehensive and special inspections have significantly increased the level of disclosure of environmental information in various regions and have extensively mobilized public participation. The inspection teams have implemented a 24-hour reporting hotline and post office boxes in different regions using Weibo and WeChat. The inspection teams also require local governments to rectify problems, deal with public opinion in each case (Jia & Chen, 2019), and publish the information in the inspection column of newspapers, TV stations, and government websites. These practices have substantially increased the disclosure level of environmental information by local governments. For local polluting firms, the regulatory pressure of CEPI can affect corporate environmental governance practices through penalties and incentives (DiMaggio & Powell, 1983). Under the penalty mechanism, each firm unconditionally complies with environmental standards and regulations (Berrone et al., 2013; Deephouse, 1996). Given the superior powers of the government, failure to comply with regulations can lead to high costs for firms. Under the incentive system, by complying with standards and regulations, firms have the opportunity to obtain multiple incentives, such as subsidies, tax breaks and access to scarce, and necessary resources (DiMaggio & Powell, 1983). Therefore, firms are leaning toward active environmental governance strategies given the threats and opportunities of regulatory pressure.

An important source of information for CEPI is the petitions and reports of the public. Public participation enables the inspection team to quickly understand the details that are difficult to grasp through regular surveys and fully understand the local governments’ environmental governance efforts. Therefore, bottom-up information gathering primarily relies on environmental public participation, including two types of individual and organized. Most existing studies show that organized public participation improves the effectiveness and sustainability of environmental governance. For instance, Wu and colleagues (2020) found that the participation of ENGOs resulted in better environmental quality than individual participation. The reason is that public participation requires a matching institutional environment (Ho, 2001), and individuals require sufficient organizational power (Meligrana et al., 2011). Specifically, organized public participation in environmental governance has the following advantages:

First, organized public participation utilizes natural advantages and motivation to report firms and create public opinion pressure, forcing firms to attach importance to environmental protection during production and operation (González & Pazó, 2008). It is essential for firms to maintain contact with the public when conducting business. The normative pressure from the public has formed standards, norms, and behaviors in the business process of firms (Roxas & Coetzer, 2012). If firms do not comply with these environmental standards and norms, succumb to normative pressures, or implement environmental governance strategies, they will be subject to boycotts, isolation, and social protest. Instead, compliance with standards and norms brings a good reputation to the firms and increases the probability of their long-term survival and profitability (Colwell & Joshi, 2013).

Second, organized public participation promotes the disclosure of government environmental information, thereby effectively supervising “government-firm collusion” and forcing local governments to increase the environmental regulations of firms and increase environmental information disclosure within their jurisdictions. An example is the report of the Pollution Information Transparency Index published by the Institute of Public and Environmental Affairs (IPE). For many years, the IPE has promoted environmental governance in China through information disclosure. It incorporates public opinion into environmental supervision to promote environmental information disclosure and improve environmental governance mechanisms (Anderson et al., 2019).

Third, as an informal form of environmental regulation, organized public participation is a useful complement to formal environmental regulation such as government policy tools (Fu & Geng, 2019; Morgan, 2012). Public participation can facilitate local implementation of environmental governance policies by the central government, especially in developing countries where formal environmental regulation is generally less effective (Blackman, 2010). In addition, CEPI encourages grassroots governments to respond positively to the environmental demands of the public (Li et al., 2020) and ensures proper channels for public participation in environmental governance and fairness of their participation in the process (M. Wang, 2021). These actions promote public participation and indirectly encourage firms to minimize pollution. Some scholars have pointed out that the incentive effect of CEPI on public participation is the result of three interactive strengthening mechanisms, namely, political mobilization, the demonstration of national regulatory capacity, and social learning. The top-down inspections consistently encourage public participation in bottom-up supervision. Instead of ending after previous inspections (Tian et al., 2016), this incentive effect may remove institutional barriers and ensure that public participation becomes a long-term environmental governance mechanism, thus affecting government and corporate environmental governance strategies (Jia & Chen, 2019).

In general, top-down CEPI and bottom-up public participation exert external institutional pressures on corporates, namely, regulatory and normative pressure. The two institutions can promote and complement each other in environmental practice and act together to influence corporate environmental governance behavior. The compound external pressure forces firms to increase their environmental protection initiative and adopt environmental governance strategies such as green innovation or increasing environmental protection investment to avoid the adverse impact of pollution incidents on their social reputation. Based on this analysis, the following is hypothesized:

Sample and Data

The unique model of environmental decentralization, the complex organizational structure of the government, and the multilevel administrative hierarchy in China provide a good setting for our research (Chang & Wu, 2014; Child et al., 2007; Luo et al., 2017). We choose the IPE, an ENGO, and CEPI to start our research. They provide a rare opportunity to weigh and supplement the current theoretical and empirical literature. First, the similarities between ENGOs and the CEPI are that they encourage local governments to disclose environmental information, exert pressure on polluting firms, and promote the interaction of public participation and central government oversight. Therefore, we identify corporate environmental governance strategies under the dual pressure of public supervision and CEPI from a policy perspective. Second, ENGOs and CEPI represent different environmental participants (i.e., the public and the central government; L. Liu et al., 2012), which helps to deconstruct different subjects in the environmental governance system. This setting allows us to examine the specific mechanism of government behavior from two aspects: government penalties and government guidance.

