Is capital structure associated with corporate social responsibility?

Abstract

Based on a total of 1,590 listed non-financial firms on the Taiwan Stock Exchange and the Taipei Exchanges covering the period of 2007 ~ 2020, this study examines whether a firm's capital structure is affected by its corporate social responsibility (CSR) performance. While existing research has explored the impact of a firm’s CSR performance on various financial and non-financial consequences, this study argues that firm engaging in CSR is putting greater emphasis on the financial and bankruptcy risks arising from the use of debt financing and to maintain firm’s sustainability, firm with better CSR performance tends to reduce the use of debt. Through descriptive statistics, correlation analysis and multiple regression estimation, principal outcome shows that firm with better CSR performance tends to use less debt financing and inter-temporally reduce the use of debt.

Introduction

In addition to continuing to pursue profits, more and more firms are also actively undertaking corporate social responsibility (CSR) to enhance the corporate sustainability. At the same time, the endless international versus domestic environmental pollution and food safety scandals, such as the long-term counterfeiting of diesel engine by the Germany Volkswagen Group, the waste water pollution incident by the ASE Inc. and oil incident by the Ting-Hsin International Group etc., all showed that the violation of laws, regulation and expectation of the consumers and the society may result in serious loss and decline in stock price, and ultimately, all of above corporate wrongdoings have negative influence on the sustainability. In order to comply with the international trend, most of firms either passively or actively participate in CSR initiatives, hoping to maintain and establish the corporate reputation and image to improve the value and sustainability of the firm.

Nowadays, the academics have engaged in research on examining the relationship between firm’s CSR performance and economic consequences, e.g., the impacts of CSR on the quality of accounting reporting (Choi et al., 2013; Kim et al., 2012), marketing (Hildebrand et al., 2011; Luo & Bhattacharya, 2006) and business strategy (McWilliams et al., 2006). The various consequences of CSR engagement has also been extensively discussed in the finance research, such as the relationship between social responsibility and firm performance (Jiao, 2010; Kim & Statman, 2012; McWilliams & Siegel, 2000; Shen & Chang, 2009; Waddock & Graves, 1997; Wu & Shen, 2013), firm’s risk (Godfrey et al., 2009; Lee & Faff, 2009), and the insurance effects of CSR was also proposed and proved (Peloza, 2006; Minor & Morgan, 2011). Some studies explored the benefits and costs of CSR on various capital market benefits and costs (EI Ghoul et al., 2011; Dhaliwal et al., 2011; Goss & Roberts, 2011; Cheng et al., 2014; Chang et al., 2014).

While existing studies have discussed how CSR affects various aspects of economic consequence of firms, there are still relatively few studies discussing the impacts of CSR firm’s capital structure decision, namely, choice of the equity financing versus debt financing. To the author's knowledge, Girard-Potin, Jimenez Garces and Louvet (2011) studied the relationship between CSR performance and capital structure by European firms, and the empirical result showed that firms with poor CSR performance tends to have higher degree of debt. Based on European listed firms, Matthijs Jan van der Leest (2018) found that firms higher corporate social performance significantly experience lower debt ratios. In fact, while firm with better CSR performance tends to have better financial versus stock market performance and lower risks, making firms have lower costs of equity and cost of debt (Goss & Robert, 2011; EI Ghoul, 2011), firm with better CSR performance tends to reduce risk-taking (Harjoto & Laksmana, 2018), so they use less debt financing that may result in greater bankruptcy and litigation risk. At present, there is a relative lack of research on the relationship between CSR performance and capital structure of public-traded firms in Taiwan’s financial market, and the government authorities in Taiwan's financial market have also followed the global trend of paying attention to the sustainability, and have successively issued several policy norms and guidelines on CSR for public-traded firms. Based on the data of non-financial industry listed firms on the Taiwan Stock Exchange (TWSE) and Taipei Exchange Market (TPEx) in recent decades, this study examines the effects of CSR performance on corporate capital structure to fill the research gap.

