1. Introduction
Chinese President Xi Jinping announced in September 2020 that China will achieve peak carbon by 2030 and carbon neutrality by 2060, paving the way for China to embrace the green transformation of the manufacturing sector more aggressively. With this as a backdrop, China has issued the Green Credit Guidelines (hereafter referred to as the Guidelines), in February 2012, which has drawn researchers’ attention to examine the Guidelines’ impact on firm performance. However, the available literature on the effect of the Guidelines mainly focuses on the disincentive impact on financial performance [
1,
2]. There is a lack of research on whether the Guidelines may lead to a green transformation and the effects of such transformation. This study aims to fill this gap by exploring the relationship between the Guidelines and green innovation. Green innovation, defined as “technological innovation activity that promotes green technology and improves the ecological performance”, is crucial to evaluate firms’ environmental performance [
3,
4].
The role of green credit in green innovation is mostly based on two theoretical mechanisms: the “compliance cost effect” and the Porter’s hypothesis. The compliance cost effect is that environmental regulations will increase production costs and pollution control costs, which will have a crowding-out effect on R&D investment activities, and thus reduce green innovation, while Porter’s hypothesis suggests that reasonable environmental regulations can encourage green innovation, which in turn will improve production management efficiency and enterprise competitiveness, offsetting the increased costs brought about by environmental regulations [
5,
6]. These two contradictory theories inspired much research on the impact of the Guidelines.
The Guidelines, which are China’s first regulatory guideline on green credit, mandate that banks and financial institutions use enterprises’ environmental and social risks as the credit operation threshold and actively encourage green credit. In particular, the Guidelines use the incentive and constraint effects of credit to internalize the external costs associated with business environmental emissions. First, the credit constraint effect increases the threshold restrictions and transaction costs for backward manufacturing to receive credit financing; second, the credit incentive effect is to give interest rate assistance for green innovation involving green transformation and increase clean investment. As opposed to earlier environmental regulatory policies characterized by administrative penalties such as sewage charges, the most essential purpose of the Guidelines is to encourage heavily polluting firms to engage in green innovation through the allocation of credit resources.
Lately, researchers are interested to study the mediating mechanism between the Guidelines and green innovation. R&D investment continues to be a go-to mediator in academic research [
7]. However, R&D investment is considered more from the perspective of technical innovations involving backward production, making it an unsuitable metric for green innovation. Environmental investment, which may promote pollution control, emissions reduction, cleaner production, and green technology creation, is a reasonable choice compared to R&D investment. However, environmental regulations, and how strict they are, affect every investment decision a firm makes [
8]. Therefore, as the Guidelines will inevitably affect corresponding changes in environmental investment by heavily polluting enterprises, will this effect be facilitative or inhibitive? Will facilitated or hindered environmental investment promote or diminish green innovation in heavily polluting enterprises? In other words, does environmental investment serve as a mediator between the Guidelines and green innovation? Is it completely mediated or partially mediated?
This study uses a DID model and a sample of Chinese listed industrial enterprises from 2009 to 2017 to empirically investigate the impact of the Guidelines on green innovation in heavily polluting enterprises, as well as the mechanisms via which this impact is exerted. The result of baseline regression reveals that the Guidelines have a significantly positive effect on a firm’s green innovation. The dynamic regression results show the volatility of green innovation while testing the parallel trend hypothesis. The findings remain consistent even after addressing any endogeneity issues through different regression techniques, particularly the improved PSM-DID method. The mediating effects model is used to reveal the partial mediating role of environmental investment. Additionally, the Moderation model results indicate that the positive influence of green credit policy is more substantial in state-owned firms and in regions with high levels of green finance development.
This study may have the following marginal contributions: First, environmental investment’s role as a mediator between the Guidelines and green innovation in heavy polluters is disclosed for the first time, and our findings expand on the green credit literature. Second, an examination of heterogeneity demonstrates that the effect of the Guidelines is influenced by the type of ownership and the level of regional green financial development, which helps to identify the variation in the effects of the Guidelines. Ultimately, our findings inform the development of green credit policy to encourage green innovation and advance sustainable development.