Specifically, we construct our sample of Chinese listed heavily polluting firms from 2013 to 2018 with various data sources. The industry distribution and detailed information of the heavily polluting sample firms are reported in Table A1 in the appendix. To measure corporate environmental governance strategies, we use data from the China Stock Market and Accounting Research (CSMAR) database and the Chinese Research Data Services (CNRDS), which provide detailed information on green patents and environmental investment. The CEPI information comes from the “Central Environmental Protection Inspection” column published on the Ministry of Ecology and Environment of the People’s Republic of China website. The public participation data are retrieved from the report of the Pollution Information Transparency Index issued jointly by the IPE. We also obtain relevant data from the CSMAR database, “China Environmental Statistical Yearbook,” and “China City Statistical Yearbook.” As the second round of CEPI started in September 2019, we set the year 2018 as the ending point of the sample interval. After a routine screening process for deleting missing data, financial and insurance firms, and delisting firms in our sample interval, we obtain a preliminary sample containing 1,505 firm-year observations.

Variable

Model

We construct Model 1 following Gilje and Taillard (2016) to test the compound impact of public participation and CEPI on corporate environmental strategies after the exogenous policy shock of CEPI:

G r e e n i , j , t / I n v e s t i , j , t = α + β 1 P u b i , j + β 2 P o s t C E P I j , t + β 3 P u b i , j * P o s t C E P I j , t + γ X + ε ,

(1)

where i indices firms, j indices cities, and t indices years. T G r e e n i , t represents the number of green patent applications and is the proxy variable of the corporate “source-control” strategy. I n v e s t i , t represents environmental investment and is the proxy variable of corporate “end-treatment” strategy. P u b j , t represents public participation. P o s t C E P I j , t , a multiple-period Difference-In-Differences estimator equals 1 in the years after CEPI, and 0 otherwise. The coefficient β 2 therefore indicates the impact of CEPI on corporate environmental governance strategies. A positive and significant β 2 suggests that CEPI exerts a positive effect on corporate environmental governance strategies while a negative and significant β 2 indicates that CEPI inhibits corporate environmental governance strategies. Finally, X is a series of control variables and ε is a random disturbance term.

β 2 + β 3 >0 indicates that public participation and CEPI have a compound effect on corporate environmental governance strategies in the same direction. Thus, the coefficients β 2 + β 3 measure the compound effect of public participation and CEPI on the choice tendency of corporate environmental governance strategy: “source-control” strategy or “end-treatment” strategy. Robust standard errors are clustered at the firm level to correct for firm-specific autocorrelation and heteroscedasticity in the estimation errors.

Summary Statistics

Table 1 provides the summary statistics of the main variables. The average value of the firms’ green innovation is 1.188, the 10% quantile is 0, the median is 0.693, and the 90% quantile is 3.045. The average value of the firms’ environmental investment is 0.183, the 10% quantile is 0, the median is 0, and the 90% quantile is 0.440. The green innovation level and environmental investment expenditure differ substantially for different firms.

OPEN IN VIEWER

Table 1.

Summary Statistics.

Variable	M	SD	P10	P50	P90
Green	1.188	1.300	0	0.693	3.045
Invest	0.183	0.529	0	0	0.440
Pub	0.840	0.367	0	1	1
PostCEPI	0.459	0.498	0	0	1
Pub*PostCEPI	0.389	0.488	0	0	1
Size	23.029	1.434	21.231	22.940	24.950
Growth	0.049	0.242	–0.205	0.071	0.297
Density	13.391	1.150	11.997	13.328	14.969
Cash	0.062	0.065	–0.016	0.061	0.143
Performance	0.034	0.058	–0.022	0.026	0.106
Market	1.490	0.716	1.059	1.274	2.029
SOE	0.178	0.382	0	0	1
GDP	11.418	0.712	10.514	11.344	12.198
Expenditure	15.741	1.528	14.459	15.666	17.634

Note. Definitions of all variables detail in Table A3 in the appendix. All firm-level continuous variables are winsorized at the 1% level in both tails of the distribution. SD = Standard Deviation; CEPI = Central Environmental Protection Inspection; GDP = gross domestic product; SOE = State-Owned Enterprises.