Through descriptive statistics, correlation analysis and multiple regression estimation, the empirical result generally shows that firm’s CSR performance does affect corporate financing decision. Firm with better CSR performance tends to have lower debt use, and those firms with better CSR performance also tend to reduce debt use over time. The empirical result of the research helps government authorities to comprehend whether the policies and regulations strengthening and encouraging the public-traded firms to fulfill CSR should be continuously implemented, and the investors can also understand whether to allocate more money to the firms fulfilling CSR by knowing how CSR affects the financial risk of their investment targets. The next section is literature review and hypothesis development. "Variable, econometric model, firm samples and data source" section introduces variables, econometric model, samples and data sources. "Empirical result" section presents empirical result and discussion, and the last section is the conclusion.

Literature review and hypothesis development

The development of CSR and related norms

The concept of CSR is generally believed to have been first proposed by Sheldon (1924). Sheldon (1924) indicated that CSR refers to the responsibility of entrepreneurs in meeting the needs of all types of people in the industry. Bowen (1953) proposed "social responsibilities of the businessman" and defined CSR as an obligation that entrepreneurs should consider social values and corporate goals when making policies and taking corporate actions. Elkington (1997) put forward the theory named "Triple Bottom line", indicating that firm should consider the result of economic, social and environmental aspects. In addition to ensuring the financial strength, a firm must also pay attention to environmental protection and respect from the society.

Several multinational institutions have also put forward famous frameworks and guiding principles of CSR for global business world, including the United Nations’ Global Compact, OECD’s Guidelines for Multinational Enterprises, the Global Reporting Initiatives (GRI), ILO Conventions, Global Sullivan Principles, Social Responsibility SA8000, International Standards Organizations ISO 14000 and AA1000, World Business Council for Sustainability and Development (WBCSD), and above guidelines are commonly known as the Global Eight (McIntosh et al., 2003).

In response to international trends, Taiwan's securities authorities, Financial Supervisory Commission of the Executive Yuan (FSC) and TWSE have strengthened the CSR performance of publicly traded firms, promoted the sound development of social responsible investment in the capital market, and enhanced the competitiveness of publicly traded firms and soundness of financial markets development. In 2010, they announced the "Corporate Social Responsibility (Sustainable Development) Best Practice Principles for TWSE/TPEx Listed Companies" and the "Ethical Corporate Management Best Practice Principles for TWSE/GTSM Listed Companies", it is clearly stated that public-traded firms should pay attention to the improvement and strengthening of social responsibility information disclosure and reduce information asymmetry. The FSC also emphasizes on CSR disclosure standards, such as Article 10 of the "Regulations Governing Information to be Published in Annual Reports of Public Companies" and Article 31 of the "Regulations Governing Information to be Published in Public Offering and Issuance Prospectuses", the firm should disclose its commitment (such as the firm’s commitment to environmental protection, social participation, social contribution, social welfare, consumer rights, human rights, safety and health, etc.), the systems and measures adopted and the information disclosure of CSR, so that the investing public may understand the firm's fulfillment of CSR.

The FSC announced in 2014 that starting from 2015, the listed firms on the TWSE and TPEx (food industry, financial industry, chemical industry and firm with paid in capital of more than 10 billion NTD) are required to prepare CSR reports, representing firm's efforts and implementation of CSR. At the same time, the TWSE also encouraged firms to devote resource to CSR and launched a number of CSR indexes, including (1) the "Taiwan Employment 99 Index", consists of firms sharing greater proportion its revenue to employee, which is in line with the government's policy of reducing the unemployment rate and encouraging firms to hire more employees. (2) the "Taiwan High Salary 100 Index", consists of firms hiring greater number of employees, which in line with the spirits of making profits, enterprises can also take care of and give back to improves employees' salaries and benefits. (3) The "Taiwan Corporate Governance 100 Index", consists of firms following better and advanced corporate governance guidelines, showing that firm with good corporate governance are suitable for investors' long-term investment.¹

The benefit and cost of CSR

The core value of CSR refers to firm’s not only create profits to be responsible for the interests of shareholders, but also take into account the rights and interests of stakeholders such as customers, employees, suppliers, society and the environment. While more and more firms have engaged in CSR in practice, in order to judge whether it is worth investing in CSR, and comprehend whether put resources in CSR bring enough benefits to cover costs for firms in various aspects, the academics has examined how CSR engagement affects various aspects of economic consequences.

