How Does Green Insurance Affect Green Innovation? Evidence from China

Abstract

In the context of the rapid development of green finance, this paper examines the impact of green insurance on green innovation. Based on panel data of listed firms from 2008 to 2020, we find that green insurance significantly increased firms’ green patent applications. The mechanisms driving this positive relationship between green insurance and green innovation include that the insured firms are able to obtain more resources, are more willing to take risks, and are more likely to have a long-term vision. Further analysis shows that green insurance can enhance a firm’s environmental performance by promoting green innovation. This study deepens our understanding of green insurance and enriches the research related to green finance.

1. Introduction

In the face of severe environmental and climatic challenges, firm behaviors that give rise to severe social and environmental problems have become an increasing cause for concern [1,2]. Green innovation is considered an effective way to achieve sustainable growth while simultaneously reducing negative environmental impacts [3,4,5], and ways to promote green innovation have received increasing attention from scholars and policymakers.

Since green innovation is risky, costly, and long-term [4,6,7], whether a firm chooses a green innovation strategy tends to depend on whether it can effectively integrate green innovation activities into its economic activities. Several recent studies show that green finance, such as green credit, green funds, and green bonds, can provide financial support for a firm’s green developments, thus promoting their green innovation [3,8,9,10,11]. Building on this research, in the present paper, we focus on another important but understudied green financial instrument—i.e., green insurance [12,13,14]—and investigate its impact on green innovation, which helps to enrich green finance research from the perspective of green insurance.

In contrast to green financial instruments, which provide direct financial support [8,13], green insurance integrates insurance institutions into the firm’s environmental management. These insurance institutions are more familiar with the relevant environmental regulations and standards and have accumulated rich environmental management and industry knowledge based on their long history of environmental assessments of similar enterprises. More important is that insurance institutions integrate the insured firm’s environmental risks into their own risk management systems through an insurance contract [4,8], which results in their involvement and interest in the firm’s environmental management activities, and environmental supervision and inspection. Existing research shows that stakeholders with strong expertise and high levels of engagement in the firm’s environment management play an increasingly important role in promoting firms’ environmental responsibility. Further, these stakeholders are more likely to demand positive and substantive environmental actions [15,16]. However, few studies investigate the impact of professional stakeholders, such as green insurance, on firms’ green innovation.

We predict that the high level of expertise and deep involvement of insurance institutions in firms’ environmental management will promote insured firms’ green innovation. This is mainly because, first, insurance institutions not only promote the diffusion and application of green knowledge and technology but also promote the firm’s green image and enable it to attract more resources; second, insurance institutions can effectively reduce the insured firm’s environmental pressures by preventing and sharing their environmental risks, which increases the insured firm’s risk taking and encourages adoption of more proactive solutions to address environmental problems; third, green insurance institutions aimed at sustainable development can encourage the firm to adopt a long-term orientation and invest in long-term projects. Therefore, we predict that green insurance helps insured firms to incorporate environmental improvement actions into their existing economic activities, thereby promoting firms’ green innovation.

The potential contributions of this paper are mainly in the following two aspects: first, this paper significantly advances green finance research from the perspective of green insurance. In contrast to the existing studies, which focus on green credit, green funds, and green bonds [8,9,17,18], we try to show how green insurance affects the insured firms’ green innovation through the intensive and professional involvement of an insurance institution. Second, our study adds to our knowledge about the determinants of corporate green innovation [19,20]. Specifically, green insurance allows the insured firms access to more knowledge and resources, increases their level of risk taking, and promotes a focus on long-term goals. These findings show that green insurance enables the insured firms to integrate green innovation in their economic activities.

The rest of the paper is organized as follows: Section 2 briefly reviews relevant literature and develops the main hypothesis, Section 3 describes the methodology and the sample, Section 4 reports the empirical results, Section 5 tests the potential mechanisms, Section 6 reports the results of a further analysis, and Section 7 concludes the paper.

2. Literature and Hypothesis

2.1. Literature Review

Green innovation refers to new or improved processes, practices, systems, and products that benefit the environment and help to reduce or avoid the negative effects of pollution [21,22,23]. Therefore, green innovation is considered a fundamental way to solve long-term climate problems and achieve sustainable development [24]. Since green innovation is costly, risky, and long-term [4,6,7], firms lack an incentive to engage in green innovation [25], and how effectively to promote green innovation is receiving increased attention from scholars and policymakers.

