Corporate Greenwashing Unexpectedly Caused by the Green Credit Policy: A Comparison between Environmental Sustainability Information Disclosure and Actual Environmental Protection Investment from China’s Listed Companies

Abstract

This paper employs a continuous difference-in-differences model to compare environmental sustainability information disclosure with the real environmental investment amount of listed companies in China before and after the implementation of the Green Credit Guidelines, and it investigates the influence of Green Credit Policy on corporate environmental sustainability information greenwashing. The results show that the Green Credit Policy unexpectedly leads to the greenwashing of environmental sustainability information of listed companies. After the implementation of the Green Credit Policy, the amount of positive environmental information disclosed by listed companies has increased faster than that of corporate environmental investment. Further research on the intermediary effect shows that the reason why the Green Credit Policy leads to greenwashing is that the Green Credit Policy greatly reduces the availability of bank credit for enterprises on the restricted list, and these enterprises will relax the financing constraints caused by the policy in the future and thus adopt the greenwashing behavior. Heterogeneity analysis shows that these effects become smaller in state-owned enterprises and enterprises with large commercial financing scales, as they are less affected by financing constraints of the Green Credit Policy. The policy goal of the Green Credit Policy is to limit the financing of polluting projects and promote the sustainable development of enterprises. However, by comparing the corporate environmental sustainability information disclosure and the actual amount of environmental investment, this paper finds that the Green Credit Policy has unexpectedly led to corporate greenwashing. This result goes against the policy goal of sustainable development. Commercial banks should be more cautious about the environmental sustainability information voluntarily provided by enterprises to correct these distorted results.

1. Introduction

Green finance aims to restrict the financial supply to polluting enterprises by differentiating the allocation of financial resources to projects with different environmental risks and guiding enterprises to transform projects to clean production and realize sustainable development [1]. The former China Banking Regulatory Commission (CBRC) issued the Green Credit Guidelines (hereinafter referred to as the Green Credit Policy) in 2012. This policy requires commercial banks to categorize and manage their credit-granting businesses according to the environmental risks posed by their customers. For customers with significant environmental risks, the policy mandates the implementation of measures to mitigate these risks; failure to do so results in the denial of credit. Conversely, credit applications that support green, low-carbon, and sustainable projects are prioritized under equivalent conditions.

The implementation of the Green Credit Policy was expected to impose stricter credit constraints on polluting enterprises [2,3]. These enterprises, once categorized as restricted due to their manufacturing activities that potentially harm the environment [4,5], must submit reports to verify the reduction of their environmental risks to obtain bank credits [6,7]. Consequently, polluting enterprises in need of bank credits are incentivized to demonstrate their mitigation efforts and might even resort to greenwashing their environmental sustainability reports in order to be to be more easily classified by banks as unrestricted enterprises [8,9]. Since the Green Credit Policy targets banks, the influencing mechanism in corporate debt financing differs from other environmental regulation policies directly aimed at polluting enterprises [10,11]. By regarding the promulgation of Green Credit Policy as a quasi-natural experiment, we can exclude the interference of other environmental policies and establish a causal relationship with more internal validity [12,13].

Data from the China Banking Regulatory Commission indicates that since the policy’s implementation, the percentage of green loans in total bank loans has steadily increased, reaching 8.7% in 2019. However, instances of environmental violations among enterprises have not shown a corresponding decline, suggesting possible manipulation in environmental reporting. This underscores the need to understand whether the Green Credit Policy inadvertently encourages greenwashing practices. In this context, firms may enhance their environmental sustainability disclosure to appear more compliant and responsible.

Grounded in legitimacy management theory, this paper aims to investigate whether external institutional pressure, specifically the Green Credit Policy, leads to greenwashing in corporate environmental sustainability reports. We employ a continuous difference-in-differences (DID) model to analyze data from listed companies in China from 2008 to 2023. By treating the promulgation of the Green Credit Policy as a quasi-natural experiment, we can isolate its effects and establish a causal relationship with greater internal validity. The DID model allows us to compare changes in environmental sustainability information disclosure and actual environmental investments before and after the policy’s implementation, providing robust empirical evidence on the policy’s impact.