Generally, there are two competing views regarding the impacts of CSR. The social impact hypothesis, proposed by Cornell and Shapiro (1987) and Preston and O' Bannon (1997), summarizes positive association between CSR and economic consequences. Firm's good performance in CSR meeting stakeholders from all levels of the society, thus improving corporate reputation and trust, which will have a positive help to the firm's financial performance. Porter and Kramer (2002) believed that firm incorporating CSR into corporate strategies, and a definitely positive CSR strategy not only enhances the positive image, but also makes firm gain competitive advantages and increased profitability. Elliott et al. (2014) found that investors tend to evaluate firms CSR engagement and are willing to invest in firms with better CSR performance thus result in firm’s better stock market performance. Firms paying attention to the performance on CSR tend to improve reputation, and positively affect compensation and market evaluation (Argenti & Druckenmiller, 2004; Dowling, 2006; Brammer & Pavelin, 2006). Not only that, firms that are good at CSR can reduce corporate risks, reduce corporate cost of capital, and increase investors base, and increase the willingness of investors’ stock-holding (El Ghoul et al., 2011; Fauzi et al., 2007; Graves & Waddock, 1994; Sharfman & Fernando, 2008). Several channels could explain this positive effect such as providing better working place improves employee productivity (Turban & Greening, 1997); donation to the public benefits increases social reputation, trust (Bowman & Haire, 1975; Alexander & Bucholtz, 1978) and brand image and product competitiveness (Porter & van der Linde, 1995; Fombrun et al., 2000). Studies by Moskowitz (1972), Parket and Eibert (1975) and Soloman and Hansen (1985) also claimed that CSR leads to more benefits than the cost incurred, suggesting that there is positive relation between CSR and financial performance.

Waddock and Grave (1997), Shen and Chang (2009) and Wu and Shen (2013) pointed out that CSR performance has a significant positive impact on performance and value. Increase in the management of CSR initiatives leads to the improvement of firm’s operating results and thus obtain better stock market performance (Griffin & Mahon, 1997; Margolis & Walsh, 2003; Orlitzky et al., 2003). Firm’s CSR performance also affects firm-specific risk indicators such as earnings volatility, leverage usage, and the stock return market risk (Orlitzky & Benjamin, 2001).CSR helps to improve firm's risk management and acts as performance insurance (Chen et al., 2015; Godfrey et al., 2009; Koh et al., 2014; Minor & Morgan, 2011; Peloza, 2006). CSR also has an impact on the firm’s stock price crash risk (Kim et al., 2014), and also affects the company’s financing costs in the financial market, including the cost of equity funds and debt funds (Dhaliwal et al., 2011; El Ghoul et al., 2011; Goss & Roberts, 2011). CSR has benefit on earnings reporting quality (Kim et al., 2012), financial market trust (Lins et al., 2017), analyst recommendations (Ioannou & Serafeim, 2015), credit rating (Chang & Shen, 2014), analyst forecast error (Dhaliwal et al., 2012), and access to financing (Cheng et al., 2014) and other financial consequence indicators.

However, Friedman (1970) believes that the market itself creates social welfare, and opposes firms engaging in CSR. Friedman (1970) suggested that the only responsibility of firm is to maximize firm’s profits. When a firm engages in CSR, it incurs costs. Aupperle et al. (1985) indicated that if a firm allocates resource to CSR such as environmental protection and charity. Firms put resource in social activities may lose competitiveness in the industry by increasing operational costs and reducing profitability (Preston & O'Bannon, 1997; Hess et al., 2002). Some studies have pointed out that although engaging in CSR helps firms build a good reputation, the reputation may not bring performance improvement (McWilliams & Siegel, 2000; McWilliams et al., 2006; Porter & Kramer, 2006; Hillenbrand & Money, 2007). From the perspective of agency theory, managers who want to gain social recognition in pursuit of personal interests or reputation are prone to over-invest in CSR activities, resulting in the abuse of corporate resources and result in decline of performance and value, in turn, damage shareholders' wealth (Beltratti, 2005; McWilliams et al., 2006; Jizi et al., 2014).²