The greater awareness of environmental problems has triggered research on the factors influencing green innovation [26]. Some studies explore the impact on green innovation of external organizational factors, such as institutional pressure [27,28], customer demand [21,29], environmental groups [30,31], and environmental policy [32,33]. External pressure is forcing firms to take more responsibility for the environment by developing green products and improving green processes [34,35]. Some studies explore internal factors promoting organizational green innovation, such as managers [19,36], firm strategy [20,37], firm resources and capabilities [38,39], etc.

Existing studies have shown that green finance contributes to environmental improvements [9,13,17,25]. Some studies suggest that green finance instruments such as green credit, green funds, and green bonds are providing financial support for firms’ green developments and innovation activities [9,40,41]. For example, green credit provides funding for energy-saving, environmentally friendly, and eco-friendly firm projects through the provision of differentiated loan funds by banks [3,10]. Similarly, green bonds are considered reliable signals of a company’s commitment to the environment and provide information for investors about the firm’s environmental commitment [9].

Green insurance, as an important component of green finance, is used widely in environmental governance and risk control worldwide [14,42]. The literature suggests that green insurance can be effective for reducing or even avoiding environmental pollution events by improving the environmental management and strengthening the supervision of the insured firms [43]. The rapid development of green finance has resulted in investigations of the effects of financial instruments, such as green credit and green bonds, on green innovation [4]. However, research on the impact of green insurance on corporate green innovation is almost non-existent, with the exception of Wang and Nie [42], who investigate the innovation effect of green insurance based on a game theory model and case studies. In the present paper, we explore the influence of green insurance on green innovation.

2.2. Institutional Background

In order to control effectively for environmental risks, in 2013, China’s Ministry of Environmental Protection issued the Guidance Opinions on the Pilot Policy of Compulsory Liability Insurance for Environmental Pollution aimed mainly at those industries experiencing higher numbers of pollution accidents with major environmental consequences. The guidance not only clarifies the responsibilities and obligations of insurance institutions and insured firms but also elucidates the scope of green insurance, the insurance contract terms, rate determinations, risk assessment, etc. The implementation of green insurance policy improves environmental risk prevention, environmental governance, and increases the firm’s compensation in the case of an environmental accident.

In most cases, before a firm is insured, the insurance institution will conduct a comprehensive investigation and assessment of the firm’s environmental risks and reach a consensus with the insured firms about the potential environmental risks, insurance terms, insurance rates, etc., before signing an insurance contract. The final insurance contract sets out the responsibilities and obligations of both parties. During the period of insurance, the insurer will conduct inspections, training, and rectification to prevent potential environmental risks and accidents by the firm. For example, insured firms in Jiaxing City have to undergo an annual “environmental physical examination”. Between 2013 and 2017, there were 498 incidents of environmental experts being called on to provide on-site services, which resulted in identification of 5030 hidden dangers. If an environmental accident occurs, the insurance institution will investigate the cause of the accident, process the environmental accident claims, and oversee environmental restoration work. To sum up, green insurance involves the insurance institution in the firm’s environmental management. To avoid environmental accidents and maximize their interests, the insurance institutions participate actively in corporate environmental governance through environmental risk assessments, insurance terms, regular supervision, premium rates, and insurance claims, which serve to promote corporate environmental actions.

Following implementation of the Guidance Opinions, between 2014 and 2015, a total of 14,400 firms in China had subscribed to green insurance, to an insured amount of CNY 26.373 billion, equivalent to a nearly 93-fold increase in the risk protection capacity of the insured firms. In 2016, the Guidance Opinions on Building a Green Finance System issued by the Chinese government further encouraged and supported insurance institutions to innovate in green insurance products and services and suggested the direction for the development of green insurance.

