1. Introduction
Climate change is one of the most critical issues in global society today, and its associated problems have influenced corporate businesses and assets significantly [
1]. Scientific evidence shows that greenhouse gas (GHG) emissions play a significant role in global warming and climate change [
2]. Global efforts such as the “United Nations Framework Convention on Climate Change” and the “Paris Agreement” have been made to address climate concerns. In addition, most developed countries have announced “Net Zero” plans to reduce net GHG emissions to zero. Corporations are among the largest emitters of GHGs and thus have a responsibility to address this issue. Accordingly, the questions concerning the way corporations should be governed and the way governance structures determine Environmental, Social, and Governance (ESG) performance have been raised [
3,
4]. Scholars have investigated the role that corporate governance plays in GHG emissions, as effective governance practices can help companies reduce their carbon footprint and mitigate climate risks [
2,
4]. Among firm-related governance factors, board composition, institutional ownership, laws and regulations, accounting and auditing, and stakeholder pressure have been studied [
5,
6,
7]. Further, researchers have attempted to understand the relation between carbon performance and financial performance, but have obtained heterogeneous results. For example, Refs. [
8,
9] demonstrated a positive relation between reduced GHG emissions and financial performance, but Refs. [
10,
11] found a negative association between the two. In this study, we attempt to examine the role that board independence plays in reducing GHG emissions as well as its effect on the relation between GHG emissions and financial performance.
Carbon performance is described generally as measures or processes that reduce GHG emissions [
12]. A large investment is required to improve carbon performance with ambiguous consequences that affect various stakeholders in different ways [
1]. Corporate managers tend to make short-term decisions in response to market pressures to increase investment value and stock prices [
13]. According to the previous literature [
3,
6], internal directors are more inclined to pursue short-term economic goals, while outside directors who represent various stakeholders tend to adopt a long-term perspective. Therefore, a diversified and unbiased board is required to alleviate conflicts of interest among various stakeholders [
7]. However, few empirical studies have explored the relation between board independence and GHG emissions. Moreover, existing studies have investigated the effects of board characteristics on carbon disclosure rather than GHG emissions. In general, carbon disclosure is a binary variable that takes a value of 1 for firms that disclose their carbon footprint information. Many empirical studies have been conducted to understand the association between GHG emissions and financial performance [
9,
11,
14] but have obtained inconclusive results. Thus, some scholars have attempted to employ various moderators, including governance-related variables, to understand when reduced GHG emissions improve financial performance. However, thus far, few studies have investigated board independence to understand GHG emissions and financial performance. In this study, we focus on board independence as our main corporate governance measure because corporate governance experts consider board independence an important factor that can bridge the gap between managers’ and shareholders’ interests [
13]. The purpose of this study is to fill that gap by investigating the relation between board independence and GHG emissions and board independence’s effect on the relation between carbon and financial performance.
This study employs board independence data together with GHG emissions and financial data of listed firms that operate in South Korea. Board independence is defined as the proportion of outside directors on the board. GHG emissions data are obtained from the National GHG Emissions Information System. As the amount of GHG emissions are quantitative, not qualitative data, a more precise analysis can be conducted [
15]. We apply Tobin’s q as a market-based measure of financial performance, and use return on equity (ROE) as an accounting-based metric. The sample contains 1404 firm-year observations for 156 listed companies from 2011 to 2019. The effects of board independence on GHG emissions and on the relation between GHG emissions and financial performance are investigated using multivariate regression models.
The results show that board independence is related positively with a reduction in GHG emissions. This positive relation is consistent with the finding of [
1], which also found a positive relation between independent directors and carbon performance. In addition, our evidence shows that companies that produce more GHG emissions have better financial performance, but board independence weakens that positive relation. These findings suggest that independent directors are more likely to consider the benefits and long-term financial consequences of reducing GHG emissions. When we divide the sample into two groups, one with a higher percentage of independent directors versus one with a lower percentage, firms with a higher proportion of outside directors show a negative relation between GHG emissions and financial performance. In addition, the patterns are more significant when a market-based measure of financial results is employed rather than an accounting-based measure. These results imply that independent directors tend to pursue sustainable development by balancing financial versus environmental performance. Further, they appreciate the financial benefits resulting from reducing GHG emissions from a long-term perspective.