The determinants of capital structure

Modigliani and Miller (1958, 1963) put forward several theories about capital structure, which claimed that under the complete capital market and without considering the existence of income tax, the firm's cost of capital would not be affected by the capital structure, and the value of firm is irrelevant to the capital structure, namely, equity financing versus debt financing. Modigliani and Miller (1963) pointed out that when a firm takes income tax into account, the interest expense of its debt can be viewed as an expense to offset the income tax burden, therefore, the higher the degree of debt use, the lower the cost of capital and the more the value of the firm. Modigliani and Miller (1963) and Miller (1977) proposed that the interest expenditure of debt financing has the effect of tax shield and affects the capital structure of firm. Jensen and Meckling (1976) proposed the agency theory, pointing out that debt financing cause a conflict between shareholders and creditors. When firm borrows more, greater potential financial and bankruptcy risk affect shareholders' rights and interests, but it can also reduce the conflict between managers and shareholders. By paying interests from borrowing, it can reduce the agency problem. Therefore, a firm will consider the agency costs versus the tax shield benefits of debt financing to achieve the optimal capital structure, and the trade-off theory also considers the trade-off between the tax shield of debt and the bankruptcy cost.

The pecking order theory of Myers and Majluf (1984) indicated that a firm tends to choose to borrow first, and then consider issuing shares to raise funds. There are also well-documented studies about capital structure, including the effects of personal tax (Miller, 1977), non-debt tax shield (DeAngelo & Masulis, 1980), agency cost (Jensen, 1986), stock return (Welch, 2004), analyst report (Chang et al., 2006), liquidity (Lipson & Mortal, 2009), information asymmetry (Autore & Kovacs, 2010; Bessler et al., 2011) and employee relations (Bae et al., 2011; Verwijmeren & Derwall, 2010).

Baskin (1989) took the top 500 enterprises in the Fortune magazine as the research sample and found that the profitability is negatively correlated with the debt ratio. Firm with better profitability tends to use internal retained earnings to meet the fund demand instead of borrowing externally, resulting in a low debt ratio. Hall, Hutchinson and Michaels (2000) took 3,500 UK small and medium-sized firms as samples in 1995, and found that profitability, growth, value and size positively affect long-term debt use, but profitability, value and size negatively affect short-term debt use. Panno (2003) used logit and probit regression models to find that the choice of financing for firm was internal financing in priority to external financing, while the debt ratio was positively correlated with scale and profitability, and negatively correlated with liquidity, financial leverage and other bankruptcy risk indicators. Based on the data of firm in developing economies such as india from 1991 to 2007, Kumar and bodla (2014) found that borrowing cost, scale, asset mortgage price and liquidity are important determinants of capital structure.

Pai (2007) studied the influencing factors of capital structure of 214 listed firms in Taiwan from 1986 to 2005, and found that there was a significant positive correlation between firm size, operational risk, growth rate and total debt ratio, furthermore, profitability, long-term investment and total debt ratio are significantly and negatively correlated. Lu (2012) studied the determinants of the capital structure of Taiwan's IC industry, and found that Taiwan's IC industry has a positive correlation between capital structure and firm size, mortgaged assets and operating risks, and a negative correlation between profitability and firm uniqueness. Chao et al. (2012) conducted empirical research on 400 listed firms in Taiwan from 2001 to 2010, and found that firm size has negative impact on debt ratio, while growth and sales of related parties have positive impact on debt ratio. The research result of Yang et al. (2015) showed that the size of the board of directors, the firm size, the number of board meetings and the shareholding ratio of directors and supervisors of Taiwan's listed construction industry are positively related to the capital structure, and the CEO duality and equity concentration are negatively related to the capital structure.

CSR and capital structure

Past literature has shown that when firms are willing to take CSR, they are less likely to violate the interests of stakeholders. Bènabou and Tirole (2010) indicated that through firm’s engaging in CSR, various stakeholders have more opportunity in participating in corporate decision-making, formally and informally, which can reduce the possibility of short-term speculative behaviors by the management or controlling shareholders. The research results of Kim et al. (2012) also found that firms that implement CSR are less likely to engage in earnings management, means that CSR encourages firms to improve financial information transparency. Gao et al. (2014) pointed out that the top management of firm with good CSR performance are less likely to engage in insider trading than firms with worse CSR performance. Cai et al. (2011) and Kong et al. (2022) also confirmed that firm with better CSR performance are more able to alleviate the phenomenon of pay inequity in terms of the pay gap between the management and employees. Eccles et al. (2014) stated that firms with higher CSR are more inclined to establish a decision-making process in which stakeholders can participate more and disclose firm’s non-financial-related information. It can be seen from the above literature that CSR indeed helps to alleviate the agency problem by taking more care of stakeholders’ interest instead of just shareholders or controlling shareholders’ interests.