2.3. Research Hypothesis

As mentioned above, since green innovation is risky, costly, and long-term [4,7,36], whether the firm chooses a green innovation strategy tends to depend on whether it can effectively integrate green innovation activities into its economic activities. We link green insurance with green innovation characteristics and reveal how green insurance integrates green innovation into corporate decision making. We expect that green insurance has a positive impact on green innovation for the following reasons:

First, green innovation activities require not only green knowledge but also sufficient resources [44], and green insurance can help the insured firm to acquire the necessary knowledge and financing resources. This is mainly because, on the one hand, the insurer accumulates a wealth of expertise and knowledge about environmental management practices in the long-term environmental assessments of large numbers of similar firms. These insurers can promote insured firms’ environmental management and innovation activities by providing industry-related green knowledge, such as relevant environmental technical knowledge. On the other hand, green insurance sends a positive signal to external stakeholders that the firm is environmentally aware and committed to building a green image in line with the environmental expectations of stakeholders, such as banks and investors [45,46,47]. This can help insured firms improve their environmental legitimacy, thereby reducing financing constraints and accessing the resources needed to innovate [27].

Second, green insurance can work to diversify the firm’s environmental risks and alleviate the firm’s preventive motives for future major environmental risks [42,48]. Existing research suggests that excessive concern about risk and uncertainty can decrease innovation output (e.g., Roper and Tapinos [1]; Ren and Wang [36]), and management of potential innovation risks has become crucial for green innovation. In our context, green insurance institutions help the insured firms to identify “blind spots” through comprehensive risk examination [49], which improves their anti-risk ability [43]. At the same time, green insurance allows the firm to improve its compensation liability in the case of an environmental pollution accident [43]. These risk-reducing mechanisms provide the firm more freedom to conduct green innovation activities, tolerate higher risks, and proactively explore new solutions.

In addition, green insurance contributes to a long-term corporate environmental investment orientation. On the one hand, insurance institutions integrate the firm’s environmental risks into their own strategic framework and encourage the firm to pay attention to environmental issues based on environmental protection, training, and supervision. This encourages the insured firms to commit to long-term sustainable development and fundamentally address environmental issues through green innovation. On the other hand, the (usually) long insurance contract term imposes on the firm long-term environmental constraints and compliance pressures, which are difficult for the firm to respond to through one-off or temporary pollution control solutions [50]. Green innovation, as a radical pollution solution to achieve a win–win economic and environmental outcome, is an effective option to cope with this persistent pressure from insurance institutions. Therefore, green insurance can stimulate firms to undertake more long-term efforts, such as green innovation, to achieve more effective and enduring solutions to pollution problems. Therefore, we hypothesize that (shown in Figure 1):

Hypothesis 1.

Green insurance has a positive impact on firms’ green innovation.

3. Methodology

3.1. Samples and Data

China’s Guidance Opinions on the Pilot Policy of Compulsory Liability Insurance for Environmental Pollution issued in 2013 provides us with an opportunity to study the impact of green insurance on green innovation. For our initial sample, we took 14,400 identified insured firms published on the Ministry of Environmental Protection official website. Our sample selection process continued as follows.

First, we restricted our sample to heavy polluting industries because these sectors include the majority of insured firms. Second, we matched the list of insured firms published by the Ministry of Environmental Protection with listed companies from the heavy polluting sectors (i.e., we matched the names and locations of insured firms with the names and locations of listed companies and their subsidiaries). This provided us with an insured listed company sample, which we used as our treatment group. As our control group, we used other uninsured listed companies in the heavy polluting industry sector. Finally, we excluded firms with missing values for financial data. We obtained a final sample of 12,496 firm–year observations.

3.2. Empirical Model

We model the impact of green insurance on green innovation as follows:

$${LnGP}_{ict}={\alpha }_0+{\beta }_1{Insurance}_{ict}+{\beta }_2{Controls}_{ict}+{Ind}_c+{Year}_t+{\epsilon }_{ict}$$

(1)

In the above equation, subscript i is the firm, c is the industry, and t is the year. The dependent variable $Ln{GP}_{ict}$ is green innovation. ${Insurance}_{ict}$ is a dummy variable for the firm’s participation in green insurance, and it equals 1 for an insured firm and 0 otherwise. The coefficient of ${\beta }_1$ reflects the impact of green insurance on green innovation. If ${\beta }_1$ is positively significant, this indicates that green insurance has a significantly positive influence on green innovation. ${Controls}_{ict}$ is a set of control variables that are described below. The variables ${Ind}_c$ and ${Year}_t$ are, respectively, industry and year fixed effects. ${\epsilon }_{ict}$ is the error term.

3.3. Variables

Dependent variable. Drawing on previous studies [44,50], green innovation is measured as the logarithm of the number of green patent applications plus 1. Based on China’s patent classification, green patents are further divided into green invention patents and green utility model patents [26]. In the robustness test, we use the horizontal patent value to measure the firm’s green innovation.