Main Findings

A critical assumption for using the multiperiod Difference-In-Differences method is the satisfaction of a parallel trend hypothesis. If the treatment group is not subject to policy intervention, its time trend should be the same as that of the control group. We employ the event study method to conduct parallel trend testing, and we use the first period before the policy shock as the base period (as shown in Table A4 in the appendix). The results show that when the explained variable is Green or Invest, the dummy variable coefficients of each period before the policy shock are not significantly different from 0, indicating that the parallel trend hypothesis is satisfied.

Before testing Hypothesis 1, it is necessary to examine the impact of public participation (as shown in Table 2). Columns 1 and 4 show that the green innovation of firms located in cities belonging to the evaluation subject of the IPE is significantly higher than that of other firms; however, there is no significant difference in environmental investment between these firms and other firms. Because it takes time to apply for a green patent, the green innovation levels of the firm in the subsequent two phases are taken as explained variables to test the relationship between them and the explanatory variable. Table 2 shows that the results of the subsequent two phases of green innovation are consistent with the above results, indicating that public participation has a long-term impact on the future green innovation of firms.

OPEN IN VIEWER

Table 2.

Public Participation and Corporate Environmental Governance Strategies.

Variable	(1)	(2)	(3)	(4)
	Green	Greent+1	Greent+2	Invest
Pub	0.227** (2.20)	0.286*** (2.62)	0.314*** (2.62)	0.005 (0.11)
Size	0.567*** (11.71)	0.557*** (10.93)	0.563*** (10.33)	–0.009 (–0.66)
Growth	0.019 (0.17)	0.339** (2.57)	0.209 (1.27)	0.006 (0.11)
Density	–0.049 (–0.80)	–0.054 (–0.84)	–0.060 (–0.87)	0.011 (0.45)
Cash	0.966* (1.91)	0.651 (1.15)	0.832 (1.28)	1.060*** (4.53)
Performance	1.028 (1.52)	1.566** (2.03)	0.789 (0.89)	–0.182 (–0.57)
Market	–0.078 (–1.49)	–0.090 (–1.64)	–0.061 (–1.07)	–0.078*** (–3.83)
SOE	–0.069 (–0.58)	–0.053 (–0.42)	–0.120 (–0.90)	–0.013 (–0.21)
Constant	–11.388*** (–10.60)	–11.053*** (–9.91)	–11.037*** (–9.11)	0.303 (0.85)
Year FE	Y	Y	Y	Y
Industry FE	Y	Y	Y	Y
Observations	1505	1261	1027	1505
R 2	.448	.445	.433	.072

Note. Green_t+1 represents the green innovation level of the firm in the next phase, and so on. Due to the serious lack of data, we exclude the sample of Chifeng, Lhasa, and Baotou. We report t-statistics in parentheses and cluster robust standard errors at the firm level in this table. FE = fixed effect; SOE = State-Owned Enterprises.

*

, **, *** indicate significance at 10%, 5%, and 1% levels, respectively.

Furthermore, we test whether CEPI further affects corporate environmental strategies with Model (1). As shown in Table 3, the interaction term (Pub*PostCEPI) coefficient of public participation and CEPI is significantly positive when the explanatory variable is the next phase of green innovation and the current environmental investment. The coefficient is still positive after adding the single variable PostCEPI, showing that the policy impact of CEPI caused public participation to affect corporate environmental governance strategy further. In other words, public participation and CEPI have a compound impact on corporate environmental strategies, encouraging firms to adopt green innovation and increase environmental investment. A comparison of the sum of the two coefficients indicates that firms are more inclined to adopt “source-control” strategies after the implementation of CEPI.

OPEN IN VIEWER

Table 3.

Public Participation, CEPI, and Corporate Environmental Governance Strategies.

