Can Green Credit Policy Promote the High-Quality Development of China’s Heavily-Polluting Enterprises?

Abstract

Prior literature on the green innovation effects of green credit policies is extensive. However, few scholars have focused on the impact of green credit policies on the high-quality development of heavily-polluting enterprises. Based on this, this study employs the difference-in-differences (DID) model to explore the causal relationship between the Green Credit Guidelines (Guidelines) issued in 2012 and the high-quality development of heavily-polluting enterprises. Additionally, we test whether the effect of upgrading human resources in enterprises strengthens this causal relationship. Our findings suggest that the implementation of the Guidelines has significantly promoted the development quality of heavily-polluting enterprises and the promotion effect is more significant in enterprises with higher development quality, state-owned enterprises, large-scale enterprises, and enterprises in the western region of China. Further research reveals that the effect of upgrading human resources in enterprises has reinforced the positive impact of Guidelines on the high-quality development of enterprises. From the perspective of high-quality development of enterprises, in this paper, we expand the research into the effects of green credit policy, providing a decision-making reference for the promotion and improvement of subsequent green credit policy in the future.

1. Introduction

At present, environmental issues have increasingly gained prominence [1]. Transforming the mode of economic development and pursuing high-quality progress has become the consensus of society as a whole. To balance environmental protection, transform economic development patterns, and achieve high-quality economic development, governments around the world have generally put forward goals of promoting the green economy and sustainable development, in which green finance policy is an important component. The essential goal of green finance is to promote sustainable development through financial products and services. This necessitates that the financial sector embraces environmental protection, energy conservation, and low-carbon strategies as crucial decision-making considerations when investing and financing. Green finance plays a dual role: firstly, it directs funds towards green industries, facilitating their growth and sustainability; secondly, it enhances the oversight of financial institutions, enabling them to assume greater social responsibility and contribute to environmental preservation. In essence, green finance represents another manifestation of the sustainable development goals within the financial realm.

Green credit is the earliest form of green financial instrument and currently serves as a fundamental component of green finance [2]. The primary objective of green credit policy is to guide and allocate social capital by implementing differentiated loan policies, ultimately promoting the coordinated development of financial institutions, enterprises, and society [3,4]. In 2012, the China Banking Regulatory Commission (CBRC) officially issued the “Green Credit Guidelines” (Guidelines), which require financial institutions to strengthen credit incentives for green enterprises with environmental protection, social, and governance as evaluation indicators, according to relevant national economic and industrial policies. By providing financial support and risk protection through green credit policy, enterprises can pursue transformation and upgrading, thereby fostering high-quality and sustainable development [5].

The high-quality development of enterprises in China exemplifies the concrete implementation of the new development concept, which emphasizes innovation, coordination, greenness, openness, and sharing as its core principles. This approach not only facilitates the transformation of the economic structure but also provides significant support for the construction of an ecological civilization. However, during the process of transitioning the economic development model and striving for high-quality development, enterprises often encounter challenges in terms of funding, technology, and management. For example, enterprises require a large amount of funds to promote the transformation of their economic structure, but traditional financial institutions have relatively little support for green industries such as environmental protection, clean industry, and new energy industry, making it difficult to meet the needs of enterprises. Therefore, to expedite the transformation and upgrading of enterprises and achieve high-quality development, there is an urgent need for the support of green finance. Such support will provide institutional guarantees for the high-quality and sustainable development of enterprises through efficient resource allocation and funding mechanisms. It is also one of the motivations for this study.

With the continuous improvement in China’s financial system, existing studies have focused on the micro-level effects of the green credit policy on enterprises. It has been proved that the green credit policy can have a significant impact on technological innovation and green transformation of heavy polluters through the channels of credit constraints, agency costs, and investment benefits [6]. Empirical research conducted using high-pollution enterprises as a sample has indicated that the green credit policy contributes positively to enterprise innovation [7]. Additionally, scholars have suggested that the green credit policy can enhance the quality of enterprise innovation, thereby generating certain economic benefits for enterprises [8,9]. However, most of the previous literature takes environmental regulation policies as the entry point and is based on Porter’s green competition theory, which studies the impact of the green credit policy on enterprise technological innovation [10], investment [11], financing costs [12], pollution emissions [13], and financial institution loan interest rates [14]. There is little literature to explore the impact of the green credit policy on the high-quality development of enterprises as well as specific impact mechanisms from a micro perspective.

On the other hand, in the previous studies studying the high-quality development of enterprises, more scholars have explored the factors affecting the high-quality development of enterprises from the perspectives of the market environment [15], the market mechanism [16], and modern enterprise system [17]. However, fewer scholars have explored the impact of green credit policy on the high-quality development of enterprises from the perspective of environmental protection. Wu [18], taking listed manufacturing companies from 2008 to 2020 as a sample, examined the relationship between digital transformation, technological innovation, and the high-quality development of enterprises. The final result shows that both digital transformation and technological innovation can significantly promote the high-quality development of enterprises, and technological innovation has a mediating role. Zhang [19], based on the survey data of China’s business environment provided by the World Bank, conducts an in-depth study of the impact of business environment optimization on the high-quality development of the manufacturing industry. The research results show that business environment optimization has a significant promoting effect on the high-quality development of the manufacturing industry. Li [20] demonstrates through empirical research that economic policy uncertainty is detrimental to the high-quality development of enterprises. Therefore, it can be seen that the previous literature has rarely considered environmental protection when studying the high-quality development of enterprises.

In addition, nowadays there is some disagreement in the existing literature on whether the implementation of the green credit policy is beneficial to enterprise development. Some scholars argue that the implementation of the policy increases the difficulty of enterprise financing and reduces the loan amount available to enterprises, which will be detrimental to the development of enterprises. Zhong et al. [12] used the DID method to evaluate the impact of the green credit policy on firms’ debt-financing costs. The research results indicate that after the implementation of the green credit policy, banks have significantly increased the loan interest of high-polluting enterprises and increased their financing difficulties, which has a significant financing penalty effect on polluting firms. Xu et al. [21], using panel data from 52 green enterprises and 81 polluting enterprises in China from 2001 to 2007, employed the fixed effect model based on the Hausman test and the mediating effect analysis method to examine the asymmetric effects of green credit policy and its development on debt-financing costs for different types of firms. The research findings indicate that the green credit policy and its development significantly increase the financing costs of polluting enterprises, and these effects exhibit regional heterogeneity. Yao et al. [22] examined the effect of the green credit policy on the performance of listed firms in China, and the results show that the green credit policy significantly reduces the performance of heavily polluting enterprises by increasing their financing constraints and lowering their investment levels.

Meanwhile, another group of scholars, based on the “Porter hypothesis”, believes that proper environmental regulations are conducive to forcing firms to reform and innovate thereby achieving further development. Yu [23], from the perspective of external credit and internal environmental concerns, examined the impact of the green credit policy on corporate innovation, using panel data and DID models for listed A-share enterprises in Shanghai and Shenzhen from 2007 to 2020. The research findings reveal a significant increase in the number of green patent applications by heavily polluting enterprises after the implementation of the green credit policy. Furthermore, the enhancement of external business credit and internal environmental concerns serve as two important mediating mechanisms influencing corporate green innovation. Wang et al. [6] used the DID method to test the green innovation performance of green credit-restricted industries relative to non-green credit-restricted industries after the implementation of green credit policies. The research results found that after the implementation of green credit policies, the agency costs of enterprises will be reduced, and investment efficiency will be improved, which is conducive to enhancing green innovation performance in green credit-restricted industries. Chen et al. [24] used green credit guidelines as a quasi-natural experiment to explore the relationship between credit constraints and commercial credit, as well as the resulting corporate environmental governance effects. The research findings reveal that although the green credit policy has increased the difficulty of obtaining loans for enterprises, it has also adjusted their industrial structure, and increased their investment in environmental governance, which is beneficial for enhancing the market competitiveness of polluting enterprises.