Independent variable. Green insurance (Insurance) is a dummy variable that is coded 1 if the sample firm is in the list of insurance firms published by China’s Ministry of Environmental Protection and 0 otherwise.

Control variable. Firm size (Size) is often considered an important determinant of firm-level innovation [47]. We use the logarithm of annual assets to measure firm size. Firms have different goals at different stages of development, and we therefore include firm age (Age) in our model [46]. In China, state-owned as opposed to private firms tend to assume more social and environmental responsibility [36]; we measure state ownership through dummy variables (SOE) and the firm’s performance (ROA) and cash flow (Cash) to measure the firm’s profitability and resource status. We control the ratio of shares owned by the largest investor to the total shares outstanding (TOP) and consider the proportion of independent directors on the firm’s board to measure board independence (Independence). The firm’s market performance is measured by firm’s earnings per share (EPS). Finally, we control for industry- and year fixed effects in the benchmark regression model.

3.4. Descriptive Statistics

Table 1. Descriptive statistics.

Variable	All Sample				Uninsured Firms		Insured Firms		t-Test
	Mean (1)	SD (2)	Min (3)	Max (4)	Mean (5)	SD (6)	Mean (7)	SD (8)	Difference (9)
LnGP	0.65	1.1	0	7.81	0.57	1.01	1.06	1.4	−0.50 ***
LnGIP	0.28	0.72	0	7.62	0.22	0.62	0.54	1.05	−0.32 ***
LnGUP	0.52	0.98	0	6.59	0.46	0.91	0.82	1.25	−0.37 ***
Insurance	0.17	0.38	0	1	0	0	1	0	−1
Size	22.1	1.35	14.76	28.64	21.97	1.25	22.72	1.58	−0.75 ***
Age	17.86	5.68	2	45	17.9	5.69	17.64	5.64	0.26 *
SOE	0.4	0.49	0	1	0.37	0.48	0.52	0.5	−0.15 ***
ROA	0.07	0.09	−0.2	0.7	0.07	0.09	0.07	0.08	0
Cash	3.03	0.07	2.13	3.28	3.03	0.07	3.06	0.08	−0.03 ***
Top	0.35	0.15	0	0.9	0.35	0.15	0.37	0.17	−0.02 ***
Independence	0.04	0.02	0	1	0.04	0.02	0.04	0.01	0.00 ***
EPS	0.36	0.76	−16.46	21.98	0.36	0.78	0.38	0.65	−0.02

Note: the above table shows the descriptive statistics of the sample variables. Column (9) reports the results of the t-test for all variables between the non-insured firms and the insured firms, with ***, and * representing significance levels of 1%, and 10%, respectively.

reports the descriptive statistics of the variables. Columns (1) to (4) show the mean, standard deviation, and minimum and maximum values of our variables. The average log of firm green patent applications is 0.65. Column (9) shows that the logarithms of patents among insured firms are significantly larger than those of uninsured firms. In addition, correlation coefficient analysis shows that there is a significant positive correlation between green insurance and firm green patents, but further empirical analysis is needed.

4. Empirical Analysis

4.1. The Impact of Green Insurance on Green Innovation

Table 2. The impact of green insurance on green innovation.

Variables	LnGP (1)	LnGIP (2)	LnGUP (3)	GP (4)	GIP (5)	GUP (6)
Insurance	0.166 ***	0.126 ***	0.119 ***	6.326 ***	5.266 ***	2.373 ***
	(0.026)	(0.020)	(0.023)	(2.070)	(1.675)	(0.637)
Controls	Y	Y	Y	Y	Y	Y
Constant	−8.215 ***	−4.685 ***	−6.558 ***	−119.991 ***	−44.373 ***	−67.300 ***
	(0.424)	(0.271)	(0.383)	(23.549)	(15.670)	(9.808)
Year fixed	Y	Y	Y	Y	Y	Y
Industry fixed	Y	Y	Y	Y	Y	Y
Observations	12,496	12,496	12,496	12,496	12,496	12,496
R-squared	0.322	0.236	0.306	0.224	0.173	0.168

Note: the standard errors clustered at the firm level are reported in parentheses. *** significant at 1%.