More specifically, Girerd-Potin et al. (2011) studied the correlation between CSR performance and capital structure, and found that firms that are good at CSR tend to issue equity securities to raise funds due to the advantage of reducing securities costs, and CSR performance and debt ratio show a significant negative correlation. Hong and Kacperczyk (2009) found that tobacco, gambling and alcohol firms are prone to litigation costs due to the norms of social standards, which is reflected in the firm's cost of capital. In addition, the firm's stock price will be undervalued on the market, so that the firm may have higher financial leverage, that is, using more debt financing. Pijourlet (2015) found that when enterprises implement social responsibility, it is negatively correlated with capital structure. For financing decisions, they tend to issue equity securities, mainly because the implementation of CSR reduces information asymmetry and lower capital costs.

Based on the discussion of the above studies, it is generally showed that the firm's capital structure is highly related to firm’s CSR performance. For firms that are good at CSR tend to reduce corporate financial and bankruptcy risk, increase the issuance of equity securities to raise the required funds in financing decisions, and then reduce the firm's debt ratio. Firm with good CSR performance have greater incentives to reduce the bankruptcy risk of the firm in order to maintain the interests of all stakeholders, not just the interests of shareholders, so they tend to use less debt. The checking hypothesis is:

• Hypothesis: the better the performance of CSR, the lower the degree of debt use.

Variable, econometric model, firm samples and data source

Variable

Explained variable-capital structure

This study refers to the existing literature (Fama & French, 2002) and employs eight proxy variables for capital structure. First, leverage ratio (lev), defined as the total liabilities divided by total equity. Second, whether the leverage ratio is greater than the average in current period (levab), when the leverage ratio is higher than the average of all firms in in current period, levab is 1, and 0 otherwise. Third, the difference of firm's leverage ratio (levdif) between the two consecutive years, defined as the leverage ratio of year t minus the leverage ratio of year t-1. Fourth, whether the difference of leverage ratio is positive (levdifd), if the leverage ratio of year t is greater than the leverage ratio of year t-1, levdifd is 1, and 0 otherwise. Fifth, debt ratio (debt), defined as the total liabilities divided by total assets. Sixth, whether the debt ratio is greater than the average in current period (debtab), when the debt ratio is higher than the average of all firms in in current period, debtab is 1, and 0 otherwise. Seventh, the difference of firm's debt ratio (debtdif) between the two consecutive years, defined as the debt ratio of year t minus the debt ratio of year t-1. Eighth, whether the difference of debt ratio is positive (debtdifd), if the debt ratio of year t is greater than the debt ratio of year t-1, debtdifd is 1, and 0 otherwise. The larger the value of these eight variables, the more inclined the firm's capital structure is to use debt financing, and vice versa.

Main explanatory variable-CSR performance

Refers to the concept of social contribution value per share proposed in the guidelines on strengthening the social responsibility of listed firms and environmental information disclosure of listed firms of Shanghai Stock Exchange issued by Shanghai Stock Exchange in 2008, this study calculates the total amount created by the firm for main stakeholders, including after-tax earnings (to the shareholders), taxes (to the government), interests (to the creditors) and employee salaries and benefits (to the employees), and this total amount is defined as firm’s social contribution value (scv). In addition, to exclude scale effect, the social contribution value is divided by the total assets and then is defined as the social returns of assets (sroa), and the social contribution value is also divided by the number of shares outstanding and then is defined as the social contribution value per share (scvps). Greater value of social contribution value, social return rate of assets and social contribution value per share refers to better CSR performance.