This study makes several significant contributions to the literature and policy discourse on green finance and corporate environmental behavior. By focusing on the impact of the Green Credit Policy, our research provides empirical evidence on how external institutional pressure can lead to greenwashing practices, thus adding to the existing literature on corporate environmental sustainability disclosure and greenwashing. The findings highlight the need for policymakers to consider potential unintended consequences when designing and implementing green finance policies, thereby informing more effective regulatory frameworks that mitigate the risk of greenwashing. Additionally, the use of a continuous DID model in this context offers a novel approach to analyzing policy impacts, enhancing the rigor and validity of the findings. This study also provides insights specific to developing countries, particularly China, where rapid industrialization and environmental challenges necessitate robust regulatory interventions. Our findings may be instrumental for other emerging economies facing similar issues.

2. Literature Review and Hypotheses

2.1. Corporate Environmental Sustainability Information Disclosure and Greenwashing

Corporate environmental sustainability information disclosure can reduce the information asymmetry within and outside enterprises. Linking environmental responsibility with corporate self-interest through channels such as social reputation and external supervision is essential to encourage environmental responsibility and reduce pollution emissions [14]. However, some companies selectively disclose environmental information, which diminishes the effectiveness of environmental information disclosure diminish and causes it to evolve into a speculative behavior, that is, greenwashing [15,16]. Greenwashing is a deceptive tactic for enterprises to selectively reveal positive environmental information and deliberately dissimulate negative environmental information [17], resulting in a lack of practical action to address environmental issues [18,19]. Scholars have discussed the economic and environmental consequences of greenwashing [20,21] as well as the various drivers behind it [22,23]. This research focuses on how external policies pressure corporations into corporate greenwashing in China, which supplements the empirical literature on factors influencing corporate greenwashing and the impact of external institutions on enterprise behavior, especially in developing countries.

2.2. Legitimacy Management Theory

Legitimacy management theory views disclosure as a means for companies to manage their legitimacy. Max Weber defined legitimacy management as the establishment and cultivation by a system of social activities of a generalized belief in the meaning of its existence [24]. An organization’s public perception and understanding determines its legitimacy, and a firm’s social image is heavily influenced by its environmental performance. Firms must maintain their legitimacy by disclosing environmental information to manage the public’s perception of their environmental performance. The empirical literature of legitimacy management theory also supports that polluting enterprises tend to disclose internal corporate information selectively [25,26]. Due to information asymmetry, commercial banks, as outsiders of enterprises, need to spend money to obtain the factual environmental sustainability information of enterprises, which is not in alignment with the cost–benefit principle. Therefore, the primary basis for judging the environmental risk of enterprises is still the information disclosed by enterprises. Enterprises facing financial constraints under the Green Credit Policy may not disclose accurate environmental information as regulations require. Instead, they may present misleading environmental reports after greenwashing to receive more favorable treatment from banks during the assessment review. Thus, our research empirically tests whether enterprises will adopt greenwashing behavior to mitigate the impact of the Green Credit Policy. It is a test of the applicability of legitimacy management theory in the field of enterprise environmental information disclosure.

2.3. Theoretical Framework and Hypothesis

Under the restrictions of the Green Credit Policy, the availability of bank credit to enterprises decreases. Enterprises may respond in two ways: one is to manage the legality of environmental sustainability information by greenwashing to pass the green credit audit of banks, thereby improving the availability of bank credit. The other is for enterprises to genuinely improve their environmental performance by increasing environmental investment, allowing them to be excluded from the negative list of Green Credit Policies. The following figure illustrates this theoretical framework (Figure 1).

This paper measures the greenwashing degree of an enterprise by comparing the positive degree of environmental information disclosed by the enterprise with the actual amount of environmental investment of the enterprise. According to the theoretical framework, we propose two alternative hypotheses and one mediating-effect hypothesis.

H1a: 

Compared with increasing environmental investment, enterprises restricted by the Green Credit Policy are more willing to disclose more positive environmental information.

H1b: 

Compared with disclosing positive environmental information, enterprises restricted by the Green Credit Policy are more willing to increase environmental investment.

H2: 

The Green Credit Policy works by reducing the availability of bank credit for listed companies.

3. Methods and Data

3.1. Variable Design

3.1.1. Dependent Variable

Most of the existing literature uses item scoring to measure the level of environmental information disclosure. The earliest relevant literature by Wiseman (1982) evaluates corporate environmental disclosure by scoring whether firms disclose environmental expenditures, environmental litigation, emissions and abatement actions, and other environmental matters in their annual reports [27]. Later, Cormier and Magnan (1999) and Cormier et al. (2005) [28,29] designed more detailed scoring criteria based on the above ideas for evaluating corporate environmental disclosure. Drawing on the existing literature, this paper proposes an indicator called positive environmental information disclosure (PEID) to measure the positive degree of environmental information disclosed in the annual report of enterprises. As shown in

Table 1. Construction of positive disclosure indicators.