reports the estimated results for the impact of green insurance on green innovation. The dependent variables in columns (1) to (3) present the logarithms of number of green patent applications (LnGP), green invention patent applications (LnGIP), and green utility model patents (LnGUP), respectively. Column (1) shows that the coefficient of green insurance (Insurance) is significantly positive at the 1% level, indicating that green insurance promotes green innovation substantially. The estimated coefficient shows that green insurance contributes approximately 24.6% of the improvement, which is also significant in terms of economic significance. Similarly, the estimation results in columns (2) and (3) show that green insurance has a significantly positive effect on green invention patents and green utility model patents. In fact, some recent studies have found that green insurance has a positive impact on green innovation, such as Wang and Nie [42], Ning et al. [51], and Zhu et al. [52]. Overall, we provide firm-level new empirical evidence for green insurance and green innovation and enrich our understanding of the determinants of corporate green innovation.

We further use the horizontal value of green patents to measure firm green innovation. The estimated results are reported in columns (4) to (6). We find that green insurance significantly promotes insured firms’ green patents, green invention patents, and green utility model patents. In terms of the coefficient, green insurance increased 6.3 patent applications, 5.3 invention patents, and 2.4 utility model patents of insured firms. This evidence suggests that green insurance promotes insured firms’ green innovation, which supports our hypothesis.

4.2. Endogenous Analysis

4.2.1. Control for Time-Varying Industry Effects

Since omitted variables can lead to biased estimation results, we further conduct several supplementary tests to alleviate such concern. Although in our baseline regression model we control for industry-level fixed effects, some unobserved time-varying industry factors, such as technological developments in different industries and policy shocks, might be influencing our estimation and could lead to biased results [36]. To alleviate this concern, we include in the interaction dummies for industry×year to account for potential time-varying variables related to industry selection.

Table 3. Control for time-varying industry effects.

Variables	LnGP (1)	LnGIP (2)	LnGUP (3)
Insurance	0.165 ***	0.122 ***	0.121 ***
	(0.026)	(0.019)	(0.023)
Controls	Y	Y	Y
Constant	−8.366 ***	−4.732 ***	−6.653 ***
	(0.443)	(0.291)	(0.394)
Year fixed	Y	Y	Y
Industry fixed	Y	Y	Y
Industry×year	Y	Y	Y
Observations	12,496	12,496	12,496
	0.336	0.252	0.325

Note: the standard errors clustered at the firm level are reported in parentheses. *** significant at 1%.

reports the results for the above estimations. The estimation results controlling for time-varying industry trends suggest that green insurance boosts green patent applications significantly, which is consistent with the results of our baseline model. Excluding potential time-varying variables related to the industry alleviates our endogeneity concerns.

4.2.2. Difference-in-Difference Analysis

Our estimation results might be affected also by certain firm-level factors, such as unobservable factors related to our choice regarding insurance companies. We estimate a difference-in-difference (DID) model based on a quasi-natural experiment where we consider insured firms as the treatment group and uninsured firms as the control group. This allows us to compare changes to green innovation among insured and uninsured firms and helps to eliminate firm-level confounding factors.

Table 4. DID estimation result.

Variables	LnGP (1)	LnGIP (2)	LnGUP (3)
Treat×Post	0.341 ***	0.119 ***	0.333 ***
	(0.048)	(0.038)	(0.042)
Treat	0.006	0.100 ***	0.064 *
	(0.034)	(0.026)	(0.048)
Post	2.065 ***	2.376 ***	1.918 ***
	(0.131)	(0.131)	(0.247)
Controls	Y	Y	Y
Constant	−7.752 ***	−4.285 ***	−6.243 ***
	(0.412)	(0.254)	(0.375)
Year fixed	Y	Y	Y
Industry fixed	Y	Y	Y
Observations	12,496	12,496	12,496
	0.360	0.326	0.328

Note: the standard errors clustered at the firm level are reported in parentheses. ***, and * significant at 1%, and 10%, respectively.

reports the results for the above estimations. The results based on the DID model indicate that green insurance significantly promotes insured firms’ green patents, green invention patents, and green utility model patents.

4.3. Robustness Test

4.3.1. Alternative Models

Since green patents are non-negative and there are some years when many firms do not apply for any patents and a small number of firms account for a large number of patents, we re-estimate our baseline model using Tobit, Poisson, and Negative Binomial models [36].