In the above research, there is a lack of empirical research on whether the green credit policy can promote the high-quality development of enterprises, and very few scholars have explored this aspect, which also serves as one of the motivations for this paper. Furthermore, the aforementioned scholars have not provided a definitive answer to whether the green credit policy can promote enterprise development. As a result, some scholars support the government and relevant agencies in continuing to implement the green credit policy to facilitate enterprises’ green transformation and innovation. However, some scholars believe that the green credit policy has negative effects on enterprise performance. Although scholars have conducted diverse research on the impact of the green credit policy on enterprises, there is still a need for further supplementation in the existing literature.

To fill the above gap and enrich the existing literature, this research uses the quasi-natural experiment provided by the exogenous policy of the “Green Credit Guidelines” (Guidelines) issued by the China Banking Regulatory Commission (CBRC) in 2012 to test the causal relationship. Furthermore, this paper selects 23,618 listed industrial enterprises as the research sample and adopts the difference-in-differences (DID), moderating effect, fixed panel models, and panel quantile regression to examine the microeconomic effects of the green credit policy on the high-quality development of enterprises.

This paper attempts to answer the following questions: Does the implementation of the Guidelines in 2012 impact the high-quality development of enterprises? How large is the actual effect of the policy? Are enterprises with different degrees of development quality affected by the Guidelines in the same way? Will the effect be heterogeneous due to other factors? Does the effect of upgrading human resources in enterprises contribute to the positive effect of the Guidelines on the quality of development of enterprises?

There are several justifications for conducting this study. Firstly, this study addresses an important issue of promoting the green transformation and development of enterprises in the context of the dual carbon goal, which is a crucial national policy in China. Secondly, this paper proves the effectiveness of the green credit policy, which has important policy implications for promoting sustainable economic growth in China. Thirdly, this study identifies the heterogeneity of the promotion effect of the green credit policy on different types of enterprises, which can guide policymakers in designing and implementing more targeted policies to promote the green transformation and development of enterprises. Lastly, this study explores the role of human resource upgrading in contributing to the positive effect of the green credit policy on the high-quality development of enterprises, which can provide insights for enterprises looking to improve their human resource management.

Compared with the existing literature, the contributions of this paper are as follows: (1) In the current literature on green credit policies, fewer scholars have focused on the impact of green credit policies on the quality of enterprise development. At the same time, in the literature on the high-quality development of enterprises, there are almost no scholars who use the DID method to study the impact of green credit policies on the high-quality development of enterprises from the perspective of environmental protection. This study fills this gap through empirical research and provides a new research perspective and evidence for understanding the impact of green credit policies on enterprises. (2) Previous literature on whether the green credit policy can promote enterprise development has yielded controversial results. This paper once again proves through empirical research and various robustness tests that the green credit policy can promote enterprise development and enhance the quality of enterprise development, providing a strong empirical basis for policymakers. (3) In the previous literature on the effects of green credit policies, the heterogeneity of enterprises in different quantiles is seldom considered, and the regression results are also vulnerable to extreme values, which lead to unstable results. To address these issues, this study employs an unconditional panel quantile regression model to analyze the impact of green credit policy on the quality of enterprise development at different quantiles. This approach yields comprehensive and robust research results, enabling policymakers to design more differentiated policy recommendations. (4) This paper analyzes heterogeneity in terms of ownership, size, and region of enterprises, revealing the differences in the impact of green credit policies on different types of enterprises, which provides an important reference for the implementation of green credit policies. (5) To maximize the effectiveness of the policy, this study explores the role of human resource upgrading in the impact of green credit policy. This finding not only enriches the research in related fields but also promotes the improvement in human resource management by enterprises and enhances the quality of their development.

The important results of this paper are as follows: (1) This paper confirms that the green credit policy can promote the high-quality development of heavily-polluting enterprises. (2) Through panel quantile analysis, this paper finds that the green credit policy has a significant positive impact on enterprises with higher development quality, but has no significant impact on enterprises with lower development quality. (3) This paper finds that after the implementation of the green credit policy, state-owned enterprises, large-scale enterprises, and enterprises located in western regions have significantly improved their development quality, while others have not been significantly affected by the green credit policy. (4) This paper proves that the upgrading of human resources in enterprises has strengthened the positive impact of the green credit policy on the high-quality development of enterprises.

The remainder of this paper is organized as follows. Section 2 shows the background for financial institutions and develops the theoretical hypotheses. Section 3 introduces the study’s empirical design. Section 4 presents the empirical results. The panel quantile analysis is conducted in Section 5. The heterogeneity analysis results are presented in Section 6. The moderating effect analysis of the effect of human resource upgrading by enterprises is shown in Section 7. The conclusions are proposed in Section 8.

2. Background and Development of Hypotheses

2.1. Background

China’s green credit policy can be divided into two main stages. The first stage primarily aimed to address the international “Equator Principles”. Specifically, in 1995, the People’s Bank of China issued the “Notice on Implementing Credit Policies and Strengthening Environmental Protection Work”. This notice mandated that financial institutions at all levels incorporate national environmental protection policies into their credit operations and utilize credit policies as a vital tool for environmental protection, participating in comprehensive decision-making processes regarding economic development. In the second stage, the green credit policy witnessed rapid development. In January 2012, the China Banking Regulatory Commission officially released the “Green Credit Guidelines” (Guidelines). These guidelines require financial institutions to implement differentiated credit policies for enterprises with significant environmental pollution risks, enforce real-name management for enterprises with significant social and environmental risks, and withhold credit from enterprises with noncompliant environmental and social performance. Moreover, credit granting can be terminated for projects that have already been approved but pose significant environmental hazards. The issuance of the Guidelines signifies that the country has recognized the value of the market-oriented green credit policy and marks a new phase in its development, specifically in promoting energy conservation and emission reduction within enterprises. It is also regarded as the first domestic regulatory document dedicated to green credit and forms the core of China’s green credit system [25,26].

2.2. Research Hypothesis

The slogan “We need both green waters and green mountains, as well as gold and silver mountains” signifies China’s commitment to achieving high-quality economic development while simultaneously preserving the ecological environment. However, the policies implemented in China, primarily centered around administration, taxation, and technology, have not successfully achieved environmental optimization goals and have had limited impact on promoting the green transformation of enterprises [6]. Consequently, financial instruments play a crucial role in fundamentally eliminating environmental pollution and attaining goals related to ecological and environmental management.

The green credit policy is a fundamental component of green finance, aiming to restrain the uncontrolled expansion of high-polluting enterprises and encourage their transition and upgrade towards environmentally friendly practices, thus enhancing the quality of their development [27]. Different to the common mandatory and market-based environmental regulations, the green credit policy can be considered a combination of financial policies and environmental regulatory tools, representing a novel complement to traditional environmental regulatory policies [28]. From the perspective of the quality of enterprise development, the green credit policy promotes the development quality of enterprises through both “incentives” and “penalties”. Firstly, it provides more favorable loan terms, such as lower interest rates or longer repayment periods, to enterprises that meet environmental standards, significantly reducing financing costs and generating more business opportunities for them [29]. Enterprises that acquire green credit can demonstrate their social responsibility, thereby establishing a positive corporate image, attracting consumers and investors, and enhancing their development quality [30]. Additionally, the green credit policy incentivizes enterprises to invest in environmental technologies and resource reduction, facilitating a decrease in their negative environmental impact, avoiding environmental violations, and ultimately reducing environmental risks [31]. Secondly, as an environmental regulatory tool, the green credit policy exerts a “penalty effect” by imposing higher interest rates and limitations on financing for enterprises that neglect environmental protection measures or face heightened environmental risks. Consequently, these enterprises encounter difficulties in obtaining sufficient financial support, limiting their potential for growth and expansion. To secure bank loans, these enterprises are compelled to undergo a transformation, pursue technological innovation, and ultimately improve the quality of their development [5]. Hence, based on the above information, we make Hypothesis 1.