Variable	(1)		(2)		(3)		(4)
	Green		Greent+1		Greent+2		Invest
PostCEPI	–0.114 (–1.30)	–0.260* (–1.83)	0.247*** (2.84)	–0.038 (–0.24)	0.024 (0.21)	0.272 (1.41)	–0.049 (–0.74)	–0.136 (–1.65)
Pub*PostCEPI		0.170(1.36)		0.330** (2.08)		–0.285(–1.51)		0.101* (1.80)
Size	0.574*** (6.32)	0.568*** (6.20)	0.358*** (3.43)	0.342*** (3.30)	0.205* (1.84)	0.218* (1.90)	–0.081** (–2.03)	–0.084** (–2.07)
Growth	–0.098(–1.15)	–0.096(–1.12)	0.296*** (2.77)	0.308*** (2.88)	–0.055(–0.49)	–0.067(–0.58)	0.016(0.23)	0.017(0.24)
Density	–0.025(–0.27)	–0.024(–0.26)	–0.032(–0.47)	–0.031(–0.46)	0.058(0.67)	0.056(0.65)	0.017(0.81)	0.018(0.80)
Cash	0.455(1.07)	0.499(1.16)	–0.039(–0.09)	0.041(0.09)	–0.129(–0.25)	–0.173(–0.34)	0.022(0.08)	0.048(0.17)
Performance	0.113(0.23)	0.141(0.28)	0.503(0.97)	0.556(1.05)	–0.489(–0.75)	–0.563(–0.86)	0.005(0.02)	0.023(0.07)
Market	–0.132* (–1.76)	–0.127* (–1.73)	–0.162* (–1.97)	–0.150* (–1.87)	0.031(0.29)	0.026(0.24)	–0.018(–0.67)	–0.015(–0.56)
SOE	0.083(0.56)	0.081(0.54)	0.295** (2.24)	0.298** (2.24)	0.033(0.22)	0.035(0.24)	0.027(0.23)	0.026(0.22)
GDP	–0.053(–0.70)	–0.059(–0.75)	–0.115(–1.36)	–0.127(–1.41)	0.032(0.13)	0.044(0.18)	–0.049(–0.66)	–0.053(–0.71)
Expenditure	0.018(1.02)	0.021(1.14)	0.019(0.76)	0.023(0.93)	–0.019(–0.11)	–0.009(–0.05)	–0.001(–0.11)	0.000(0.03)
Constant	–11.449*** (–4.81)	–11.307*** (–4.70)	–5.640** (–2.25)	–5.247** (–2.08)	–4.472(–0.97)	–5.028(–1.08)	2.267* (1.86)	2.352* (1.90)
Year FE	Y	Y	Y	Y	Y	Y	Y	Y
Firm FE	Y	Y	Y	Y	Y	Y	Y	Y
Observations	1505	1505	1261	1261	1027	1027	1505	1505
R 2	.187	.189	.172	.179	.125	.128	.030	.032

Note. Green_t+1 represents the green innovation level of the firm in the next phase, and so on. Due to the fixed effects model and individual fixed effects, Pub is absorbed in the regression. We report t-statistics in parentheses and cluster robust standard errors at the firm level in this table. CEPI = Central Environmental Protection Inspection; SOE = State-Owned Enterprises; GDP = gross domestic product; FE = fixed effect.

*

, **, *** indicate significance at 10%, 5%, and 1% levels, respectively.

Endogeneity and Robustness Checks

Robustness checks: Replacing the research model

To ensure the reliability of the previous results, we refer to the multiple policy composite effect evaluation model of Gray and colleagues (2014) and construct Model 2 based on data from 2006 to 2012 as follows:

T G r e e n i , j , t / I n v e s t i , j , t = β 0 + β 1 P u b i , j + β 2 C E P I i , j + β 3 P u b i , j * P E _ 2008 j , t + β 4 P u b i , j * P E _ 2016 j , t + β 5 C E P I i , j * P E _ 2016 j , t + γ X + ε ,

(2)

where i indices firms, j indices cities, and t indices years. P u b i , j and C E P I i , j denote public participation and CEPI, respectively, as used in Model 1. P E _ Y E A R j , t is a time dummy variable indicating the start of the policy (including PE_2008 and PE_2016). β 5 measures the changes in corporate environmental governance strategies in cities that have implemented CEPI based on ENGOs participating in environmental governance. We also compare β 5 with different explained variables to determine whether firms tend to adopt a “source-control” strategy or an “end-treatment” strategy after CEPI. The results are shown in Table A7 in the appendix. The coefficients of CEPI*PE_2016 are all significantly positive, and the coefficients in Columns 1 and 2 have higher significance than those in Column 4, which supports the previous results.

Government Penalties

Public participation and the combination of public and environmental policies may affect corporate environmental strategies through government penalties, encouraging firms to adopt “source-control” strategies. Faced with pressure from the public and higher-level governments, local governments must improve their environmental governance efficiency and implement appropriate penalties for local polluting firms to urge them to strengthen pollution control, such as pollution charges, project restrictions, water and power outages, and other means. To verify the effectiveness of penalties as a transmission mechanism, we select the pollution charge (Charge) as a proxy variable and use the operating income for standardization. Although pollution charges are part of a firm’s costs and reduce profit, they can encourage managers to reflect on the firm’s green development shortcomings. Thus, penalties compensate for the inherent shortcomings of the firms’ environmental governance mechanism and the inability of the organization to implement changes (Ambec & Barla, 2002). Eventually, pollution charges cause firms to adopt “source-control” strategies such as green innovation.