Table 5. Alternative models.

Variables	LnGP (1)	LnGIP (2)	LnGUP (3)
Panel A: The results of Tobit model
Insurance	0.235 ***	0.290 ***	0.178 ***
	(0.051)	(0.063)	(0.055)
Panel B: The results of Poisson model
Insurance	0.126 ***	0.216 ***	0.107 ***
	(0.028)	(0.042)	(0.033)
Panel C: The results of Negative binomial model
Insurance	0.127 ***	0.210 ***	0.096 ***
	(0.029)	(0.045)	(0.034)
Controls	Y	Y	Y
Year fixed	Y	Y	Y
Industry fixed	Y	Y	Y
Observations	12,496	12,496	12,496

Note: *** significant at 1%.

reports the estimation results. Panel A presents the results of the Tobit model and shows that green insurance significantly promotes the green patent, green invention patent, and green utility model patent applications. Panel B presents the results for the Poisson model, showing that green insurance significantly and positively promotes green patent applications. Panel C provides similar results to our baseline model. Overall, our results are robust to different models.

4.3.2. Alternative Variables

We further check the robustness of our variable measures. First, since innovation activities show a certain lag [36], we run the regressions, including lagged independent variables. Second, to measure the dependent variables, total patents, invention patents, and utility model patents are used in turn as alternative measures of green patents.

Table 6. Alternative variables.

Panel A: Results of Lagged Independent Variables
Variables	LnGP (1)	LnGIP (2)	LnGUP (3)
Insurance	0.182 ***	0.131 ***	0.133 ***
	(0.028)	(0.021)	(0.025)
Controls	Y	Y
Year fixed	Y	Y	Y
Industry fixed	Y	Y	Y
Observations	11,052	11,052	11,052
	0.317	0.237	0.303
Panel B: Results of Alternative Dependent Variables
Variables	LnPatent (1)	LnIPatent (2)	LnUPatent (3)
Insurance	0.162 ***	0.240 ***	0.108 ***
	(0.034)	(0.031)	(0.033)
Controls	Y	Y	Y
Year fixed	Y	Y	Y
Industry fixed	Y	Y	Y
Observations	12,496	12,496	12,496
	0.389	0.354	0.402

Note: the standard errors clustered at the firm level are reported in parentheses. *** significant at 1%.

reports the estimation results. Panel A for the lagged variables presents results that are similar to our baseline model results. Panel B reports the estimation results for patents, invention patents, and utility model patents as alternative measurements. Overall, the results are robust.

5. Mechanism Analysis

In this section, we further analyze the potential channels for the influence of green insurance on green innovation, i.e., resources acquisition, risk taking, and long-term orientation. Note that these mechanisms are not necessarily mutually exclusive.

5.1. Resource Acquisition

As discussed above, green insurance can promote the insured firms’ resource acquisition. To test this mechanism, we first used the SA financing constraint index, where a higher value means a higher constraint on external financing [53]. We also consider firms’ financing scale and cost as alternative variables. Since green innovation requires long-term financial support, we use share of long-term borrowing in the firm’s total assets as the credit scale and measure financing cost as the firm’s average current interest expenditure divided by total debt in the current period [54].

Table 7. Impact of green insurance on resource acquisition.

Variables	SA (1)	Financing Sale (2)	Financing Cost (3)
Insurance	−1.144 ***	6.033 ***	−0.000
	(0.068)	(0.269)	(0.005)
Controls	Y	Y	Y
Year fixed	Y	Y	Y
Industry fixed	Y	Y	Y
Observations	12,491	12,496	8859
R-squared	0.601	0.366	0.004

Note: the standard errors clustered at the firm level are reported in parentheses. *** significant at 1%.

presents the estimation results. Column (1) presents the results for the effect of green insurance on the SA index and shows that green insurance significantly reduces the SA index at the 1% level, indicating that green insurance helps to reduce the level of financing constraints; column (2) reports the effect of green insurance on the level of financing scale and shows that green insurance significantly increases the size of long-term loans; column (3) shows the effect of green insurance on financing costs but shows no statistically significant effect on this aspect. To sum up, green insurance helps firms obtain more resources, reflected mainly by the scale of financing. The result is similar to the study of Ning and Yuan [51], who also find that green insurance can reduce corporate financing constraints.