Hypothesis 1.

The implementation of the green credit policy contributes to the high-quality development of enterprises.

Currently, environmental issues are becoming increasingly prominent, and more and more enterprises have incorporated green development into their strategies. However, enterprises with different levels of development quality are affected differently by the green credit policy. Enterprises with higher-quality development are more willing to take social responsibility [32], are more aware of the importance of protecting the environment, and are more proactive in energy conservation and emission reduction [33], which also makes them more willing to actively accept the green credit policy. Secondly, enterprises with higher-quality development usually have more advanced environmental protection technologies and products, while at the same time they have fewer environmental risks [34], and thus they are more likely to meet the criteria for the green credit policy. In addition, enterprises with higher development quality usually have a better financial status and more stable cash flows [35,36], thus making it easier to obtain green credit. In short, enterprises with higher development quality have better environmental awareness, technological levels, and financial performance, which makes it easier for them to obtain green credit. It also promotes their transformation to a low-carbon, environmentally friendly, and sustainable economy, ultimately achieving high-quality development. Therefore, the following research hypothesis is proposed.

Hypothesis 2.

Compared with enterprises with lower development quality, the implementation of the green credit policy significantly improves the development quality of enterprises with higher development quality.

The implementation effect of the green credit policy may be influenced by the scale, property rights, and location of enterprises. Firstly, based on the classification by property rights, state-owned enterprises have greater advantages in terms of guarantees and financing compared to private enterprises, and they bear more social responsibilities and national policy-oriented tasks. Therefore, financial institutions provide more adequate credit funds to state-owned enterprises [12]. State-owned enterprises, as the backbone of China’s socio-economic development and the “vanguard” of the construction of a national ecological civilization, to a certain extent, represent the government as its “spokesperson”. Therefore, state-owned enterprises have a stronger willingness to protect the ecological environment. State-owned enterprises also have greater financial strength and technological advantages, and a higher quality of disclosure concerning environmental information [37]. Compared to private enterprises, state-owned enterprises have more comprehensive environmental management systems, such as well-developed organizational structures and personnel-allocation mechanisms. These factors contribute to the strong potential for green transformation in state-owned enterprises [38]. Secondly, from the perspective of enterprise scale, larger enterprises often have a greater repayment ability and lower financing constraints, making it easier to obtain bank credit support. The “punishment effect” of the green credit policy is also more significant for larger polluting enterprises [12]. Larger enterprises generally have more mature technology and have a certain ability to control the market, which can ensure the profitability of the enterprise. However, small-scale enterprises are generally unwilling to assume environmental responsibility or engage in technological innovation due to their high operational risks [39]. Finally, due to the vast land area of China, there are significant differences in resource conditions and a severe imbalance in regional development. With the introduction of the green credit policy, financial institutions determine whether to provide credit based on the environmental performance of enterprises. Due to the previous emphasis on economic development and less attention to environmental pollution control, the problem of environmental pollution in the western region of China is very severe [40]. Furthermore, the eastern and central regions of China have higher levels of economic development and a more favorable financial market environment compared to the western region. In contrast, the development of the financial market in the western region is relatively less, and enterprises in this region have a stronger reliance on bank credit [41]. Therefore, we propose Hypothesis 3.

Hypothesis 3.

Compared to other enterprises, state-owned enterprises, large-scale enterprises, and enterprises in the western region are more affected by the green credit policy.

Human resources are of strategic importance to the development of enterprises and are an important part of their strategic resource management. In the era of the green economy, human resources have become the key to enterprises and play a vital role in the long-term development of enterprises. As the quality of employees improves and daily management becomes more scientific and rational, enterprises will reduce environmental damage, lower pollutant emissions, and ultimately enhance their environmental performance in the production process [42]. The upgrading of human resources in enterprises can help them obtain better policy benefits from the “incentive” environmental regulation system, such as R&D expenses deductions, tax incentives, and direct financial subsidies. This not only greatly enhances the motivation of enterprises to implement green innovation but also ensures better performance in green innovation [43]. The decisions made by leaders can influence employees’ environmental behavior and environmental values [44]. High-quality leaders generally possess diverse professional qualities and practical experience. They can quickly make scientific and appropriate judgments in the face of complex and ever-changing external environments, promoting the implementation of sustainable development concepts and technological innovation in enterprises [45]. They can not only help enterprises cope with complex and dynamic environmental issues but also enable them to see the long-term and strategic competitive advantages of green environmental protection [46]. Therefore, we propose Hypothesis 4.

Hypothesis 4.

The effect of upgrading human resources in enterprises strengthens the positive effect of the green credit policy on the high-quality development of enterprises.

3. Empirical Design

3.1. Data Sources

As China’s green credit policy conducted in 2012 is an exogenous variable, this study selects industrial enterprises listed in Shanghai and Shenzhen A-shares from 2009–2019 as the research samples. ST, ST*, PT, and PT* companies, financial and insurance companies, companies with missing main-variable data, companies with a gearing ratio greater than 1 and less than 0, and companies with other abnormal financial indicators were deleted to ensure the reliability and consistency of the sample data. In the end, 3540 companies were included, totaling 23,618 valid observations. The data source of this research mainly includes the following three parts: corporate data and corporate financial data were all from CSMAR and WIND databases. Regional characteristic data were obtained from the China Statistical Yearbook and China Urban Statistical Yearbook. Missing data for a particular year were collected manually from the RESSET database and the China Economic Network.

3.2. Variable Selection

3.2.1. Explained Variable: Labor Productivity

The explained variable in this paper is the development quality of the enterprise. Generally speaking, both total factor productivity (TFP) and labor productivity (LP) can be used to measure the development quality of real enterprises [47,48]. Although TFP is more comprehensive and covers more information, it is susceptible to the impact of measurement methods and parameter settings. Although LP is not as comprehensive as TFP, its measurement method is simpler and more comparable. Moreover, it has been proved that there is a long-term stable positive relationship between TFP and LP [49]. Therefore, drawing on the previous literature [50,51,52], LP is adopted in this paper to measure the quality of real enterprise development, calculated by dividing an enterprise’s gross industrial product by the number of employees [52].

3.2.2. Explanatory Variable: The Green Credit Policy

This study regards China’s green credit policy as an exogenous explanatory variable, which is expressed as an interaction between ${Policy}_{it}\ast {Treat}_{it}$. When an industrial enterprise belongs to heavy-polluting firms, the value of ${Treat}_{it}$ is 1. Otherwise, it is 0. The list of heavy-polluting firms is based on the Guidelines on Industry Classification of List Firms revised by the China Securities Regulatory Commission in 2012, and the List of Listed Firms for Environmental Verification Industry Classification in 2008 [11,26]. A total of 14 industries are classified as heavy-polluting industries including: thermal power, steel, cement, electrolytic aluminum, coal, metallurgy, building materials, mining, chemicals, petrochemicals, pharmaceuticals, light, textiles, leather (The exact codes of heavy-pollution industries are: B06, B07, B08, B09, B11, C17, C18, C19, C22, C25, C26, C27, C28, C29, C30, C31, C32, C33, D44). ${Policy}_{it}$ represents the time dummy variable, which is bounded by the implementation of the green credit policy in 2012. The value of ${Policy}_{it}$ for 2012 and beyond is 1, otherwise, it is 0, $i$ represents different enterprises, and $t$ represents time. In the rest of this paper, for simplicity, the acronym “$did$” will be used to represent ${Policy}_{it}\ast {Treat}_{it}$.