Table 4 reports the regression results of the mechanism analysis. As presented in Column 1, the coefficient of public participation Pub is significantly negative. The coefficient of Pub*PostCEPI in Column 2 is not significant, indicating that the compound effect of public participation and CEPI on corporate environmental strategies is not affected by the governments’ penalties.

OPEN IN VIEWER

Table 4.

Regression Results of Mechanism Analysis.

Variable	Government punishment effect		Government guidance effect
	(1)	(2)	(3)	(4)
	Charge	Charge	Subsidy	Subsidy
Pub	–1.058** (–2.28)		–0.035(–0.48)
PostCEPI		0.130(0.50)		–0.176(–0.87)
Pub*PostCEPI		–0.159(–0.63)		0.281* (1.73)
Size	–0.018(–0.19)	–0.012(–0.13)	0.009(0.34)	0.037(0.34)
Growth	–0.309** (–2.48)	–0.311*** (–2.59)	–0.177(–1.60)	–0.200(–1.41)
Density	0.090(1.49)	0.090(1.21)	–0.053(–1.53)	–0.051(–0.59)
Cash	–0.220(–0.39)	–0.262(–0.45)	0.142(0.42)	–0.130(–0.20)
Performance	–0.597(–0.84)	–0.621(–0.95)	–0.806* (–1.95)	–0.862(–1.15)
Market	–0.125(–1.44)	–0.130(–1.26)	0.029(0.84)	0.126(1.16)
SOE	0.440(1.01)	0.442(0.94)	0.076(1.22)	–0.150(–0.77)
GDP	0.089(1.17)	0.096(1.11)	–0.037(–0.45)	–0.073(–0.54)
Expenditure	–0.017* (–1.84)	–0.019* (–1.66)	0.042** (1.97)	0.046* (1.75)
Constant	–0.462(–0.24)	–0.970(–0.40)	0.767(1.37)	1.254(0.36)
Year FE	Y	Y	Y	Y
Industry/Firm FE	Y	Y	Y	Y
Observations	1,505	1,505	1,505	1,505
R 2	.836	.030	.475	.039

Note. We report t-statistics in parentheses and cluster robust standard errors at the firm level in this table. CEPI = Central Environmental Protection Inspection; SOE = State-Owned Enterprises; GDP = gross domestic product; FE = fixed effect.

*

, **, *** indicate significance at 10%, 5%, and 1% levels, respectively.

Government Guidance

Public participation and the combination of public and environmental policies may affect corporate environmental strategies through government guidance, encouraging firms to adopt “source-control” strategies. Shleifer and Vishny (1994) found that government officials tried to influence firms through subsidies. Studies have shown that government guidance based on subsidies and other forms can encourage firms to implement environmental governance measures. On one hand, government guidance can provide financial support for firms to purchase new technology and equipment (Montmartin & Herrera, 2015), and environmental protection subsidies can directly support firms’ green innovation. On the other hand, government guidance can substantially reduce firms’ risks during innovation (Stiglitz, 2015). Firms use green innovation to carry out long-term environmental governance; thus, they face unknown challenges, such as market fluctuations, financial crises, and technical shortcomings. Government subsidies can send a positive signal in this case (Manso, 2011).

To test whether the transmission mechanism is effective, we select the government environmental protection subsidy (Subsidy) as the proxy variable of government guidance in the regression and use the operating income for standardization. The coefficient of Pub in Column 3 in Table 4 is not significant, indicating that public participation does not promote green innovation by increasing government environmental protection subsidies.

Test of the Policy Timeliness of CEPI

The first round of CEPI (including the pilot) is divided into five batches, with a time span of 2 years. The pressure may show an attenuation trend. Thus, policy deterrents may weaken over time (Jia & Chen, 2019). The subsequent groups of inspected firms may be less willing to adopt a “source-control” strategy due to the imitation effect (Popp, 2010). Therefore, we divide the sample into 2016 and 2017 for another analysis; the results are shown in Table 5.

OPEN IN VIEWER

Table 5.

Regression Results of Heterogeneity Tests.