Control variables

In addition to CSR performance variables, this study considers other determinants of corporate capital structure. Referring to existing studies such as Kumar and Bodla (2014), Hall et al. (2000), Baskin (1989), Frank and Goyal (2003), DeAngelo and Masulis (1980), Flannery and Rangan (2006), Crutchley et al. (1999), Panno (2003), Kumar and Bodla (2014) and Lee and Yeh (2004), capital structure determinants include firm's size (natural logarithm of total assets: asset), market to book value (market value of common equity to book value of common equity: mtb), fixed assets ratio (non-current assets to total assets: tang), depreciation expense ratio (depreciation to total assets: dep), R&D expense ratio (R&D expense to net sales: rd), institutional shareholdings (the number of shares hold by institutional investors divided by the number of shares outstanding: insthold), directors’ shareholding ratio (the number of shares hold by directors divided by the number of shares outstanding: dirhold) and the shareholding pledge ratio of directors (the ratio of all directors' pledged shares to total shares hold by all directors: pledge). Finally, while the research samples cover 31 industries and 14 years, and considering the differences of capital structure in various industries and years, the 30 industry dummy variables (INDUSTRY) and 13 yearly dummy variables (YEAR) are incorporated into the regression equation. The abbreviations and brief definitions of the above variables are summarized in Table 1.

Table 1 Abbreviations and definitions of variables

Econometric model

This research employs multiple regression estimation to examine the effects of CSR performance on capital structure. The regression equation is:

$$\mathbf{CAPITAL}_{i,t}=\beta_{0}+\beta_{1} \cdot\ \mathbf{CSR}_{i,t}+\beta_{2}\cdot asset_{i,t}+\beta_{3}\cdot mtb_{i,t}+\beta_{4}\cdot tang_{i,t}+\beta_{5}\cdot dep_{i,t}+\beta_{6}\cdot rd_{i,t}+\beta_{7}\cdot insthold_{i,t}+\beta_{8}\cdot dirhold_{i,t}+\beta_{9}\cdot pledge_{i,t}+\varepsilon_{i,t}+\gamma \mathbf{INDUSTRY}_{i}+\delta \mathbf{YEAR}_{t}+\varepsilon_{i,t}$$

(1)

where subscript i and t represent firm i in year t, respectively. CAPITAL is a vector of capital structure variable, including leverage ratio (lev), whether the leverage ratio is greater than mean (levab), yearly change in leverage ratio (levdif), whether yearly change in leverage ratio is positive (levdifd), debt ratio (debt), whether the debt ratio is greater than mean (debtab), yearly change in debt ratio (debtdif) and whether the yearly change in debt ratio is positive (debtdifd). CSR is a vector of CSR performance variable, including social contribution value (scv), social return on assets (sroa) and social contribution value per share (scvps). The control variables include firm size (asset), the ratio of market value to book value (mtb), the ratio of fixed assets to total assets (tang), the ratio of depreciation to total assets (dep), the ratio of R&D expense to net sales (rd), institutional investors’ shareholding ratio (insthold), directors’ shareholding ratio (dirhold), directors’ shareholdings pledge ratio (pledge), vector of industry dummies (INDUSTRY) and vector of yearly dummies (YEAR). The regression equation is pooled-OLS estimated.

Sample selection and data source

This study takes listed non-financial industry firms on the Taiwan Stock Exchange (TWSE) and the Taipei Exchange (TPEx) (excluding the firms of banking, insurance, billing, securities and financial holdings) as the research samples, with a total of 1,590 firms. The data is yearly ranged from 2007 to 2020, which forms 22,260 firm-year observations for each variable (yet some variables have missing data). The data source of firm’s financial characteristics and governance variables is Taiwan Economic Journal (TEJ) database.

Empirical result

Descriptive statistics and correlation analysis

Table 2 reports descriptive statistics of full samples (panel A), the samples of firm with above-median scvps (panel B) and the samples of firm with below-median scvps (panel C). The difference in mean of each variable is reported in rightmost column. Observing the mean differences of eight capital structure variables between two sub-samples, it can be found that they are all negative, and most of them reaches 1% statistical significance level, indicating that the sample of firms with better CSR, on average, have significantly lower leverage ratios, lower probability that the leverage ratio is higher than the averages, a smaller number of leverage ratio increases relative to the previous year and a lower probability that the leverage ratio increases relative to the previous year. There is also a lower probability that the debt ratio is higher than the average of the current year, the debt ratio increases less compared with the previous year, and there is a lower probability that the debt ratio increases compared with the previous year.