Items	Disclosure Project Name	Element
pd1	ISO certification	Discloses whether ISO14001 or ISO9001 is audited.
pd2	Pollutant emission compliance	Discloses the compliance with pollutant emission standards.
pd3	Environmental guideline	Dislcoses the enterprise’s environmental philosophy, environmental policy, environmental management organization structure, etc.
pd4	Environmental education	Discloses the enterprise’s participation in education related to environmental protection.
pd5	Environmental activities	Discloses the enterprise’s participation in environmental protection activities.
pd6	Environmental honors	Discloses honors or awards in environmental protection.
pd7	“Three simultaneous” policy	Discloses the implementation of the “three simultaneous” policy, which includes design, construction, and production at the same time.

, if enterprises disclose the information of any items, the value of the corresponding item is 1; otherwise, it is 0. The score of PEID is the sum of the values assigned to each corresponding item.

As an audit item in the annual report, the environmental investment of listed companies are audited by well-known audit institutions, and these data are highly reliable, which is one of the few indirect indicators that can objectively measure the improvement of the company’s real environmental performance. We use the ratio of PEID to the natural logarithm of real environmental investment amount plus one of the listed companies to measure the greenwashing degree of enterprises, and the formula is shown in (1).

$${\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{h}\mathrm{i}\mathrm{n}\mathrm{g}}_{\mathrm{i},\mathrm{t}}={\displaystyle \frac{{\mathrm{P}\mathrm{E}\mathrm{I}\mathrm{D}}_{\mathrm{i},\mathrm{t}}}{\mathrm{l}\mathrm{n}({\mathrm{e}\mathrm{n}\mathrm{v}\mathrm{i}\mathrm{r}\mathrm{o}\mathrm{n}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{i}\mathrm{n}\mathrm{v}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}+1)}_{\mathrm{i},\mathrm{t}}}}$$

(1)

When an enterprise discloses more positive environmental information and actually invests less in the environment, it shows that the higher the degree of greenwashing of the enterprise, the higher our ${\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{h}\mathrm{i}\mathrm{n}\mathrm{g}}_{\mathrm{i},\mathrm{t}}$ index.

3.1.2. Independent Variable

This paper uses the cross-multiplier Gcp to measure the extent to which firms are affected by the Green Credit Policy, as shown in Equation (2).

$${\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}={\mathrm{T}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{t}}_{\mathrm{i}}\times {\mathrm{E}\mathrm{F}\mathrm{D}}_{\mathrm{i},\mathrm{t}}$$

(2)

In Equation (2), Treat is the grouping term, which takes 1 if enterprise i belongs to a green credit-restricted industry; otherwise, it takes 0. The Green Credit Guidelines determine which industries are considered green credit-restricted based on industry codes. EFD represents a company’s external financing dependence, calculated by dividing the capital expenditure of enterprises minus net operating cash flow by capital expenditure. Among them, capital expenditure is calculated by subtracting the cash received from selling fixed assets, intangible assets, and other long-term assets from the cash paid by the company for constructing fixed assets, intangible assets, and other long-term assets. Firms susceptible to the Green Credit Policy should be those more dependent on external financing. The reason is that enterprises with sufficient internal cash flow and small capital expenditure will have little impact on their production and operation activities, even without additional external financing. Even if a company falls under the category of restricted enterprises in the Green Credit Policy, it will not be affected if it has enough internal cash flow, has low capital expenditure, and does not rely on external financing. Thus, the policy identification variable Gcp constructed in this paper can better reflect the degree of enterprises affected by the Green Credit Policy.

3.1.3. Control Variables

In order to control the impact of other characteristics of enterprises on environmental information disclosure, referring to the existing literature [30,31,32], firm-level control variables are added: (1) Firm size (lnsize), measured by the total assets of the enterprise plus one to take the logarithm of the total assets of the enterprise. (2) Age of the enterprise (lnage), measured by the logarithmic number of the year minus the establishment year of the enterprise. (3) Nature of property right (soe), state-owned enterprises take 1, otherwise 0. Whether a state-owned enterprise is a state-owned enterprise is judged according to whether the actual controller of the enterprise is state-owned. (4) Return on assets (roa), the ratio of the enterprise’s net profit to the average total assets of the year. (5) Financial level (level), the ratio of the enterprise’s liabilities to its assets at the end of the year. (6) Stock ownership concentration (concentration), the proportion of shares held by the top 10 shareholders to total shares. (7) The company’s transparency (transparency), which is concluded by analyzing the latest annual report through its T&D score [31,32]. Since all the econometric models in this paper control for industry fixed effects, industry-related variables have been absorbed.