Variable	Inspection years				Property rights		Geographical area		Fiscal investment
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
	2016	2017	2016	2017	SOE	Non-SOE	Eastern region	Midwestern region	High	Low
PostCEPI	0.494*** (2.67)	0.586*** (4.41)	0.356(1.60)	0.226(0.85)	–0.460(–1.47)	0.012(0.07)	–0.209(–1.11)	0.097(0.42)	–0.208(–1.19)	0.030(0.11)
Pub*PostCEPI			0.173(0.92)	0.462* (1.72)	0.685** (2.27)	0.304* (1.74)	0.579*** (3.08)	0.044(0.19)	0.579*** (3.27)	0.175(0.69)
Size	0.406*** (2.93)	0.320** (2.34)	0.396*** (2.85)	0.308** (2.26)	0.147(0.44)	0.346*** (2.93)	0.241(1.65)	0.365** (2.34)	0.212(0.82)	0.166(1.11)
Growth	0.274** (2.43)	0.441** (2.13)	0.285** (2.52)	0.430** (2.10)	0.490** (2.18)	0.280** (2.28)	0.315** (2.07)	0.312** (2.15)	0.212(1.21)	0.481*** (2.92)
Density	0.177** (2.17)	–0.252*** (–3.34)	0.176** (2.17)	–0.243*** (–3.35)	0.098(0.54)	–0.067(–0.85)	0.061(0.66)	–0.048(–0.44)	–0.015(–0.13)	–0.154(–1.25)
Cash	–0.160(–0.27)	–0.107(–0.16)	–0.132(–0.23)	0.010(0.02)	–0.941(–0.98)	0.164(0.32)	0.922* (1.68)	–0.556(–0.81)	0.284(0.42)	–0.317(–0.46)
Performance	0.520(0.83)	1.018(1.27)	0.590(0.95)	0.921(1.15)	0.597(0.55)	0.733(1.11)	0.706(1.20)	0.068(0.08)	–0.404(–0.56)	1.325(1.30)
Market	–0.156(–0.75)	–0.175** (–2.19)	–0.140(–0.67)	–0.167** (–2.16)	–0.053(–0.15)	–0.127(–1.47)	–0.656*** (–3.12)	0.009(0.11)	–0.409(–1.38)	–0.037(–0.46)
SOE	0.310* (1.79)	0.404* (1.91)	0.309* (1.77)	0.390* (1.93)	––	––	0.332(1.64)	0.320* (1.75)	0.599*** (2.77)	0.416* (1.70)
GDP	–0.163(–0.30)	–0.079(–0.90)	–0.203(–0.36)	–0.101(–1.04)	–0.141(–0.98)	–0.128(–1.16)	–0.076(–0.47)	–0.043(–0.42)	–0.098(–0.82)	–0.735* (–1.90)
Expenditure	0.022(0.87)	–0.136(–0.90)	0.023(0.91)	–0.108(–0.72)	0.034(0.62)	0.021(0.73)	0.034(1.53)	–0.128(–0.74)	–0.698(–1.40)	–0.097(–0.53)
Constant	–8.961(–1.40)	–0.090(–0.02)	–8.301(–1.27)	–0.149(–0.04)	–2.489(–0.31)	–4.821(–1.63)	–4.128(–1.09)	–4.512(–1.00)	8.594(0.78)	8.585(1.26)
Year FE	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Firm FE	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Observations	692	569	692	569	239	1,022	675	586	469	566
R 2	.215	.155	.217	.167	.170	.185	.270	.114	.252	.148
p-value	.092 *		.289 **		.381 *		.738 *		.404 **

Note. All the explained variables are Green_t+1, which represents the green innovation level of the firm in the next phase. Columns 1–4 represent the regression results of the test of the policy timelines of CEPI with the sample divided into two groups of 2016 and 2017. Columns 5 and 6 report the results of heterogeneous effects of public participation and CEPI in SOEs and non-SOEs. Columns 7 and 8 report the results of heterogeneous effects of public participation and CEPI in the eastern region and midwestern region. Columns 9 and 10 report the results of heterogeneous effects of public participation and CEPI in firms with High or Low fiscal investment in environmental protection by local governments. Robust standard errors are clustered at the firm level, and t-statistics are reported in parentheses. CEPI = Central Environmental Protection Inspection; SOE = State-Owned Enterprises; GDP = gross domestic product; FE = fixed effect.

*

, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively. The p value is used to test the significance of the difference in CFLOW coefficient between groups and obtained by self-sampling (Boot-strap) 1,000 times.

The coefficient of PostCEPI indicates that inspection has a significantly positive impact on corporate green innovation in 2016 and 2017, and the coefficient in 2017 is slightly larger than that in 2016. The coefficient of Pub*PostCEPI is significantly positive in 2017, demonstrating that CEPI only had a significant positive impact on firms inspected in 2017. Therefore, there is no imitation effect of the policy. The policy deterrent of CEPI is maintained for a long time, continuously affecting corporate environmental governance strategies and promoting green innovation.

Heterogeneity of Property Rights: SOEs Versus Non-SOEs

We must consider the impact of corporate property rights on environmental governance under the unique institutional background of China. SOEs and non-SOEs may respond differently to environmental strategy decisions in the face of public participation and CEPI. Therefore, we divide the sample into SOEs and non-SOEs. The results are listed in Table 5. Columns 5 and 6 show that green innovation is significantly higher for SOEs than non-SOEs after the CEPI. A possible explanation is that CEPI focuses on “inspecting the government” and encourages local governments to reduce their environmental responsibilities (Li et al., 2020). Compared with non-SOEs, SOEs have a political foundation and are more active in adopting green innovation pressured by public participation and CEPI (Gan et al., 2018; i.e., they adopt “source-control” strategies to meet the requirements of government policies).