Following a literature review in
Section 2, the research hypothesis is presented in
Section 3, the foundational models, variables, and data are introduced in
Section 4, the empirical analysis is described in
Section 5, the mediating mechanism is analyzed in
Section 6, the moderating effect is analyzed in
Section 7, and the implications and future directions are presented in
Section 8.
4. Research Design and Statistical Analysis
4.1. Sample Selection and Data Sources
This study includes all listed industrial firms between 2009 and 2017 as its initial sample, excluding those having gearing ratios of less than 0 or larger than 1, non-length arm’s listings (including ST, ST*, and PT), and those with missing relevant data. We matched the patent classification numbers acquired from the China Research Data Service Platform (CNRDS) with the International Patent Classification Green List produced by WIPO in 2010. Patents are sorted into two groups, “green” and “non-green”, based on the match outcome, with the former broken down further into invention and utility model patents. The main financial data comes from the CSMAR database and the shareholder data are from the RESSET database. A total of 7911 yearly data were gathered after applying tail reduction at the 1% and 99% levels to the major continuous variables.
4.2. Variable
Corporate green innovation (
Patent). Referring to Li and Lu (2015) [
44], this study uses the number of green patent applications to measure green innovation. The total number of green innovations (
Total) is calculated by adding the numbers of patent applications for green inventions and green utility models; the former provides a measure of the green innovation’s quality (
Inva) while the latter provides an indication of its quantity (
Uma). To fix the data’s right-skewed distribution, we take the natural logarithm to get
LnTotal,
LnInva, and
LnUma, respectively.
An enterprise’s dummy variable (Treat). According to the Guidelines, the former CBRC has specified the types of environmental and social risks in the Key Evaluation Indicators for the Implementation of Green Credit, and the industries with environmental and social risks of Category A are identified as heavily polluting industries in this study. The nuclear power industry, hydroelectric power industry, water conservancy, inland river and port engineering construction, coal mining and washing, oil and gas mining, ferrous metal mining and selection, non-ferrous metal mining and selection, non-metallic mining and selection, and other mining industries make up the Class A industries. If an enterprise belongs to the above 9 industries, it is identified as a heavily polluting enterprise with (experimental group), otherwise (control group).
A period dummy variable (Post). The Guidelines were officially published in February 2012, and the years before 2012 are used as the experimental period in this study. If the sample year is after 2012, Post = 1, otherwise .
Control variables. Following previous studies [
45,
46], in this study, enterprise size (
Size), the ratio of assets to liabilities (
Debt), proportion of fixed assets (
Ppe), return on assets (
ROA), growth (
Growth), cash flow (
Cash), equity level (
Topl), enterprise age (
Age), employee count (
Employee), independence of the board (
Board), and Tobin Q (
TQ) are selected as control variables.
Table 1 displays descriptive statistics for the primary variables.
Appendix A explains in detail what each variable means.
4.3. Model
In order to effectively mitigate the problems of endogeneity and omitted variables, this study analyzes the impact by constructing a DID model, as follows:
In Equation (1), denotes the firm and denotes the year. With enterprises needing time to file patent applications, we choose Patent with a lag of one period as our explanatory variable. The control variable is represented by X, the firm fixed effect by , the year fixed effect by , and the random disturbance term by . reflects the impact of the Guidelines on green innovation in heavy polluters.
6. The Mediating Effect of Environmental Investment
This study examines the possible environmental investment channel behind the Guidelines to encourage green innovation in highly polluting enterprises, and with a significant
of model (1), constructs a mediating effect model below:
is the mediating variable, denoting environmental investment, and total environmental investment and green transition investment are denoted by
TEi and
GEi, respectively. Drawing on [
44],
TEi and
GEi were manually collected from the debit additions related to environmental protection in the construction-in-progress detail item table of the corporate annual report.
GEi includes both investments in green industries and investment in green upgrades to traditional production. Green industry investment is judged according to whether the investment project belongs to the green industries and their subheadings in the Green Industry Guidance Catalogue (2019 Edition) issued by the National Development and Reform Commission; green upgrading investment is defined according to the Green Manufacturing Project Guidelines (2016–2020) issued by the Ministry of Industry and Information Technology on the green upgrading of traditional manufacturing. Therefore, enterprises’ investments in five areas, including green industries, clean production, efficient and low-carbon use of energy, efficient use of water resources, and recycling of resources, are all considered green transition investments.