3.2.3. Control Variables

To exclude interference from other economic variables that will affect the effect of the policy, referring to the existing literature [12], this paper also utilizes a series of control variables at enterprise and provincial levels. On the one hand, at the enterprise level, drawing on the existing literature [12,20,23], this paper uses controls for the following variables: enterprise size (SIZE), enterprise liabilities (LIABILITY), enterprise longevity (AGE), asset-liability ratio (LEVERAGE), operating revenue (OR), net profit (NP), asset cash ratio (CFO), net profit margin of total assets (ROA), Tobin Q (TQ), government subsidies (SUBSIDY), the largest holder rate (HOLD). On the other hand, at the provincial level, referring to the existing literature [53,54], this paper also uses controls for the following variables: level of local economic development (In_PERGDP), level of foreign investment (FDI), talent capital (EDU), population size (POP), industrial structure (IND), consumption capacity (WAGE). The variable definitions are presented in

Table 1. Variable description.

Variable Category	Variable Name	Symbol	Variable Description
Explained variable	Labor Productivity	LP	Operating incoming/The number of employees (logarithmic)
Explanatory variable	Treat	Treat	When an enterprise belongs to heavy-polluting firms, the value of Treat is 1; otherwise, it is 0.
	Policy	Policy	If the sample year is 2012 and later, post = 1, otherwise post = 0
Control variables at the enterprise level	Enterprise size	SIZE	Total assets at the end of the year (logarithmic)
	Enterprise liabilities	LIABILITY	Total liabilities at the end of the year (logarithmic)
	Enterprise longevity	AGE	Year of survey (2012)-the year of establishment of the enterprise + 1 (logarithmic)
	Asset-liability ratio	LEVERAGE	Total liabilities at the end of the year/Total assets at the end of the year
	Operating revenue	OR	Operating revenue at the end of the year (logarithmic)
	Net profit	NP	Net profit at the end of the year (logarithmic)
	Asset cash ratio	CFO	Total cash at the end of the year/Total assets at the end of the year
	The net profit margin of total assets	ROA	Net profit/Average total assets
	Tobin Q	TQ	Tobin Q from the CSMAR database one of the indicators to measure the growth of enterprises (logarithmic)
	Government subsidies	SUBSIDY	Subsidy from government (logarithmic)
	Largest holder rate	HOLD	The shareholding ratio of the largest shareholder
Control variables at the provincial level	Level of local economic development	In_PERGDP	Per capita GDP of the province where the enterprise is located (logarithmic)
	Level of foreign investment	FDI	Total import and export of goods by foreign-invested enterprises (logarithmic)
	Local education level	EDU	The population at the college level and above in the province where the enterprise is located (logarithmic)
	Population size	POP	Total population per province at the end of the year (logarithmic)
	Industrial structure	IND	Total industrial value/Total output value
	Purchasing power	WAGE	Total wages of employees in active positions/Number of employees in active positions (logarithmic)
Mechanism variable	The proportion of highly-skilled employees in each enterprise	HIGH	The number of employees with a college degree or above/Total number of employees

.

3.2.4. Mechanism Variable

Referring to the previous literature [55], this paper divides enterprise employees with a college degree or above into highly-skilled laborers, and the rest into low-skilled laborers. The proportion of highly-skilled laborers in enterprises is used as an indicator to measure the upgrading of the enterprises’ human resources, which will explore whether the upgrading of human resources significantly contributes to the positive impact of the green credit policy on the high-quality development of enterprises.

3.3. Descriptive Statistics

As shown in

Table 2. Descriptive statistics of main variables of main variables.

	Variable	N	Mean	Std. Dev	Min	Max
Control group	LP	15,191	13.77	0.95	9.93	18.92
	DID	15,191	0	0	0	0
	Policy	15,191	0.81	0.39	0	1
	Treat	15,191	0	0	0	0
	AGE	15,191	16.39	5.95	0	61
	OR	15,191	21.37	1.49	14.78	27.98
	NP	15,191	18.74	1.51	10.34	24.87
	TQ	15,191	2.12	6.49	0.68	729.6
	CFO	15,191	0.19	0.15	−0.06	1
	SUBSIDY	15,191	16.26	1.74	2.61	22.57
	SIZE	15,191	22.08	1.33	15.72	28.34
	LIABILITY	15,191	21.03	1.75	13.59	28.06
	LEVERAGE	15,191	0.41	0.21	0.01	0.99
	ROA	15,191	0.06	0.05	0	3.59
	HOLD	15,191	3.46	0.47	0.79	4.49
	In_PERGDP	15,191	11.12	0.47	9.24	12.01
	IND	15,191	0.36	0.09	0.07	0.53
	EDU	15,191	3.28	0.57	−0.2	4.08
	FDI	15,191	15.88	1.82	2.22	17.9
	POP	15,191	8.54	0.64	5.67	9.35
	WAGE	15,191	11.11	0.4	10.09	12.02
Treated group	LP	8427	13.81	0.8	8.86	18.77
	DID	8427	0.79	0.41	0	1
	Policy	8427	0.79	0.41	0	1
	Treat	8427	1	0	1	1
	AGE	8427	16.49	5.43	1	41
	OR	8427	21.61	1.43	15.74	28.72
	NP	8427	18.83	1.58	12.61	25.74
	TQ	8427	2.06	2.06	0.7	102.4
	CFO	8427	0.16	0.13	−0.02	0.87
	SUBSIDY	8427	16.26	1.69	7.6	23.23
	SIZE	8427	22.15	1.31	17.81	28.64
	LIABILITY	8427	21.06	1.72	15.46	27.88
	LEVERAGE	8427	0.4	0.2	0.01	0.99
	ROA	8427	0.06	0.06	0	1.56
	HOLD	8427	3.49	0.46	−1.24	4.49
	In_PERGDP	8427	10.93	0.5	9.24	12.01
	IND	8427	0.38	0.09	0.07	0.53
	EDU	8427	3.21	0.71	−0.2	4.08
	FDI	8427	15.01	2.44	2.22	17.9
	POP	8427	8.54	0.67	5.67	9.35
	WAGE	8427	11	0.39	10.09	12.02

, the total sample size is 23,618, including 15,191 samples in the control and 8427 samples in the treated group, which indicates that the treated group accounts for 35.68% of the total samples. In addition, it is clear that the difference in LP between the control group and the treated group is not substantial, and the std. dev. of LP in both groups is high, especially in the control group, whose std. dev. is 0.95. It is indicated that Chinese enterprises’ labor productivity is extremely uneven.

Table 3. Pearson correlation coefficients of main variables.

	LP	did	AGE	OR	NP	TQ	In_PERGD	|IND	EDU	FDI	POP
LP	1
DID	0.053 ***	1
AGE	0.176 ***	0.120 ***	1
OR	0.508 ***	0.084 ***	0.170 ***	1
NP	0.355 ***	0.049 ***	0.096 ***	0.722 ***	1
TQ	−0.051 ***	−0.006	−0.003	−0.122 ***	−0.085 ***	1
In_PERGD	0.172 ***	−0.005	0.209 ***	0.062 ***	0.123 ***	−0.013 **	1
|IND	−0.130 ***	−0.042 ***	−0.148 ***	−0.104 ***	−0.133 ***	−0.002	−0.414 ***	1
EDU	−0.064 ***	0.003	0.041 ***	−0.031 ***	−0.022 ***	−0.013 **	0.098 ***	0.484 ***	1
FDI	0.047 ***	−0.136 ***	0.033 ***	−0.026 ***	0.022 ***	−0.008	0.614 ***	0.175 ***	0.592 ***	1
POP	−0.109 ***	0.007	0.001	−0.056 ***	−0.055 ***	−0.012 *	−0.115 ***	0.606 ***	0.931 ***	0.461 ***	1

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

reports the Pearson correlation coefficients. However, due to space limitations, only the correlation coefficients of some main variables are reported. It can be shown that the correlation coefficient between LP and DID is 0.053, and it is significant at the 1% level, which indicates that the green credit policy increases an enterprises’ labor productivity. In other words, it improves the quality of enterprise development.