Heterogeneity of Economic Conditions: Eastern Region Versus Mid-Western Region

Differences in economic conditions and resource endowments exist in different regions; therefore, firms may make different choices in environmental governance strategies. We divide the sample into the eastern and mid-western regions. As shown in Table 5, the coefficient of Pub*PostCEPI is significantly positive for the sample of the eastern region, indicating that public participation and CEPI have a significant compound impact on corporate environmental governance strategies. However, no significant impact is found in the sample of the mid-Western region. The eastern region has a higher level of economic development than the mid-Western region and has significant environmental problems. Therefore, firms in this region are more likely to adopt “source-control” strategies under the dual pressure of public participation and CEPI. However, firms in the midwestern region may be less regulated due to a greater need for economic development (Chen et al., 2016; B. Zhang & Cao, 2015; B. Zhang et al., 2018).

Heterogeneity of Fiscal Investment in Environmental Protection

Local governments in different cities have different levels of enthusiasm for environmental governance, resulting in differences in annual financial investments for energy conservation and environmental protection. To test whether these differences affect corporate environmental governance strategies, we use the median “energy-saving and environmental protection expenditure” of each city as the threshold. Moreover, we divide the sample into high and low environmental protection expenditures (excluding the sample in municipalities). The results show that the coefficient of Pub*PostCEPI is significantly positive in areas with a high level of financial investment in environmental protection and not significant in areas with low levels of financial investment in environmental protection (as shown in Table 5). When the financial investment in environmental protection is high, local governments are more focused on the environmental governance of firms in their jurisdictions. Therefore, local governments are more likely to encourage local firms to achieve environmental governance goals and adopt a long-term “source-control” strategy to continuously improve green innovation.

Discussion of Hypothesis Tests

In this study, principle-agent theory, as well as institution theory and its derivations, are drawn upon to conceptualize and test an integrated model by examining the impact of the interaction between CEPI and public participation on corporate environmental governance strategies. We also investigate the underlying mechanism and conduct a heterogeneity analysis.

The results empirically support Hypothesis 1 that the interaction between CEPI and public participation has a positive compound effect on the corporate “source-control” strategy. Before testing Hypothesis 1, we examine the impact of public participation. Our findings indicate that public participation encourages firms to adopt “source-control” strategies (i.e., increase their green patent research investment). The IPE has released the report of the Pollution Information Transparency Index of relevant cities every year since 2008, indicating that public participation in the oversight of urban environmental governance and corporate pollution discharge behavior has been continuous. Only by adopting long-term green innovation can firms meet their sustainable development requirements. Based on Table 3, the reasons why the interaction between CEPI and public participation enhances green innovation in firms may be as follows: (a) CEPI protects the public’s right to report the environmental governance of firms, thereby deterring firms from committing environmental violations; and (b) CEPI has a trend of standardization and normalization, which can produce long-term positive policy effects that improve environmental performance by overcoming fundamental problems in the environmental governance system (Jia & Chen, 2019). Therefore, the adoption of long-term “source-control” strategies by firms is more in line with the requirements of long-term inspections.

Furthermore, we investigate the underlying mechanism of the two aspects: government penalties and government guidance. The results reveal that the pressure of public participation causes firms to improve their green innovation by reducing government penalties, and the compound effect of CEPI and public participation encourages firms to adopt “source-control” strategies due to government guidance. The possible explanation is that firms may not have sufficient funds and capabilities to compensate for their environmental governance when faced with huge fines and charges. However, the potential penalties may cause managers to consider the green development shortcomings of firms. It is also difficult to adopt “source-control” strategies without sufficient funds. Therefore, due to public pressure, local governments lower the pollution charges imposed on firms to have sufficient funds to carry out long-term environmental governance and continuously improve green innovation. However, after the enactment of CEPI, the government may appropriately increase pollution discharge charges for heavily polluting firms, which explains why the compound effect is not affected by the government’s penalties. As firms have faced intense inspection pressure after CEPI, it has been difficult for local governments to achieve pollution control by only reducing the pollution charges of firms. Therefore, local governments are more willing to encourage and guide heavily polluting firms to adopt environmental governance strategies through environmental protection subsidies. Due to the injection of government funds, firms have sufficient capital for green innovation.