5.2. Risk Taking

The basic function of green insurance is to disperse the insured firms’ environmental risks, thus increasing the their risk taking level. We test the risk taking mechanism by following previous work and using two types of measures to capture the firm’s level of risk taking. We use the firm’s R&D expenditure, which is often used to measure the firm’s risk strategy [55,56]. R&D investments are risky due to the low probability of research success and uncertain benefits. We measure the firm’s overall risk based on its asset–liability ratio and overall returns volatility [57].

Table 8. Impact of green insurance on risk taking.

Variables	R&D (1)	LEV (2)	Return Volatility (3)
Insurance	0.339 **	0.091 ***	0.035 ***
	(0.144)	(0.004)	(0.003)
Controls	Y	Y	Y
Year fixed	Y	Y	Y
Industry fixed	Y	Y	Y
Observations	12,496	12,496	12,496
R-squared	0.450	0.396	0.349

Note: the standard errors clustered at the firm level are reported in parentheses. *** and ** significant at 1% and 5%, respectively.

presents the results of the estimations. Column (1) presents the results for the effect of green insurance on R&D investment and shows that green insurance significantly promotes R&D investment at the 1% level, which indicates that green insurance helps to increase the level of risk taking; column (2) reports the effect of green insurance on financial leverage and shows that green insurance significantly increases the level of financial leverage; column (3) shows that green insurance significantly increases the volatility of firm earnings. These findings suggest that green insurance not only increases the firm’s risk strategy but also promotes the overall risk taking level. Wang and Nie [42] also find that the mechanisms driving this positive relation between green insurance and green innovation include reducing innovation risk.

5.3. Long-Term Orientation

Another potential mechanism for green insurance to promote green innovation might be through long-term cooperation between the insurance institution and the firm, which promotes the firm’s long-term orientation. We measure the firm’s long-term orientation in two ways. First, we use the firm’s long-term investment [58], including tangible and intangible assets [59], which we measure as the sum of capital expenditure (CAPEX). Second, long-term commitment is sometimes measured by ISO14001 certification and corporate social responsibility [60]. ISO14001 certification provides an objective valuation of the firm’s pro-environment commitment, and ISO14001 audit shows that the company values long-term benefits. Commitment to corporate social responsibility reflects the importance that the firm attaches to long-term sustainable management [60].

Table 9. Impact of green insurance on long-term orientation.

Variables	CAPEX (1)	ISO14001 (2)	CSR (3)
Insurance	0.723 ***	0.033 ***	4.720 ***
	(0.024)	(0.011)	(0.025)
Controls	Y	Y	Y
Year fixed	Y	Y	Y
Industry fixed	Y	Y	Y
Observations	12,390	11,471	12,390
R-squared	0.376	0.062	0.772

Note: the standard errors clustered at the firm level are reported in parentheses. *** significant at 1%.

reports the estimation results. Column (1) presents the effect of green insurance on long-term corporate investment and shows that green insurance contributes significantly to long-term investment at the 1% level, suggesting that green insurance increases firms’ long-term orientation. Column (2) reports the results for the effect of green insurance on ISO14001 certification and shows that green insurance significantly improves corporate certification. Column (3) reports the impact of green insurance on social responsibility and shows that green insurance significantly increases corporate social responsibility. These results suggest that green insurance promotes a stronger commitment to long-term and sustainable development.

6. Further Analysis

6.1. The Impact of Green Insurance on the Green Innovation Quality

We find that green insurance can increase the number of green patent applications; however, whether green insurance contributes to higher-quality green innovation is unclear. Therefore, we investigate the effect of green insurance on green innovation quality. Patent citations are commonly used to measure innovations’ quality [61,62]. We measure innovation quality using total citations and weighted patent citations after scaling based on patent category citations [62,63].

Table 10. The impact of green insurance on the green innovation quality.

Variables	Ln(Citations) (1)	Ln(WCitations) (2)
Insurance	0.129 ***	0.123 ***
	(0.025)	(0.024)
Controls	Y	Y
Year fixed	Y	Y
Industry fixed	Y	Y
Observations	12,496	12,496
R-squared	0.234	0.226

Note: the standard errors clustered at the firm level are reported in parentheses. *** significant at 1%.

presents the results and column (1) reports the effect of green insurance on the overall number of total citations, and column (2) presents the results for the effect of green insurance on weighted citations. The results show that green insurance contributes significantly to an increase in green patent citations, which suggests that green insurance contributes also to higher green innovation quality. This is similar to the results of our estimation based on green invention patents, often used by scholars to characterize innovation quality.