3.4. Model Construction

Difference-in-Difference (DID) is used to infer the effects of an intervention by comparing the differences between the treated group (intervened) and the control group (not intervened) before and after the intervention. More specifically, the core of DID is to divide the sample into a control group (before/after the intervention) and a treatment group (before/after the intervention), and then calculate the differences within each group before and after the intervention, and compare the difference-in-difference [56]. The difference-in-difference is DID, and this is also considered as the effect of the intervention. This method can eliminate the inherent differences and time effects between the two groups, thereby obtaining the net effect of the intervention on the treated group. It is generally widely used to assess the effects of policies, programs, or interventions on specific groups [57,58]. Compared with other methods, DID is simpler and more intuitive in form and can avoid the problem of model endogeneity caused by selection bias to a certain extent [59]. At the same time, as policies are generally exogenous shocks, the DID model can well avoid the problems of model endogeneity due to explanatory and explained variables being causally related to each other.

Therefore, this paper adopts the DID model to construct a quasi-natural experiment to study the impact of the green credit policy on the high-quality development of enterprises. The basic premise is this: due to the release of the “Green Credit Guidelines” in 2012, financing of high-polluting enterprises will be affected. This paper, taking 2012 as the dividing point, treats enterprises restricted by Guidelines from 2009 to 2019 as the treated group, while other enterprises not affected by Guidelines are included in the control group. We then measure the differences in outcome variables between the two subsamples by controlling other relevant variables. The specific model settings are as follows:

$$\begin{gathered} did={Policy}_{it}\ast {Treat}_{it} \\ LP={\beta }_0+{\beta }_1\ast did+\theta \ast \sum x+{\mu }_i+{\gamma }_t+{\lambda }_j+{\epsilon }_{it} \end{gathered}$$

(1)

where $i$, $t$, $j$ represent enterprise, year, and region respectively and the $did$ in this model is the core explanatory variable, which is expressed as the interaction item between ${Policy}_{it}$ and ${Treat}_{it}$. $LP$ is the explained variable calculated by dividing the operating incoming by the number of employees. In the following empirical research, the impact of the green credit policy on the high-quality development of the enterprise is further explored. ${\mu }_i,$ ${\gamma }_t,{\lambda }_j$ represent the individual, time, and regional fixed effects, respectively. $\sum X$ represents a series of control variables and ${\epsilon }_{it}$ stands for the residual item. This paper mainly focuses on the coefficient ${\beta }_1$. If ${\beta }_1$ is significantly greater than 0, it shows that the policy has significantly promoted the high-quality development of industrial enterprises.

4. Empirical Result

4.1. Baseline Model Result

The estimation results of Model 1 are shown in

Table 4. The impact of the green credit policy on enterprises’ LP.

	(1)	(2)	(3)	(4)
	LP	LP	LP	LP
DID	0.190 ***	−0.0188 *	0.0304 *	0.0306 **
	(0.0113)	(0.0100)	(0.0156)	(0.0152)
Constant	13.70 ***	7.865 ***	6.774 ***	7.634 ***
	(0.0137)	(0.355)	(1.978)	(2.068)
Control variable	NO	YES	YES	YES
Year FE	NO	NO	YES	YES
Individual FE	NO	NO	YES	YES
Regional FE	NO	NO	NO	YES
Observations	23,618	23,618	23,398	23,398
R-squared	0.0028	0.2986	0.8721	0.8741

Note: Standard errors are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

. Column (1) shows the regression result without control variables, Year FE, Individual FE, and Regional FE. The coefficient ${\beta }_1$ is 0.190 and is significant at the 1% confidence level. Column (2) shows the regression with control variables. Although the coefficient ${\beta }_1$ in column (2) is negative, its significance level is not very high and it is not significant economically because it is too small. Furthermore, there is no Year FE, Individual FE, and Regional FE in column (2). According to Column (3), after adding control variables, Year FE, and Individual FE, the estimated coefficient is 0.304, and the estimated coefficient is significant at the 10% confidence level. In column (4), control variables, Year FE, Individual FE, and Regional FE are all added. The coefficient ${\beta }_1$ is 0.0306, which is significant at the 5% confidence level. The estimation in column (4) shows that the implementation of the green credit policy has significantly promoted the high-quality development of heavily polluting enterprises, which verifies hypothesis 1. Columns (3) and (4) both adopt the order of Reghdfe to regress, therefore some samples are dropped because it will automatically eliminate single point values based on fixed effects.

4.2. Robust Tests

4.2.1. Parallel Trend Test

The parallel trend test is an important prerequisite for using the DID method to investigate the policy effect. Referring to previous references [6,47], the mean of the sample size is used to perform the parallel trend test, and the test result is as Figure 1. Figure 1 shows the parallel trend of labor productivity in the control and treated groups from 2009 to 2019, with the horizontal axis representing the year and the vertical axis representing the average labor productivity of the sample firms for each year. The blue line represents the mean of the treated group and the red line represents the mean of the control group. The dot line represents the year of policy implementation, namely 2012. Taking the policy issued in 2012 as the boundary, we divided the sample period into two stages: before policy implementation (2009–2012) and after policy implementation (2012–2019). From Figure 1, it can be seen that, before 2012, the changing trend of labor productivity in the control group and the treated group was almost consistent, and the difference was not significant. However, after 2012, the gap between the treated group and the control group gradually widened, and the labor productivity of the treated group experienced a significant increase, and while the labor productivity of the control group also increased, the increase was not as obvious. Therefore, the parallel trend test is satisfied.

The event analysis method is another way to test the assumption of parallel trends [60,61]. Following previous references [62,63], the following model is constructed, where ${did}_{it}^k$ represents the virtual variable of the impact of the green credit policy. The policy was implemented in 2012, therefore $k=t-2012$. Coefficient ${\beta }_k$ represents the change in the LP of enterprises from the first two periods before the policy intervention to the last two periods after the treatment. Coefficient ${\beta }_0$ denotes the effect of the current treatment period. Other variables in model (2) are the same as in model (1). If coefficient ${\beta }_k$ was not significant before the implementation of the policy, then the fact that there is no significant difference between the control group and the treated group before policy implementation is proved, and the parallel trend assumption is passed. The regression result of model (2) is produced in Figure 2.

$$LP=\alpha +{\sum }_{k=-2}^{k=2}{\beta }_k\ast {did}_{it}^k+\phi \ast \sum X+{\mu }_i+{\gamma }_t+{\lambda }_j+{\epsilon }_{it}$$

(2)

As presented in Figure 2, the coefficient ${\beta }_k$ is not significantly different from 0 before the policy (the 95% confidence interval contains a value of 0), which indicates that there is no significant difference between the treated group and the control group before the implementation of the policy. The parallel trend test is satisfied.

4.2.2. Change the Time Window

Before the implementation of the policy, the Chinese government noticed the strategic significance of green finance and successively adopted related credit policies. The green credit policy in 2012 is the most comprehensive and influential relative to other policies. Therefore, to exclude the disturbance of some other policies affecting the accuracy of the empirical results, referring to the previous literature [23], this paper assumes that 2010 is the year of implementation of the green credit policy, and constructs a counterfactual test based on this for regression analysis. Because the policy was not implemented in 2010, when the estimated result of the core explanatory variable coefficient is not significant, it is considered to have passed the counterfactual test.

Table 5. Robustness Test of Changing Policy Time.

	(1)
	LP
DID	−0.000955
	(0.0156)
Constant	7.766 ***
	(1.565)
Control variable	YES
Year FE	YES
Individual FE	YES
Regional FE	YES
Observations	23,398
R-squared	0.878

Note: Standard errors are in parentheses. *** p < 0.01.

indicates the estimated result. The estimated coefficient of DID is not significant, indicating success.