Implications for Theory

Our research makes the following contributions to theory. First, we have advanced principal-agent theory by focusing on implementing policies at different government administrative levels to provide the following insights: Two critical channels for the central government to collect local information can ensure smooth information flow and ease the principal-agent problems between the central and local governments. While researchers have recognized the central role of the government in environmental governance, they have viewed the government as a single entity and have rarely considered principal-agent issues, assuming they operate across hierarchies in a consistent and coordinated manner (Delmas & Toffel, 2008; Sharma & Henriques, 2005). However, due to environmental decentralization, local governments have their own development agendas and rarely keep pace with the central government in environmental governance (Child et al., 2007). Hence, it is vital to identify the channels for solving the intergovernmental principal-agent problems and analyze their action mechanism. Our results reveal that top-down CEPI improves the environmental information disclosure of local governments through a series of measures, and bottom-up public participation forces local governments to strengthen the environmental supervision of firms and increase environmental information disclosure within their jurisdictions. The combination of the two channels not only helps the entrusting party (the central government) monitor the entrusted party (the local government) but also supervises the wrongdoing of the firm as a third party. Therefore, our research helps to clarify the game relationship between the central government and the local governments and reveals how the supervision and the supervised relationship between the government, the public, and enterprises affect environmental governance, thus advancing principal-agent theory.

Second, we demonstrate empirical support for the compound influence mechanism of different forms of institutional pressure acting on corporate environmental governance decisions, thereby expanding institutional theory. In terms of environmental management, regulatory and normative pressures are theoretically critical for firms (Brammer et al., 2012; Colwell & Joshi, 2013; Kassinis & Vafeas, 2006) and have been the focus of recent research. However, most previous studies have examined the impact of a particular institutional pressure (regulatory or normative pressure) on corporate environmental governance (Berrone et al., 2013; Daddi et al., 2016; Lin, 2019; Walker, 2012). In fact, CEPI and public participation, as typical formal and informal institutions, respectively, are complementary to each other. Top-down inspections continue to encourage public participation in bottom-up regulation, and public participation also provides CEPI with an important source of information. Therefore, the combination of the two may generate greater power, and their effects are mutually reinforcing, which has not been fully explored in the extant literature. In addition, the government remains a key driver of corporate environmental practice (Delmas & Toffel, 2004; Marquis & Qian, 2014). Good environmental governance requires government leadership, public assistance, and positive responses from firms. Thus, the compound effect mechanism of CEPI and public participation is that local governments increase their guidance, such as by issuing government subsidies and tax incentives. In short, we have extended the current research on institutional pressures and corporate environmental governance by arguing that the interaction of CEPI and public participation promotes the corporate adoption of environmental governance strategies and by identifying their influence mechanisms.

Implications for Practice

Our results also provide some practical implications as follows. First, our results demonstrate that government regulation and public participation could play a better role in environmental governance, especially when the central government and local governments have principal-agent problems in a decentralized model. Environmental governance requires the full mobilization of all parties and the multiparty cooperation of the government, the public, and firms (Pinkse & Kolk, 2012). This finding is not only applicable to environmental governance in China but also has practical value in addressing environmental challenges on a global scale, especially in federal states with high levels of local autonomy, such as Germany and the United States (R. Wang et al., 2018).

Second, the compliance issues arising from centralized policy-making and local decentralized implementation are core impediments to good governance worldwide (Malesky et al., 2014). The central government should further unblock public opinion expression mechanisms to smooth the local information collection channels, especially to broaden the environmental appeals and complaint channels of the public. In addition, the central government should adhere to environmental policies and ensure that local environmental governance information is obtained from top to bottom. Moreover, governments should reduce pollution charges, fines, and other penalties related to environmental governance and should adopt more diversified incentives and guidance methods, such as environmental protection subsidies and environmental protection financial investment. The green innovation activities of local firms should be supported to ensure environmental “source governance.”

Third, good environmental governance involves the public. Implementing the central environmental policy requires the active cooperation of local governments and the public; therefore, the third-party oversight of public participation must be included in environmental regulations. Moreover, the concept of ecological civilization should be firmly established. The public should increase their awareness of environmental protection and cultivate ecological ethics and behavior. The implementation of policies requires the interaction of environmental protection behaviors and a wide range of encouragement of public participation. Only in this way can we improve the environmental governance system designed and operated by governments, firms, and the public to achieve “multi-subject synergy” (L. Liu et al., 2012).

Finally, firms should be aware that government inspection and public supervision are continuous. Only by attaching importance to environmental governance and pollution prevention can firms achieve sustainable development. Green innovation is the “source-control” strategy that creates a trade-off between economic and ecological developments, which enables firms to gain unique competitive advantages while trying their best to reduce negative external economic activities. Therefore, it is wise for firms to adopt green innovation strategies to comply with local environmental regulations and boost green growth in the long term.