3.2. Data Sources

This study constructs the research sample with panel data of A-share listed companies in China from 2008 to 2023. The reason for starting the sample period from 2008 is that only some listed companies disclosed environmental information before 2008. After China’s MEE issued guidelines for improving the disclosure of environmental information by listed companies in 2008, a sufficient number of listed companies disclosed environmental information, enabling a quantitative study with a large sample. The data for corporate environmental information is obtained from the China Stock Market and Accounting Research Database (CSMAR). In this paper, we treat the samples of listed companies according to following conventions: (1) ST enterprises are excluded because they are often in financial difficulties, and thus this paper does not discuss the impact of the Green Credit Policy on them; (2) financial and real estate enterprises are excluded as they are not the target of the green credit; (3) the samples with seriously missing data are excluded; (4) to avoid the influence of outliers, all data are Winsorized at 1% and 99% quartiles. The descriptive statistics of the main variables are shown in

Table 2. Descriptive statistics.

Variable	Obs.	Mean	SD	Min	Median	Max
Greenwashing	35,885	2.472	1.602	0	3.096	10.839
Gcp	35,885	1.883	1.961	0	1.979	6.306
Lnsize	35,885	23.959	1.567	19.347	22.931	26.196
Lnage	35,885	2.942	0.402	1.386	2.933	3.966
soe	35,885	0.373	0.491	0	0	1
roa	35,885	0.038	0.078	−0.384	0.038	0.274
Level	35,885	0.482	0.261	0.067	0.473	1.018
Concentration	35,885	0.316	0.013	0.139	0.585	0.952
Transparency	35,885	67.982	361.882	37	62	100

.

4. Econometric Model and Empirical Analysis

4.1. Econometric Model

This paper examines the impact of green credit policies on corporate environmental disclosure using a continuous DID approach. The green credit policy assessment literature generally identifies policy impacts by whether they are polluting industries. This paper designs a firm-level continuous difference-in-differences variable, Gcp, to identify policy impacts. For the Green Credit Policy affecting firms’ micro-behavior by creating firms’ credit constraints, Gcp contains firms’ external financing dependence variables, enabling more accurate identification of the policy impact and better internal validity than existing methods. Furthermore, it can exclude the impact of industry-level confounders, such as other environmental regulatory policies that can affect polluting industries in the same period. The econometric model of this paper is conducted as follows:

$$\begin{array}{cc}\hfill {\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{h}\mathrm{i}\mathrm{n}\mathrm{g}}_{\mathrm{i},\mathrm{t}}& ={\mathsf{\beta }\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}\times {\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}}_{\mathrm{t}}+\mathsf{\gamma }{\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}+\mathsf{\delta }{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i},\mathrm{t}}+\mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}\_\mathrm{d}\mathrm{u}\mathrm{m}\mathrm{m}\mathrm{y}\hfill \\ & +\mathrm{I}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}\_\mathrm{d}\mathrm{u}\mathrm{m}\mathrm{m}\mathrm{y}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}}\hfill \end{array}$$

(3)

In Equation (3), ${\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{h}\mathrm{i}\mathrm{n}\mathrm{g}}_{\mathrm{i},\mathrm{t}}$ represents the greenwashing degree of enterprise information disclosure. Gcp × Post is the core explanatory variable in this paper, where Post is used to differentiate between before and after the policy implementation, with 0 before 2012 and 1 after 2012. The coefficient before Gcp × Post represents the net effect of the Green Credit Policy on environmental information disclosure, and ε is a random disturbance term. Control is the firm-level control variable. The model controls for year fixed effects and industry fixed effects. i and t represent individual and time, respectively.

4.2. Baseline Regression Results

Table 3. Green Credit Policy and corporate greenwashing.