This empirical research contributes to the existing body of literature on corporate governance and carbon performance in numerous ways. First, we provide empirical evidence of the way governance-related factors influence GHG emissions as well as the relation between GHG emissions and financial performance. Other studies have used quality indices to measure carbon performance, such as participation in CDP, or aggregated quality indices such as ASSETS4 or MSCI ESG STATS. By employing the actual amount of GHG emissions at the firm level, we present quantitative information on the extent to which board independence contributes to the decrease in GHG emissions and weakens the relation between GHG emissions and financial performance. Second, our evidence suggests a moderating variable to explain the heterogeneous relation between environmental and financial performance. Many scholars have attempted to determine this relation by considering diverse variables. For example, Ref. [
16] employed firm size as a moderating variable; Refs. [
17,
18], pollution levels; and Refs. [
19,
20], environmental strategies, but few studies have investigated board characteristics in the GHG emissions context. Finally, we extend the scope of empirical research by using a South Korean sample. South Korea’s carbon emissions have continued to increase and ranked seventh worldwide as of 2019. The trend in Korea is in contrast to that in the U.S. and Europe, which have shown a decreasing trend every year. Korea’s per capita GHG emissions are 11 tons: twice that of Europe. Some researchers have examined the effect of Corporate Social Responsibility (CSR) performance or Environmental, Social, and Governance (ESG) rating on financial performance in Korea [
21,
22], but few studies have been conducted to understand GHG emissions at the firm level in South Korea. In summary, our findings suggest the importance of corporate governance structure in addressing challenges of climate change and the policy implications that certain numbers of outside directors should be maintained to balance a firm’s financial versus environmental performance.
The remainder of this study is organized as follows: In
Section 2, the theoretical backgrounds of the relation between corporate governance and carbon performance are examined and our research hypotheses are formulated. In
Section 3, we discuss the study’s methodology and empirical models.
Section 4 describes the principal findings, and finally, the research findings are discussed in
Section 5.
3. Materials and Methods
The sample data were obtained from South Korean firms that are obligated to report their GHG emissions data. According to the “Carbon Neutral Green Growth Framework Act” that was established in 2010, carbon-intensive firms that produce over 50,000 tons CO
2-eq or plants that produce over 15,000 tons CO
2-eq annually must report their GHG emissions. These firms are required to reveal their Scope 1 (direct emissions) and Scope 2 (indirect emissions) according to the Greenhouse Gas Protocol, which provides guidance on the ways to quantify and report the emissions figures [
76]. The number of firm-year observations included in the full sample is 7312 from 2011 to 2019, which includes 951 entities. These entities’ total emissions accounted for 89% of nationwide emissions in 2019. After unlisted firms and public institutions were removed, the number of firms in the sample decreased to 391, which accounts for 35% of nationwide emissions. Further, listed firms with 9-year emissions data were obtained for the final sample to employ a balanced panel to accommodate each firm’s year-by-year variations sufficiently. Finally, a sample of 1404 firm-year observations was obtained, which included 156 listed firms. In 2019, the total emission amount from these 156 firms was 222,562,867 tons CO
2-eq, which accounted for 31% of nationwide emissions. Next, the emissions data were matched with financial data from DataGuide.
Appendix A details the sample selection procedure that yielded 1404 firm-year data. The panel unit-root test results show that the GHG emissions variables are stationary.
The industry distribution of sample firms is provided in
Appendix B. Samples from firms that manufacture chemicals and chemical products constitute 17% of the total sample, followed by those that manufacture basic metals (14%); those that manufacture pulp, paper, and paper products (12%); and those that manufacture electronic components, computers, and visual, counting, and communication equipment (9%).