When both and in Equations (3) and (4) are significant, if is also significant, and the absolute value of is smaller than the absolute value of , then there is a partial mediating effect of ; if is not significant, then there is a full mediating effect of ; if at least one of and is not significant, a Bootstrap test is required; if the test result is significant, the mediating effect is considered significant; otherwise, it is not significant.
6.1. Total Environmental Investment
The results of the test for the mediating effect of
TEi are shown in
Table 5. The coefficient of
in column (1) is significantly positive, indicating that the Guidelines significantly increase the
TEi of heavily polluting enterprises. Meanwhile, the coefficients of
in columns (2), (3), and (4) are all significantly positive at the 1% significant level, provided that the
TEi coefficient is significant, indicating that the Guidelines promote green innovation through
TEi. Moreover, the absolute value of the
coefficient is smaller than the absolute value of
, indicating that the mediating impact of
TEi is only partial.
6.2. Green Transition Investment
The results of the test for the mediating effect of
GEi are shown in
Table 6. The coefficient of
in column (1) is significantly positive, indicating that the Guidelines significantly increase the
GEi of heavily polluting enterprises. Meanwhile, the coefficients of
in columns (2), (3), and (4) are all significantly positive at the 1% significant level, provided that the
GEi coefficient is significant, indicating that the Guidelines encourages green innovation via
GEi. Same as the analysis as for
TEi, the partial mediating effect of
GEi is demonstrated.
In summary, both in terms of the quality and quantity of green innovation, the Guidelines contribute to heavy polluters via
TEi and
GEi partially. Research hypothesis H2 is confirmed. This implies that the effect of the Guidelines on green innovation further affects green innovation through the indirect effect of environmental investment, in addition to the direct effect. Similarly, Ma et al. (2022) [
52] also verified that environmental investment plays a partially mediating role in the relationship between internal control and corporate green innovation. Our added value lies in the splitting of environmental investments into
TEi and
GEi, with a better research framework.
8. Conclusions and Policy Implications
This research experimentally explored the connection between the Guidelines and green innovation using data from publicly listed Chinese industrial enterprises from 2009 to 2017 and analyzed the mediation role of environmental investment. The following conclusions are drawn. First, the Guidelines can effectively encourage green innovation in heavily polluting enterprises, and the increase in quantity is more significant compared to quality. Second, environmental investment, and particularly investment in green transition, partially mediates the Guidelines’ influence on green innovation. This implies that, in addition to the direct effect of the Guidelines on green innovation, the indirect effect of environmental investment also has an impact on green innovation. Third, the moderating effect of enterprise ownership and the regional green finance development level is verified, and they enhance the implementation effect of the Guidelines.
This paper theoretically enriches the literature on the effects of green credit policy and the factors influencing corporate green innovation, and also provides important insights for improving the implementation of green credit guidelines. First, strengthen the Guidelines’ current implementation. It is more important to correctly identify green initiatives than different categories of businesses. Assistance should be provided to significantly polluting firms who want to make environmental investment for the green transformation of the industrial structure. Second, take into account the diversity of enterprises and geographic areas to moderate the uneven impact of the Guidelines. Green credit should take enterprises of different ownership and regions into account, such as extending to regions with low green financial development, private enterprises, and highly competitive industries.
Based on the findings of this study, the following are the avenues for future research. Firstly, while this study has explored the mediating impact of environmental investment, future research should explore the impact of other factors on green innovation, such as green agency costs, which arise from the conflicting interests of environmental regulation between the principal as shareholder and the agent as manager. Secondly, this paper simply extracts the green transition investment from the total environmental investment to represent how much heavy polluters are investing in pollution control at the source. Future research could find a better way of separating environmental investment into green transition investment and end-of-pipe treatment investment so as to distinguish if firms are favoring pollution prevention at the source instead of end-of-pipe treatments. Lastly, this study has focused on the Chinese context, and future research could explore whether the findings can be generalized to other countries with different political and economic systems.