Greenwashing	(1)	(2)	(3)
Gcp × post	1.138 ***	1.563 ***	2.131 ***
	(0.003)	(0.001)	(0.000)
Gcp	−0.035	−0.019	−0.011
	(0.690)	(0.383)	(0.251)
post	0.692 ***	0.121 *	0.060
	(0.006)	(0.053)	(0.732)
lnsize		0.942 ***	0.856 ***
		(0.000)	(0.000)
lnage		−0.111	0.069
		(0.563)	(0.774)
soe		0.311 ***	0.255 *
		(0.008)	(0.100)
roa		1.403	0.826
		(0.160)	(0.367)
level		0.123	−0.020
		(0.724)	(0.958)
concentration		0.002	0.002
		(0.928)	(0.920)
transparency		0.001	0.001
		(0.977)	(0.979)
_cons	−16.152 ***	−15.850 ***	−12.374 ***
	(0.000)	(0.000)	(0.000)
Year fixed effects	No	No	Yes
Industry fixed effects	No	No	Yes
N	35,885	35,885	35,885

Note: p-values in parentheses, ***, and * indicate significance at the 1%, and 10% levels, respectively. Same as below.

reports the regression results of the econometric model of Equation (3) on corporate greenwashing, which is estimated using a panel fixation effect model. The first column is the regression result of no control variables, year fixed effect, and industry fixed effect. The second column is the regression result of adding control variables and not adding year fixed effect and industry fixed effect. The third column is the regression result of adding control variables, year fixed effect, and industry fixed effect. From Table 3, it is evident from the coefficients of the interaction term in columns (1), (2), and (3) that the Green Credit Policy significantly increases greenwashing. Overall, the empirical results suggest that corporate environmental information disclosure becomes a tool for corporate legitimacy management when facing the impact of the Green Credit Policy. Enterprises will greenwash their environmental sustainability reports by disclosing more positive and less negative environmental information to cope with the adverse effects of the Green Credit Policy rather than increasing real environmental investment. This result is consistent with Hypothesis 1a of this paper.

4.3. Parallel Trend Test

The parallel trend assumption of the DID method requires that the treatment and control groups have a similar change trend before the policy occurs. This paper draws on the event study method to calculate the difference in policy effects between the treatment and control groups before and after the Green Credit Policy shock to compare the parallel trend of the two. The test model is as follows:

$$\begin{array}{cc}\hfill {\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{h}\mathrm{i}\mathrm{n}\mathrm{g}}_{\mathrm{i},\mathrm{t}}& ={\displaystyle \sum _{\mathrm{t}=2008}^{2021}}{\mathsf{\beta }}_{\mathrm{t}}{\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}\times {\mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}}_{\mathrm{t}}+\mathsf{\gamma }{\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}+\mathsf{\delta }{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i},\mathrm{t}}\hfill \\ & +{\mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}}_{\mathrm{d}\mathrm{u}\mathrm{m}\mathrm{m}\mathrm{y}}+{\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}}_{\mathrm{d}\mathrm{u}\mathrm{m}\mathrm{m}\mathrm{y}}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}}\hfill \end{array}$$

(4)

Year is a time dummy variable that takes 1 when the sample observation time is year t and 0 otherwise, and the other variables are the same as the model in Equation (3). Figure 2 show the results of parallel trend tests for greenwashing, respectively. Between 2008 and 2011, the β-estimates and 95% confidence intervals in Figure 2 are around 0, indicating that the parallel trend hypothesis is tenable before launching the Green Credit Policy.

4.4. Robustness Tests

4.4.1. Placebo Test

In order to test that the regression results are not random, the placebo test is conducted by randomly dividing the treatment group and control group. In the year the policy was launched, the same number of samples as the baseline regression treatment group are randomly selected as the treatment group for the placebo test, and the other samples are used as the control group. We use Equation (3) to regress the greenwashing and then randomly simulate them 500 times. Kernel density plots of the regression coefficients of the T-value from 500 random simulations are shown in Figure 3. The results show that the absolute value of the mean T-value of Figure 3 is less than 1.64, indicating that most of the stochastic simulation results fail the 10% significance test. Therefore, there is no significant effect of the Green Credit Policy on the corporate greenwashing after randomized grouping, i.e., the placebo is ineffective. This proves that the estimation results in Table 3 are robust.

4.4.2. Sample Selectivity Bias

Considering that the division based on the industries with pollution projects exists, this may still make the treatment group firms have specific characteristics that may lead to endogeneity problems by sample selectivity bias if these characteristics affect the firms’ environmental information disclosure. In this paper, samples with similar characteristics to the treatment group firms are selected from the control group by propensity score matching (PSM) and paired, and then DID estimation is performed to exclude possible sample selectivity bias. We use the control variables as covariates to estimate the probability that each firm would enter the treatment group in 2012, adopting a 1:1 nearest-neighbor approach to match the closest characteristics of firms for treatment group firms. Then, the DID regression is carried out for the matched control group and treatment group, and the results with the control variable omitted are shown in

Table 4. PSM-DID estimation results.