To understand the association between board independence and GHG emissions, a GHG intensity variable (
CARBON) is regressed on board independence (
B_IND), as in Equation (1):
where
i denotes firms and
t, periods. The dependent variable,
CARBON, is measured by dividing the total GHG emissions by total assets (
GHG/TA) and by sales (
GHG/SALES), respectively. The amount of GHG emissions is converted to kilograms, while a company’s assets and sales are measured in thousands of Korean won (KRW). This indicates how much GHG is produced per total asset measured by one thousand KRW and per sales of one thousand KRW, respectively. The independent variable, board independence (
B_IND), is computed as the number of outside directors divided by the total number of directors, and represents the independence in decision-making [
30]. In addition, the equation includes control variables that are known to influence carbon performance [
1,
6]. Profitability (
ROA) is determined by dividing net income by total assets. To determine leverage (
LEV), total liabilities are divided by total assets. Capital expenditure (
CAPEX) is defined as the ratio of capital expenditure to sales. The ratio of market-to-book value of equity (
MB) is also included as a control variable, as companies with a larger market-to-book ratio have more investment opportunities and are more likely to demonstrate superior environmental performance [
6]. Further, the regression model includes year and industry dummy variables. The fixed effect model eliminates unobserved industry features that are associated highly with the explanatory variables and adjusts for time-invariant industry characteristics.
In our study, the potential endogeneity problem is addressed using one-period lagged values of the relevant right-hand-side variables in the model. Because the future cannot cause the past, the right-hand-side-variables can be considered predetermined, thus alleviating the potential endogeneity problem. There may also be common third factors that are not controlled in a relation among the variables of interest. In our case, the common factors that can affect the relation between the dependent variable and the independent variables may be the unobservable firm and CEO characteristics, firm risk, executive compensation, and asset tangibility, among others. A range of techniques, such as instrumental variables, panel data models, difference-in-differences (DIDs), and regression discontinuity (RD) designs, can be considered to address the problem that arises from the presence of common factors and endogeneity. We employed a panel data model that included industry-fixed effects and year-fixed effects to do so. Fixed-effect models control for time-invariant unobservable variables by estimating the relation between variables within groups over time. This technique is particularly useful for panel data where observations are collected from the same individuals or units over time. By controlling time-invariant unobservable variables, fixed-effect models can help identify the causal effect of variables of interest. Note that more powerful tools such as DIDs and RD could not be implemented in the Korean setting because no structural changes in board independence occurred during the period that corporate carbon emissions were collected.
Hypothesis 2 examines the effect of board independence on the relation between GHG emissions and financial performance. To test the hypothesis, financial performance (
FP) is regressed on carbon intensity (
CARBON) and an interaction term of
CARBON and
B_Dummy is included, as in Equation (2):
where
B_Dummy takes the value of 1 for the firms in which board independence is 0.4 or more and takes the value of 0 otherwise. Note that the statistical analysis shows that the median value for board independence (
B_IND) is 0.4. Two different financial measures are employed for the financial performance (
FP) measure. Return on equity (
ROE) is an accounting-based measure, defined as the net income over the shareholder’s equity, and
Tobin’s q is a market-based measure, defined as the sum of the book value of equity and the book value of total liabilities divided by the book value of total assets [
36,
51]. Total GHG emissions are divided by total asset (
GHG/TA) and sales (
GHG/SALES) for carbon performance (
CARBON), as defined in Equation (1).
To alleviate endogeneity problems, the financial performance variable (
FP) is advanced by one year (
FPt+1). Endogeneity may occur with the presence of simultaneous causality, which the slack resource theory explains. According to [
77], improvement in environmental performance can be associated with an increase in financial performance, which implies that the relation’s direction is unclear. In previous research [
11,
53,
76], the independent variable of carbon performance was lagged by one year. In a similar manner, the dependent variable for this empirical research, financial performance (
FP), is advanced by one year to address endogeneity issues according to [
74]. Consequently, the reliability and robustness of the empirical analysis pertaining to the direction of the association can be enhanced. Further, the regression model includes year and industry dummy variables, as in Equation (1).
Leverage (
LEV) and capital expenditure (
CAPEX) are included as control variables, as in (1), as well as other variables known to affect financial performance, such as firm size (
SIZE), advertisement intensity (
ADV), R&D intensity (
R&D), asset growth (
ASSETS GRWTH), and sales growth (
SALES GRWTH) [
74].