Table 2 Descriptive statistics

Numerically, the samples with better CSR performance tend to have lower leverage ratio (69.02% versus 79.5%), lower probability of having above-mean leverage ratio (0.3115 versus 0.3449) and lower probability of having inter-temporal increase in leverage ratio (0.4819 versus 0.5473). Similar result is also hold for the debt ratio. The samples with better CSR performance, on average, the leverage ratio and debt ratio are decreasing compared with the previous year (decreased by -2.5771% and -0.3426%, respectively), while the samples with worse CSR performance, on average, the leverage ratio and debt ratio are indeed increasing compared with the previous year (increased by 3.9681% and 0.4437%, respectively). The result shows that firms with better CSR performance tend to raise funds by issuing equity securities, which is in line with the hypothesis of the study. Firms with better CSR performance pay more attention to the overall interests of the stakeholders and sustainability of firm, not just the interests of stockholders and short-term profitability. Firms engaging in CSR also pay more attention to the financial and bankruptcy risks arising from the use of liabilities, so it tends to use equity funds rather than debt funds in financing.

The lower triangular elements of Table 3 report the Pearson correlation coefficients among variables. By observing the intersection of the 9^th ~ 11^th row and the 1^st ~ 8^th columns, it is found that the correlation coefficients of three CSR performance variables and eight capital structure variables are all negative and reach at least 5% significance level, indicating that firm with higher social contribution value, social return rate on assets and social contribution value per share tends to have lower leverage ratio, lower probability of having above-mean leverage ratio, less increase in leverage ratio compared with the previous year, lower probability of having increasing leverage ratios, lower debt ratio, lower probability of having above-mean debt ratio, less debt ratio increase compared with the previous year, lower probability of having increasing debt ratios. The upper triangular elements of Table 3 report the Spearman rank correlation coefficient among variables, and the result is similar. The principal outcome of correlation analysis is similar to the result of descriptive statistics, such that firm with better CSR performance tends to raise funds by issuing equity securities, which is consistent with the hypothesis that CSR engagement pushes firms to reduce the use of debt financing.

Table 3 Correlation coefficients matrix

Baseline regression result

Table 4 reports the regression estimation results of the effect of CSR performance (proxied by the social contribution value: scv) on the firm's capital structure. The capital structure variables in models (1) ~ (8) are leverage ratio (lev), whether the leverage ratio is greater than the average of all firms in the current year (levab), whether the leverage ratio is relative to the previous period (levdif), whether the leverage ratio is positive relative to the previous period (levdifd), debt ratio (debt), whether the debt ratio is greater than the average of all firms in the current year (debtab), whether the debt ratio relative to the change of the previous period (debtdif) and the debt ratio relative to the change of the previous period are positive (debtdifd), respectively. Observing the estimated coefficients on scv in models (1) ~ (8), it is found that they are all negative and reach a less-than 10% statistical significance level, indicating that the firm's social contribution value has negative impact on eight variables of the capital structure. Firm with higher social contribution value tends to have lower leverage ratio, lower probability of having above-average leverage ratio, less inter-temporal increase in leverage ratio, lower probability of having positive inter-temporal increase in leverage ratio. Similarly, firm with higher social contribution value tends to have lower debt ratio, lower probability of having above-average debt ratio, less inter-temporal increase in debt ratio and lower probability of having positive inter-temporal increase in debt ratio.

Table 4 Regression result of the effects of CSR (Social Contribution Value: scv) on the capital structure

The estimation result of main explanatory variable is consistent with the prediction of hypothesis in the study, firm’s CSR performance tends to reduce the use of debt financing and use more equity funds. When firm has greater commitment and engagement in CSR and has better CSR performance, information asymmetry problem between firm and financial markets participants is less severe and the firm has a lower cost of capital. Furthermore, firm with better CSR performance pays more attention to the interests of the all stakeholders to achieve sustainability, and tends to raise required funds by issuing equity securities without bankruptcy risk, which makes the debt ratio tend to be lower.

The estimation result of control variables in Table 4 shows that the estimated coefficients on asset, market value to book value, directors' shareholding ratio and directors' shareholding pledge ratio are mostly positive and significant, indicating that firm with larger scale, higher market value to book value, directors' shareholding ratio and directors' shareholding pledge ratio tends to use more debt financing. In addition, estimated coefficients on R&D ratio and institutional investors’ shareholding ratio are mostly negative and significant, indicating that firm with higher R&D ratio and higher institutional investor shareholding ratio tends to use equity funds instead of debt.