Greenwashing	(1)	(2)	(3)
Gcp × Post	1.201 *** (0.002)	1.607 *** (0.001)	2.056 *** (0.000)
Control variable	No	Yes	Yes
Year fixed effects	No	No	Yes
Industry fixed effects	No	No	Yes
N	28,372	28,372	28,372

Note: p-values in parentheses, ***, indicates significance at the 1% levels.

. Column (1) is the regression result of no control variables, year fixed effect, and industry fixed effect; column (2) is the regression result of adding control variables and not adding year fixed effect and industry fixed effect; and column (3) is the regression result of adding control variables, year fixed effect, and industry fixed effect. The results show that the coefficients of the interaction terms after matching are still greater than 0 at the 1% significance level, which suggests that the estimation of the baseline regression is robust.

4.5. Mechanism Analysis

This study uses a mediated-effects model to test the mechanism, drawing on the two-step approach to design a mediated-effects model to test whether bank credit availability (lnCredit) mediates channels. The first step is the regression of core explanatory variables on mediating variables:

$$\begin{gathered} {\mathrm{l}\mathrm{n}\mathrm{C}\mathrm{r}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{t}}_{\mathrm{i},\mathrm{t}}={\mathsf{\beta }\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}\times {\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}}_{\mathrm{t}}+\mathsf{\gamma }{\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}+\mathsf{\delta }{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i},\mathrm{t}}+{\mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}}_{\mathrm{d}\mathrm{u}\mathrm{m}\mathrm{m}\mathrm{y}} \\ +{\mathrm{f}\mathrm{i}\mathrm{r}\mathrm{m}}_{\mathrm{d}\mathrm{u}\mathrm{m}\mathrm{m}\mathrm{y}}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em} \end{gathered}$$

(5)

The first step of the mediation effect test is conducted using lnCredit as the mediating variable. lnCredit is measured using the logarithm of the ratio of total short-term and long-term loans to tangible assets of publicly traded companies plus one. Enterprises reduce pollution and improve environmental performance through cleaner production transformation or pollution treatment, both of which require enterprises to invest capital. The improvement of environmental investment often accompanies the improvement of environmental performance. The inclusion of environmental investment as an item in the annual reports of listed companies was audited by a reputable auditing organization, which means that the credibility of the data is high, and it is one of the few indirect indicators that can objectively measure the improvement of a company’s environmental performance.

The results of regression analysis are presented in

Table 5. Mechanism tests.

	(1)	(3)
Variable	lnCredit	Greenwashing
Gcp × Post	−0.081 *** (0.001)	0.193 *** (0.001)
lnCredit		−0.502 *** (0.001)
Control variable	Yes	Yes
Year fixed effects	Yes	Yes
Industry fixed effects	Yes	Yes
N	35,885	35,885

Note: p-values in parentheses, ***, indicates significance at the 1% levels.

. Column (1) displays the regression results of Equation (5) on lnCredit; β is less than 0 at the 1% significance level, indicating that the Green Credit Policy reduces corporate bank credit availability.

The second step of the mediated-effects model is to regress the core explanatory and mediating variables on the explanatory variables:

$$\begin{array}{cc}\hfill {\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{h}\mathrm{i}\mathrm{n}\mathrm{g}}_{\mathrm{i},\mathrm{t}}=& {\mathsf{\phi }\mathrm{l}\mathrm{n}\mathrm{C}\mathrm{r}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{t}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\beta }\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}\times {\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}}_{\mathrm{t}}+\mathsf{\gamma }{\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}+\mathsf{\delta }{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i},\mathrm{t}}+\hfill \\ & {\mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}}_{\mathrm{d}\mathrm{u}\mathrm{m}\mathrm{m}\mathrm{y}}+{\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}}_{\mathrm{d}\mathrm{u}\mathrm{m}\mathrm{m}\mathrm{y}}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}}\hfill \end{array}$$

(6)

The regression results are shown in column (2) of Table 5, which are the regression results of Equation (6) on greenwashing. The coefficients of lnCredit in column (2) is negative at a 1% significance level, which indicates that the lower the availability of bank credit, the higher the degree of environmental information disclosure in greenwashing. Overall, these results suggest that the bank credit availability of corporations plays a mediating effect, which verifies Hypothesis 2 of this paper.