4.2.3. PSM-DID

Owing to large differences in the size and financial status of sample enterprises and the regions where enterprises are located, the regression result may be subject to the problem of sample selection error. To solve this problem, the PSM-DID test is used for robustness testing. The control variables at the enterprise and province levels are used as covariates. This paper uses the logit model to estimate the probability of each sample being selected in the treated group and then, respectively, adopts the nearest-neighbor matching method and kernel matching method to match the treated group with a reasonable control group. As shown in Figure 3 and Figure 4, after the nearest-neighbor matching, the standardized deviation of all variables is less than 10% and the nuclear density difference between the treated group and the control group is significantly smaller. The same conclusions can also be drawn from Figure 5 and Figure 6. After kernel matching, there is no significant difference in each characteristic variable between the treated group and the control group, which indicates that kernel matching is effective. Column 1 in

Table 6. Regression results of PSM-DID.

	(1)	(2)
	LP	LP
DID	0.0262 *	0.0262 *
	(0.0144)	(0.0144)
Constant	8.088 ***	8.184 ***
	(1.883)	(1.883)
Control variable	YES	YES
Year FE	YES	YES
Individual FE	YES	YES
Regional FE	YES	YES
Observations	23,384	23,382
R-squared	0.878	0.878

Note: Standard errors are in parentheses. *** p < 0.01, * p < 0.1.

shows the PSM-DID estimation results based on the nearest-neighbor matching, and column 2 in Table 6 shows results based on the kernel matching. Both results show that the coefficients of DID are significantly positive at the 10% level.

4.2.4. Placebo Test

Considering that random factors may also affect the accurate estimation of policy effects, this paper conducts a placebo test by randomization. In the randomization, referring to the previous paper [64], we randomly select enterprises as the treated group for the test. To enhance the validity of this test, we repeat this random process 500 times and obtain the corresponding samples. Then the DID regression is performed according to model 1. Finally, the results are shown in Figure 7 and Figure 8. From Figure 7 and Figure 8, it can be shown that the 500 estimated coefficients of the false DID are all distributed near 0, and most of the coefficients do not pass the significance test. Therefore, we confirm that the baseline results are not accidental.

5. Quantile Regression

To further present the impact of the green credit policy on the high quality of heavy-polluting enterprises, this study employs a panel quantile regression model for estimation. Furthermore, to enhance the robustness of the estimation, the unconditional quantile regression estimator is also used. The final results of the regression analysis are presented in

Table 7. The panel quantile regression.

	RIF (Q25)	RIF (Q50)	RIF (Q75)	RIF (Q90)	RIF (Q99)
DID	0.0347	0.0572 **	0.0743 **	0.0886 *	0.508 ***
	(0.0225)	(0.0224)	(0.0303)	(0.0486)	(0.164)
Control variable	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES
Individual FE	YES	YES	YES	YES	YES
Regional FE	YES	YES	YES	YES	YES
Observations	23,398	23,398	23,398	23,398	23,398

Note: Standard errors are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

.

Based on the findings presented in Table 7, it can be observed that, at the 0.25 quantile, the influence of the green credit policy on the labor productivity of heavily polluting enterprises is not statistically significant. However, for quantiles ranging from 0.5 to 0.99, the green credit policy exhibits a significant positive impact on labor productivity in heavily polluting enterprises. Notably, both the magnitude and significance of the coefficients display a notable increasing trend. These results suggest that the positive effect of the green credit policy on the high-quality development of heavily polluting enterprises primarily manifests in enterprises with labor productivity situated within the middle and high quantiles. Moreover, the positive effect progressively strengthens as labor productivity increases, whereas its impact on enterprises with labor productivity within the lower quantiles is not statistically significant.

The possible reasons for this situation may be as follows. Firstly, enterprises with low development quality generally do not require large-scale environmental investment. In addition, they generally have a narrow business scope and do not operate high polluting and energy-consuming businesses. Secondly, enterprises which have low development quality and belong to high-pollution industries are relatively small in scale. They generally create less revenue, have high operational and financial risks, lack innovative technology and talent, and have relatively simple financing channels, thus they can only rely on bank loans. Commercial banks, on the other hand, have stricter loan-approval procedures and set relatively low loan amounts for them due to the operational risks, therefore the green credit policy has less impact on them.

For enterprises with high development quality, their business scale is relatively large, their business scope is broader, and they are more likely to operate within high pollution and highenergy-consumption industries. At the same time, because these enterprises have more resources and technological advantages, they can better afford to invest in environmental protection, and they are also more likely to obtain the trust and support of banks, thus these enterprises have higher loan amounts and faster loans. In addition, to improve their reputation and establish a good corporate image, enterprises with high development quality often choose initiatives to protect the environment. Due to the greater social influence of these enterprises, the government usually imposes stricter regulatory requirements on them. Overall, enterprises with higher development quality are more capable of assuming social responsibility and protecting the environment. The green credit policy has a greater impact on these enterprises. That is the reason why the positive effect of the green credit policy gradually increases with the increase in the labor productivity of enterprises.

6. Heterogeneity Analysis

6.1. Analysis of Corporate Heterogeneity

The previous analysis shows that the green credit policy has a significantly positive impact on the high-quality development of enterprises. However, the behavioral response to the policy varies by enterprise attributes. Based on this background, this paper distinguishes enterprises with different property rights and enterprises of different sizes to explore the impact of the green credit policy. The exploration of this issue will help us understand the micro-effects of the green credit policy under different scenarios.

To further examine the impact of policy on enterprises with different property rights, a difference-in-difference-in-difference (DDD) model will be used to test the heterogeneity of property rights. The specific regression model is set as model 3. $H_{it}$ represents heterogeneous variables in property rights, which is the focus of this DDD model. The remaining variables are consistent with model 1 and the regression results are shown in column 1 of

Table 8. Analysis of corporate heterogeneity.

	SOE	Large-Scale	Small-Scale
DDD	0.200 ***	0.0899 ***	−0.000495
	(0.0309)	(0.0241)	(0.0202)
Constant	7.143 ***	11.73 ***	10.41 ***
	(2.069)	(2.989)	(2.858)
Control variable	YES	YES	YES
Year FE	YES	YES	YES
Individual FE	YES	YES	YES
Regional FE	YES	YES	YES
Observations	23,398	11,581	11,440
R-squared	0.875	0.892	0.885

Note: Standard errors are in parentheses. *** p < 0.01.

. The results show that the regression coefficient of the DDD model is 0.2 and is significant at the 1% level, indicating that the development quality of state-owned enterprises is more affected by the policy than that of non-state-owned enterprises.

The likely economic explanations are as follows: Firstly, generally speaking, the goal of modern private enterprises is to maximize shareholder wealth, but China’s SOEs have both economic-level goals, such as profit and other operating performance indicators, and political-level goals, such as promoting coordinated regional development, and social-level goals, such as stabilizing employment and ensuring the supply of public goods and services. As a result, Chinese SOEs are more motivated to implement the credit policy proactively and the positive impact of the policy on SOEs is also more pronounced [65]. Secondly, for SOEs, access to enjoy various biased policies in their production and operation is one of the important characteristics of the operation of Chinese enterprises. According to the survey data of the 2017 Report on the Business Environment Index of China’s Provincial Enterprises, SOEs are significantly better treated than non-SOEs in areas closely related to the government, such as policy openness, justice in law enforcement, and administrative efficiency [66]. Finally, when local governments face assessment pressure while allocating more social resources to SOEs, they will also transmit greater assessment pressure to SOEs. The combination of “carrot” and “stick” will have a greater impact on SOEs [67].