6.2. The Impact of Green Insurance on Environmental Performance

Our results show that green insurance significantly improves both the quantity and quality of firms’ green innovation, which is an exciting result since green innovation is a fundamental measure to solve environmental and climate issues. Therefore, we expect that green insurance will further improve corporate environmental performance through green innovation.

Due to the lack of data on firm-level environmental performance, we use firm environmental disclosure to measure environmental performance following Xiao and Shen [64]. The variables are based on environmental disclosure of waste gas, wastewater, and solid waste.

Table 11. The impact of green insurance on environmental performance.

Variables	Waste Gas (1)	Waste Water (2)	Waste Solid (3)	Waste Gas (4)	Waste Water (5)	Waste Solid (6)
Insurance×LnGP	0.054 ***	0.049 ***	0.033 ***
	(0.013)	(0.012)	(0.012)
Insurance×LnCitation				0.073 ***	0.069 ***	0.035 ***
				(0.014)	(0.014)	(0.013)
Controls	Y	Y	Y
Year fixed	Y	Y	Y	Y	Y	Y
Industry fixed	Y	Y	Y	Y	Y	Y
Observations	11,471	11,471	11,471	11,471	11,471	11,471
R-squared	0.220	0.204	0.207	0.220	0.205	0.206

Note: the standard errors clustered at the firm level are reported in parentheses. *** significant at 1%.

reports the results. Columns (1) to (6) show that the estimated coefficients of Insurance×LnGP and Insurance×LnCitation are significantly positive, indicating that green insurance promotes better environmental performance through green innovation. Some studies also find that green insurance promotes environmental performance; for example, Ning and Yuan [51] find that green insurance has significantly promoted all types of industrial firms’ emissions reduction.

7. Conclusions and Policy Implications

This paper investigated the impact of green insurance on corporate green innovation in the context of China’s 2013 Pilot Policy of Compulsory Liability Insurance for Environmental Pollution. Based on panel data of listed companies in the period 2008 to 2020 covering 12,496 firm–year observations in high-polluting industries, we find that, on average, insured firms applied for a higher number of green patents, indicating that green insurance significantly promotes green innovation. The mechanism analysis shows that green insurance can promote the insured firms’ access to resources, enhance risk taking, and promote a long-term orientation. In addition, green insurance can improve the insured firms’ environmental performance by promoting green innovation.

Based on our findings, we further propose some policy implications, which can provide some enlightenment for government departments to adopt or plan to adopt green insurance to deal with environmental challenges. First, green insurance aimed at reducing environmental risks could incentivize green innovation, which indicates that, in the context of global green transformation, green insurance can be used to encourage high-carbon firms to innovate and achieve low-carbon transformation and green development. We suggest that government departments should actively improve the policy design and product design of green insurance and maximize the professional functions of insurance institutions to stimulate green innovation.

Secondly, the mechanism analysis finds that green insurance introduces insurance institutions to rich expertise and deep engagement in the firm’s environmental management, which promotes the firm’s resources acquisition, risk taking, and long-term positioning. Evidently, green insurance is a supplement to green financial instruments that directly provide financial support. We propose to strengthen the combination of green financial instruments, such as green insurance and green loans, so as to provide more professional and financial support for those polluting firms and promote them to achieve sustainable development.

Thirdly, further analysis shows that green insurance can improve environmental performance through green innovation, which indicates that green insurance can help to achieve a win–win situation of environmental and economic benefits. We suggest that green insurance, as a policy tool, should be widely used for enterprises with different pollution levels and in different regions to encourage green innovation of high-carbon enterprises and thus promote the development of a green economy.

Future work on green insurance could include the following aspects. The present study focuses mainly on firms’ green innovation and environmental performance; future work could investigate in more detail the impact of green insurance on economic performance and competitiveness to deepen our understanding of how green insurance contributes to environmental and economic sustainable development. Future work could also consider the mutual influence and synergic effects of different green financial instruments, such as green credit, green bonds, etc., which would enable a more comprehensive understanding of the effect and impact of green finance.