4.6. Heterogeneity Analysis

4.6.1. Heterogeneity in the Property Rights

Due to the prevalence of soft budget constraints issues in SOEs [33], commercial banks may be less stringent in the implementation of the Green Credit Policy for SOEs, resulting in the difficulty of the policy to reduce the bank credit availability of SOEs and SOEs having a relative lack of incentive to greenwash environmental information. This paper examines how heterogeneity in the nature of enterprise property rights affects the interaction term of Equation (3) by introducing property rights as a moderating variable. The specific model is shown in Equation (7):

$$\begin{gathered} {\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{h}\mathrm{i}\mathrm{n}\mathrm{g}}_{\mathrm{i},\mathrm{t}}={\mathsf{\phi }\mathrm{s}\mathrm{o}\mathrm{e}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\gamma }{\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}+\mathsf{\beta }\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}\times {\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}}_{\mathrm{t}}+{\mathsf{\theta }\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}\times {\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}}_{\mathrm{t}}\times \\ {\mathrm{s}\mathrm{o}\mathrm{e}}_{\mathrm{i},\mathrm{t}}+\mathsf{\delta }{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i},\mathrm{t}}+\mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}\_\mathrm{d}\mathrm{u}\mathrm{m}\mathrm{m}\mathrm{y}+\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}\_\mathrm{d}\mathrm{u}\mathrm{m}\mathrm{m}\mathrm{y}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}} \end{gathered}$$

(7)

Column (1) of

Table 6. Heterogeneity analysis.

	(1)	(2)
Greenwashing	Property Rights	The Scale of CCF
Gcp × Post × Soe	−0.121 *** (0.001)
Gcp × Post × BF		−0.219 *** (0.005)
Control variable	Yes	Yes
Year fixed effects	Yes	Yes
Industry fixed effects	Yes	Yes
N	35,885	35,885

Note: p-values in parentheses, ***, indicates significance at the 1% levels.

reports the estimation results of the heterogeneity in the firms’ property rights. The coefficient θ is significantly negative, while the coefficient β is still significantly positive. The results suggest that the Green Credit Policy encourages firms to greenwash environmental information but that it has a weaker effect on state-owned enterprises. The nature of state-owned property rights negatively moderates the Green Credit Policy, affecting corporate greenwashing.

4.6.2. Heterogeneity in the Scale of Commercial Credit Financing

As an alternative financing tool for enterprises, commercial credit financing may also reduce the impact of green credit. For example, delaying delivery to form receivables when trading with upstream enterprises and collecting funds in advance to form advance receipts can supplement capital for the enterprise to alleviate the impact of financing constraints brought by the Green Credit Policy. However, only some enterprises are qualified to obtain large-scale commercial credit financing. This paper uses the ratio of the total amount of the enterprise’s accounts payable, notes payable, and advance receipts to the total assets to measure the enterprise’s scale of commercial credit financing (CCF) as a moderating variable to be added to Equation (3), and the specific equation is shown as follows:

$$\begin{array}{cc}\hfill {\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{h}\mathrm{i}\mathrm{n}\mathrm{g}}_{\mathrm{i},\mathrm{t}}& ={\mathsf{\phi }\mathrm{C}\mathrm{C}\mathrm{F}}_{\mathrm{i},\mathrm{t}}+\mathsf{\gamma }{\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\beta }\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}\times {\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}}_{\mathrm{t}}+{\mathsf{\theta }\mathrm{G}\mathrm{c}\mathrm{p}}_{\mathrm{i},\mathrm{t}}\times {\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}}_{\mathrm{t}}\times {\mathrm{C}\mathrm{C}\mathrm{F}}_{\mathrm{i},\mathrm{t}}\hfill \\ & +\mathsf{\delta }{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i},\mathrm{t}}+\mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}\_\mathrm{d}\mathrm{u}\mathrm{m}\mathrm{m}\mathrm{y}+\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}\_\mathrm{d}\mathrm{u}\mathrm{m}\mathrm{m}\mathrm{y}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}}\hfill \end{array}$$

(8)

Column (2) of Table 6 reports the estimation results of the heterogeneity in the scale of CCF. The coefficient θ is significantly negative, while the coefficient β is significantly positive, indicating that the Green Credit Policy promotes corporate greenwashing. However, the effect is weaker for enterprises with larger commercial credit financing scales. The larger the scale of CCF and the more able the enterprise is to use CCF to alleviate the financing constraints caused by the Green Credit Policy, the lower the degree of corporate environmental information disclosure in greenwashing. The negative moderating effect of the scale of commercial credit financing can also confirm that the impact of the Green Credit Policy on corporate environmental information disclosure is through the role of corporate bank credit acquisition.