Considering that there are differences in enterprises of different sizes, based on the previous literature, this paper divides the sample data into two subsamples for regression analysis: large-scale enterprises (enterprises with asset sizes greater than the median total assets of the sample enterprises) and small-scale enterprises (enterprises with asset size less than the median total assets of the sample enterprises). The results obtained are shown in columns (3) and (4) of Table 8. The results show that the green credit policy has a significant positive impact on the development quality of large-scale heavily-polluting enterprises, while the impact on the development quality of small-scale heavily-polluting enterprises is negative and not significant.

The possible explanations for this are: Firstly, large-scale enterprises are more likely to obtain bank loans, and the number of loans is larger, therefore they are more affected by the policy. Small-scale enterprises, on the other hand, have more difficulty in obtaining bank loans and at the same time have smaller loan amounts. As a result, the credit policy has very little impact on small-scale enterprises [11]. Secondly, there are significant differences in how enterprises of different sizes use credit funds. Large-scale enterprises are more likely to use credit funds to upgrade their technology, enhance their technological innovation capabilities, and further promote their development quality after obtaining bank credit because of their abundant resources and strong strength. As for small-scale enterprises, their viability is relatively low, their survival space is relatively limited, and their resistance to risk is relatively poor, therefore they are more likely to use credit funds for daily operations rather than technological innovation after obtaining bank credit. Therefore, the promotion effect of credit funds on the development quality in small-scale enterprises is not significant.

$$LP={\beta }_0+{\beta }_1\ast did\ast H_{it}+{\beta }_2\ast did+{\beta }_3\ast H_{it}+\theta \ast \sum X+{\mu }_i+{\gamma }_t+{\lambda }_j+{\epsilon }_{it}$$

(3)

6.2. Analysis of Regional Heterogeneity

Institutional environmental and economic development generally varies from region to region. These objective factors may affect the effectiveness of policy implementation. Thus, to test the heterogeneous effects of region, this paper classifies enterprises into three sub-samples: eastern, central, and western, for regression analysis based on the differences in the economic development level of the location of enterprises. The eastern part includes Beijing, Tianjin, Shanghai, Hebei, Shandong, Jiangsu, Zhejiang, Fujian, Guangdong, Hainan, Heilongjiang, Jilin, and Liaoning. The central region includes Shanxi, Anhui, Henan, Hubei, Hunan, and Jiangxi. The west includes Inner Mongolia, Chongqing, Sichuan, Yunnan, Guizhou, Tibet, Shaanxi, Guangxi, Gansu, Qinghai, Ningxia, and Xinjiang. Columns (1), (2) and (3) in

Table 9. Analysis of regional heterogeneity.

	East	Central	West
DID	0.0264	−0.0202	0.0813 **
	(0.0195)	(0.0325)	(0.0377)
Control variable	6.841 ***	3.445	9.928 *
	(2.221)	(8.729)	(5.329)
Control variable	YES	YES	YES
Year FE	YES	YES	YES
Individual FE	YES	YES	YES
Regional FE	YES	YES	YES
Observations	16,885	3346	3150
R-squared	0.876	0.885	0.858

Note: Standard errors are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

show the results of the regression for the eastern, central, and western samples, respectively. From the regression results, it can be found that the regression coefficients of the subsamples in the eastern and central regions are not significant, while the coefficient of the subsamples in the western region is 0.083, and is significant at the level of 5%. This indicates that the green credit policy can significantly improve the development quality of heavily polluting enterprises in the western region, while the impact on the central and eastern regions is not significant.

The likely economic explanations are as follows. Firstly, the financial industry is more developed in the eastern and central regions, where enterprises can not only obtain credit funds from commercial banks but also carry out capital turnover through other financing channels. Thus, the green credit policy has less impact on the heavily polluting enterprises in the eastern and central regions. In contrast, the financial industry in the western region is relatively backward and capital is scarce, and enterprises rely heavily on commercial banks. Therefore, the green credit policy has a significant impact on western regions. In addition, the eastern and central regions of China are more developed in terms of level of technology, and most enterprises are new technology-intensive enterprises, while the west is relatively backward in terms of level of technology, and most of the enterprises are resource-intensive and labor-intensive traditional enterprises; as a result, the western region is indeed more significantly affected by the bank credit policy.

7. Further Analysis

The above research results indicate that the green credit policy can significantly promote the improvement of labor productivity in heavily-polluting enterprises. In other words, the green credit policy is beneficial to improve the high-quality development of highly heavily-polluting enterprises. Next, this section will discuss whether the effect of upgrading human resources in enterprises can strengthen the positive effect of the green credit policy on the high-quality development of highly heavily-polluting enterprises. Drawing on previous literature [55], in this paper, the employees with a college degree or above are considered highly-skilled workers, and other employees are considered low-skilled workers. The proportion of high-skilled workers in each enterprise is calculated as an indicator to measure the upgrading of human resources in the enterprise. This proportion is then multiplied with the core explanatory variable to construct a new model for regression analysis. The specific regression model is set as model 4. ${HIGH}_{it}$ represents the proportion of highly-skilled employees in each enterprise, which is the focus of this model. The remaining variables are consistent with Model 1 and the regression results are shown in

Table 10. Analysis of the Moderating Effect.

	(1)
	LP
DID_HIGH	0.0523 ***
	(0.0183)
Constant	7.514 ***
	(2.013)
Control variable	YES
Year FE	YES
Individual FE	YES
Regional FE	YES
Observations	22,394
R-squared	0.885

Note: Standard errors are in parentheses. *** p < 0.01.

. From the regression results, it can be seen that the coefficient ${\beta }_1$ of the cross-term $did\ast {HIGH}_{it}$ is positive and significant at the 1% level. It indicates that the effect of upgrading human resources in enterprises strengthens the positive impact of the green credit policy on the high-quality development of enterprises.

The possible economic explanation is that the higher the education level of the employees in an enterprise, the higher the quality of the employees will be, and they will be more receptive to new things and better able to learn. In addition, they are more able to consider the interests of the enterprise from a long-term perspective and are more willing to actively cooperate with the government to assume corporate social responsibilities. Therefore, when the green credit policy was issued, enterprises with high-quality employees were more proactive in implementing it. As a result, these enterprises are more affected by the policy.

$$LP={\beta }_0+{\beta }_1\ast did\ast {HIGH}_{it}+{\beta }_2\ast did+{\beta }_3\ast {HIGH}_{it}+\theta \ast \sum X+{\mu }_i+{\gamma }_t+{\lambda }_j+{\epsilon }_{it}$$

(4)

8. Conclusions

Against the background of the increasingly severe sustainable development and green transformation in China, the green credit policy has received extensive attention in recent years. It guides the green development of enterprises through the allocation of credit resources and is an important policy innovation that is used to promote the transformation and upgrading of enterprises. Its close relationship with high-quality development in enterprises has not only attracted the attention of government policymakers but also attracted the attention of enterprises as an important financing channel. Accordingly, based on the data of 23,618 listed Chinese industrial enterprises from 2009 to 2019, this research focuses on the high-quality development of enterprises, taking the introduction of the green credit policy in 2012 as a quasi-natural experiment and using the difference-in-difference method to empirically investigate whether the green credit policy can promote the high-quality development of heavily-polluting enterprises. Our findings have great theoretical and practical implications for green credit policy implementation, high-quality development of enterprises, and the sustainable development of society.

8.1. Theoretical Implications

(1) The positive effect of the green credit policy on the high-quality development of enterprises has been verified using the DID model, which supports Hypothesis 1. This positive effect is generally provided through the following two channels: Firstly, financial institutions restrict loans to heavily-polluting enterprises, which can inhibit them from investing in polluting industries and force them to change their economic development mode, and finally achieve high-quality development. Secondly, when heavily-polluting enterprises invest in environmental industries, financial institutions will provide them with lower loan interest rates and other financial support to help them successfully transform and upgrade.