ADV is calculated by dividing advertising expense by sales;
R&D, by dividing research and development costs by sales;
ASSETS GRWTH, by dividing the current year’s assets growth by the previous year’s assets; and
SALES GRWTH, by dividing the current year’s sales growth by the previous year’s sales [
8,
45,
49].
ASSETS GRWTH is used in the regression model with the independent variable
GHG/TA, while
SALES GRWTH is used in one with
GHG/SALES, consistent with [
74].
5. Discussion
In this study, we evaluate the effect of corporate governance, specifically the independence of the board, on GHG emissions and the relation between GHG emissions and financial performance. In contrast to existing research [
36,
37], we quantify environmental performance by focusing on the amount of GHG emissions at the firm level. This is based on a multivariate analysis of firm-level data from 156 South Korean listed enterprises over the nine-year period from 2011 to 2019. We find that board independence is related positively with the decrease in GHG emissions and has a negative effect on the relation between GHG emissions and financial performance. Our findings are consistent with the agency theory and stakeholder theory arguments [
3,
78] that outside directors are more likely to be accountable to a wider variety of stakeholders. According to the agency theory, an independent governing body can supervise agents’ actions effectively [
26]. According to the stakeholder theory, board independence influences environmental performance favorably because shareholders influence external directors less [
1,
28]. Our results imply that an effective internal governance framework may help businesses achieve both financial and nonfinancial performance.
Our findings provide evidence that firms that produce more GHG emissions have better financial performance, which is consistent with the neoclassical view [
64]. Reducing GHG emissions is considered an additional financial burden that reduces a firm’s financial performance [
65,
67]. However, firms with more independent directors have a different relation. As Ref. [
6] showed, independent boards are more likely to care about the company’s environmental performance and to appreciate the realization of long-term investments in environmental projects. Therefore, independent boards tend to support costly environmentally friendly decisions and appreciate long-term financial performance. Our empirical findings indicate that the effects of board independence on the relation between GHG emissions and financial performance is more pronounced when financial performance is assessed using a market-based metric,
Tobin’s q. It has been reported that costs and benefits associated with environmental initiatives are not reflected instantly in financial performance [
61]. However,
Tobin’s q may be able to evaluate their predicted long-term effects [
50]. Thus, it can be considered one of the main financial performance measures in a GHG-emissions-related analysis. Reducing GHG emissions requires considerable resources without providing immediate financial benefits [
66], indicating a negative correlation between carbon and short-term financial success. Thus, a good corporate governance structure is expected to be more responsible in controlling GHG emissions.
The quantitative analysis in our research contributes to the current body of knowledge about the relation between GHG emissions and financial performance. Prior research on environmental performance has concentrated on the relation between corporate governance and environmental disclosures [
33,
79,
80] or CDP participation [
1,
30]. To the best of our knowledge, this is one of the first attempts to investigate the relation between corporate governance and actual amounts of GHG emissions at the firm level. By employing GHG emissions data to measure carbon performance, our study demonstrates the relation between corporate governance and environmental performance quantitatively. In addition, we extend the moderating analysis literature to evaluate the heterogeneous relation between environmental and financial performance by introducing a governance-related variable. Further, we extend the scope of empirical investigations by using a sample from South Korea, which is ranked as the country with the seventh largest carbon emissions as of 2019. Overall, our findings indicate the significance of the corporate governance structure in responses to climate change issues. As climate change affects everyone—corporate managers, shareholders, lawmakers, and consumers—it is important to have more independent directors to monitor firms’ decisions and meet sustainable goals.
Our results are restricted to public corporations in South Korea; therefore, this research has certain limitations. The analysis of unlisted firms or those operating in other countries may yield different results. In addition, we employ industry-fixed effect panel regression models rather than firm-fixed effect panels because of the limited number of samples as well as the persistence of board independence. The research on the underlying mechanism of corporate governance’s influence on environmental performance and its financial consequences may be more complex. Therefore, other research methods, such as case studies or survey methods, can provide a more in-depth understanding. We expect that future research on effective corporate governance mechanisms will be conducted to reduce GHG emissions and the potential damage attributable to climate change.