5. Discussion and Conclusions

This paper examines the impact of the Green Credit Policy on corporate environmental information greenwashing by using the data of listed companies in China from 2008 to 2023 and presents the following findings. First, the Green Credit Policy significantly promotes the listed companies’ greenwashing of their environmental information. This implies that some companies view environmental information disclosure as a tool to implement environmental legitimacy management when they face the impact of the Green Credit Policy. Enterprises greenwash the environmental disclosure to alleviate credit constraints caused by the policy. This finding proves the applicability of Weber’s legitimacy management theory in corporate environmental disclosure. Second, the intermediary effect test in our research shows that the reason why the Green Credit Policy unexpectedly leads to greenwashing is that the Green Credit Policy greatly reduces the availability of bank credit for credit-restricted listed enterprises, which leads these enterprises to alleviate the financing shortage by providing more positive information. Thirdly, the state-owned property rights and commercial credit financing have a heterogeneous impact on the effect of the policy. Specifically, the Green Credit Policy may have a weaker effect on promoting corporate greenwashing among state-owned enterprises than non-state-owned ones. Additionally, commercial credit financing and bank credit are substitutes, and the larger the scale of commercial credit financing, the smaller the effect of the Green Credit Policy.

Our research findings provide evidence on how finance can support environmental governance and highlight several practical implications. First of all, the unexpected policy effect of the Green Credit Policy promoting corporate greenwashing suggests that listed companies may use environmental information disclosure as a tool for managing corporate environmental legitimacy, thereby reducing the real investment in environmental performance. The Green Credit Policy should further prevent enterprises from obtaining low-cost green credit funds through environmental information disclosure at the level of institutional design. Banking supervision departments and commercial banks should improve the existing green credit management system and prevent the moral hazard behavior of corporate greenwashing under the impact of the Green Credit Policy. The key to prevention is to alleviate the information asymmetry inside and outside the enterprise. In addition to the existing on-site inspection of enterprises and written examination of environmental risks, commercial banks should strengthen third-party assessment and strive to achieve information interconnection with local environmental protection departments to obtain the real environmental information of customers applying for green credit as much as possible, and they should have a more comprehensive, true, and objective understanding of corporate environmental risks.

Secondly, the nature of state-owned property rights and commercial credit financing can weaken the role of green credit, indicating that green credit policies are not effective in all scenarios and there are certain policy blind spots. Environmental governance issues involve the interests of enterprises, governments, and people, and it is difficult to make a single environmental regulation policy be effective everywhere. The Green Credit Policy has the unique advantages of market-based policy instruments, but it also needs a combination of other types of policy instruments to make up for its shortcomings. For state-owned enterprises to fulfill their environmental responsibilities, mandatory directives from government authorities may be more effective than market-based policies such as green credit. For state-owned enterprises, the fulfilment of environmental responsibilities can be used in support of mandatory directives and policies. For enterprises with a large scale of commercial credit financing, because it is difficult to impose financing constraints on financial environmental policies, these enterprises may need to rely on other non-financial environmental policies for constraints. The formulation and implementation of environmental regulation policies should fully consider the heterogeneity among different entities and adopt a variety of differentiated and combined policy tools for collaborative governance.

The marginal contributions of our research are primarily in two aspects: Firstly, it discusses the impact of external institutional pressure on corporate information disclosure from the perspective of legitimacy management, which enriches the research in the field of corporate information disclosure and greenwashing. Secondly, it introduces cross-multiplication terms of external financial dependency and polluting industries to build a continuous DID model. This model has greater internal validity, as it can better mitigate the interference of confounding elements associated with polluting industries than the standard DID model.

However, this study also has limitations. It is limited by its focus on listed companies in China, which may not be representative of all enterprises. Additionally, the data period from 2008 to 2023 may not capture long-term trends and effects of the Green Credit Policy. Future research should explore the long-term impacts of green credit policies and include a more diverse set of companies, including small- and medium-sized enterprises. Further studies could also investigate the effectiveness of different policy combinations and their impact on various types of enterprises.