The existing literature on the impact of green credit policies on enterprise development focuses more on enterprise innovation [68,69,70], economic performance [22,71], and environmental performance [71,72], with little attention paid to the impact of these policies on the development quality of enterprises. At the same time, there are certain controversies in the academic community regarding whether green credit policies can promote enterprise development: positive promotion [73,74], and negative promotion [22,75]. These results are related to the samples and variable measurement methods chosen by scholars. This paper emphasizes that the main bodies of green credit policies are heavily-polluting industrial enterprises, which to some extent reduces sample-selection bias. In addition, in contrast to previous literature [18,20], this paper uses labor productivity to measure the development quality of enterprises, which is more consistent with the characteristics of industrial enterprises. In summary, this paper provides a new research perspective on the micro impact of green credit policies on enterprises.

(2) After the implementation of the green credit policy, compared to enterprises with lower development quality, enterprises with higher development quality are more positively affected by the green credit policy, which proves Hypothesis 2. The green credit policy usually sets some indicator requirements in terms of environmental protection and energy consumption. Enterprises with higher development quality generally perform better in energy saving and environmental protection and have greater environmental awareness, making it easier for them to meet the requirements for green credit. Therefore, the incentive effect of green credit policy on enterprises with higher development quality is greater.

Previous studies on the quality of enterprise development did not divide the quality of enterprise development into different levels, ignoring the heterogeneity of enterprise development quality at different quartiles, which tends to cause the estimation results to be affected by the extreme values of the sample, thus potentially leading to inaccurate estimation results [19,50]. To enrich this type of research and fill certain gaps, this paper considers the development level of different enterprises when analyzing the impact of the green credit policy on the high-quality development of enterprises. The panel quantile regression model is used, and the unconditional regression analysis method is adopted, which can more accurately and comprehensively estimate the impact of the green credit policy on enterprises with different levels of development.

(3) The heterogeneity analysis is carried out from three aspects: enterprise property rights, enterprise scale, and region of the enterprise. This paper found that after the implementation of the green credit policy, state-owned enterprises, large-scale enterprises, and enterprises located in the western regions of China have significantly improved their development quality, while the development quality of non-state-owned enterprises, private enterprises, and enterprises in the central and eastern part of China has not been significantly affected by green credit policy. These findings support Hypothesis 3.

Generally speaking, state-owned enterprises, large-scale enterprises, and enterprises located in the western regions of China have more resources and capabilities in environmental protection and energy conservation. State-owned enterprises are usually relatively advantaged in terms of resources and policies, and can more easily invest in the field of environmental protection and energy conservation. Larger enterprises have more financial and technological resources, which enables them to better implement environmental protection and energy conservation measures. Enterprises located in western China may receive more policy support and attention, enabling them to enjoy relevant policy dividends and thus have a certain advantage in environmental protection and energy conservation. Therefore, the green credit policy has a greater positive impact on the development quality of these enterprises.

(4) Finally, scholars have not yet analyzed the micro role of human resource upgrading in enterprises in this process. Therefore, to fill the gap and enrich such research, this paper takes the proportion of high-quality talents in enterprises as an indicator of their human resource upgrading and explores the micro effects. The results show that when the human resources of enterprises are upgraded, the positive effect of green credit on the high-quality development of enterprises is more significant, which is further in favor of Hypothesis 4. When the human resources of enterprises are upgraded, employees’ skills, knowledge, and quality are improved, and the enterprises’ capabilities in environmental protection and energy conservation are also enhanced. Therefore, when introducing green credit policies, enterprises are more capable of obtaining corresponding credit support and better implementing environmental and energy-saving measures, thereby achieving high-quality development.

8.2. Policy Implications

According to the results concluded above, it is clear that the green credit policy plays a significant role in promoting the high-quality development of enterprises and environmental protection. The analysis of it is significant, and some policy recommendations are proposed in this paper.

(1) Specific measures the government can take include the following: Firstly, continue to strengthen the implementation of the green credit policy, increase the proportion of green credit in the total credit allocation, and provide more financial support for the high-quality development of enterprises. Secondly, the government should promote regional coordinated development, provide more support to the western region in the implementation of the green credit policy, strive to narrow the regional development gap, and promote balanced development in the eastern, central, and western regions of China. For non-state-owned enterprises and small-scale enterprises, the government can strengthen policy support and guidance, and encourage them to increase investment in environmental protection and research and development of green technologies, which aim to better meet the requirements of green credit. Furthermore, the government can optimize the design of green credit policy, lower credit thresholds, increase loan speeds, and increase credit support for these enterprises. Finally, the government can provide free human resources training and consulting services to enterprises to maximize the positive effect of the green credit policy, helping enterprises improve employee skills and innovation ability.

To sum up, the government should raise the environmental awareness of society as a whole and actively establish the concept of green and sustainable development as well as raising the concept of economic structural transformation and high-quality development of enterprises to a strategic level.

(2) Financial institutions should continue to improve the implementation mechanism and establish a tracking and evaluation mechanism to ensure the effective implementation of the policies. Secondly, financial institutions should adopt differentiated credit policies based on the development quality of enterprises, for instance by prioritizing green credit support for enterprises with higher development quality and providing more favorable credit interest rates for these enterprises. Moreover, financial institutions should strengthen credit assessment and risk management for enterprises with higher development quality to reduce credit risks. Thirdly, more credit support should be provided to state-owned enterprises, large-scale enterprises, and enterprises in the western regions that meet the criteria to promote their further high-quality development. For other enterprises, financial institutions may consider adjusting the green credit policy appropriately and redesigning green credit products, which will enhance the positive impact of the green credit policy on these enterprises. Finally, based on the positive impact of human resources upgrading on the high-quality development of enterprises, financial institutions should consider the following exact recommendations to further strengthen the positive effect of green credit policy: Provide customized green credit products that are specifically designed to support the human resources upgrading of enterprises (for instance, financial institutions could offer preferential interest rates or longer repayment periods for loans used for human resources upgrading). Strengthen the credit evaluation system for enterprises, particularly in evaluating the impact of human resources upgrading on their overall development. Priority will be given to providing green credit support to enterprises with good human resource management, and certain preferences will be given in terms of credit interest rates and repayment terms.

(3) Enterprises should make full use of the green credit policy to change their economic development mode and actively transform to achieve their high-quality development. Specific measures are as follows: enterprises need to strengthen their awareness of energy conservation and environmental protection, increase investment in science and technology, and improve the quality of their development to receive support from green credit. State-owned enterprises, large-scale enterprises, and enterprises in the western region of China should pay more attention to investment in environmental protection and energy conservation, strengthen cooperation with the government, and strive for more policy support and capital investment. Other enterprises need to strengthen their building of capacity and increase their emphasis on energy conservation and environmental protection to enhance their sustainable development level. Finally, enterprises need to strengthen human resource management, improve employees’ skills and qualities, and pay attention to employees’ environmental awareness and social responsibility, which will better obtain green credit and achieve high-quality development of the enterprise.

8.3. Limitations and Directions for Future Research

This study has some limitations that may lead to further research. Firstly, the sample time interval of this paper is only up to 2019, which may limit the accurate examination of the long-term effects of the policy. Future research could expand the time interval of the sample to further investigate the long-term effects of green credit on the high-quality development of enterprises.

Secondly, the sample data in this study only includes data from listed industrial enterprises and does not include data from unlisted small and medium-sized enterprises, as well as financial and insurance enterprises. Therefore, future research can attempt to collect more comprehensive data for a more thorough analysis.

Finally, this paper mainly explores the impact of the green credit policy on the high-quality development of enterprises, but it does not delve into the specific transmission mechanism of the green credit policy on the high-quality development of enterprises. The green credit policy forces polluting enterprises to transform and upgrade by increasing their credit rates, and future research can explore the specific transmission mechanism of the policy starting from bank credit rates.