Table 5 and Table 6 report the estimation results of the effects of CSR performance (proxied by social return on assets: sroa and social contribution per share: scvps) on the firm's capital structure. Similar to the result of Table 4, in model (1) ~ (8), estimated coefficients on social return on assets and social contribution per share are both negative and reach a statistical significance level, indicating that social return on assets and social contribution per share both negatively affect eight capital structure variables. Firm with higher social return on assets and social contribution value per share tends to use less debt financing. Firm with higher social return on assets and social contribution value per share tends to have lower leverage (debt) ratio, lower probability of having above-mean leverage (debt) ratio, less inter-temporal increase in leverage (debt) ratio and lower probability of having positive inter-temporal increase in leverage (debt) ratio. Estimation result of Tables 4, 5 and 6 generally support the hypothesis of the study, firm with better CSR performance tends to raise funds through equity financing instead of debt financing.

Table 5 Regression result of the effects of CSR (Social Returns on Assets: sroa) on the capital structure

Table 6 Regression result of the effects of CSR (Social Contribution Value Per Share: scvps) on the capital structure

Additional tests

Moderating effect of family control

While a large number of firms in Taiwan's financial market are family controlled firms (Claessens et al., 2000; Yeh, 2005; Yeh et al., 2001), and the impacts of family control on firm’s performance were divided into positive views such as stewardship theory (Davis et al., 1997; Corbetta & Salvato, 2004a, b) and negative views such as agency theory (Jensen & Meckling, 1976; Fama & Jensen, 1983). Because firm’s commitment and engagement in CSR is also an investment which consumes corporate resources, the motivation and consequences of CSR engagement may be distinguished between family-controlled firms and non-family-controlled firms. Under the stewardship theory, family controlled firms are more likely to engage in CSR in order to enhance firm's social reputation, social legitimacy and corporate sustainability, then to limit itself to use debt financing. On the contrary, under the agency theory, the CSR investment of family-controlled-firms are more likely to be used as highlighting the reputation of the family or family members rather than really beneficial to firm’s sustainability, such that the relationship between CSR performance and debt use is weakened in family-controlled-firms.

In order to test the moderating effect of family control on the negative relationship between CSR and capital structure, this study incorporates the cross-product term of CSR and dummy of family-controlled-firm and re-estimates the regression. The estimated results are reported in the model (1) ~ (4) in Table 7. By observing the coefficients on cross-product term, it is shown that they are all positive and reach a statistically significance level, indicating that the effect of performance of CSR in family controlled firms are weaker than the effect in non-family-controlled-firms. Under the agency theory of family firm (Jensen & Meckling, 1976; Fama & Jensen, 1983), the non-separation of ownership and control fosters conflicts of interests between controlling shareholders and other stakeholders, thus weakens of effects of firm’s engaging in CSR on debt use. The effect of CSR performance on capital structure is relatively weak in family-controlled-firms.

Table 7 Regression result of the effects of CSR (Social Contribution Value per Share: scvps) on the capital structure– moderating effects of family control and financial tsunami

Insurance effects of CSR on capital structure

Second, according to Lins et al. (2017), the public tends to have greater confidence and trust toward firms with better CSR performance, thus during the period of loss confidence and trust of financial markets such as the financial tsunami (2008–2009) period, firm with good CSR performance achieved better profitability, sales growth and employee productivity than firms with poor CSR performance. Based on the arguments and findings of Lins et al. (2017), this study proposes that firm with better CSR tends to reduce more of debt use because firm’s profitability and equity price decline is less in period of losing confidence, namely, insurance effect of CSR emerges. To test the moderating effect of financial tsunami period on the negative relationship between CSR and capital structure, this study incorporates the cross-product term of CSR and dummy of financial tsunami period and re-estimates the regression. The estimated results are reported in the model (5) ~ (8) in Table 7. By observing the coefficients of cross-product term of CSR and dummy of financial tsunami period, it is shown that they are all negative and reach statistically significance level, indicating that better CSR performance has greater effects in reducing the use of debt in financial tsunami period. The debt-reducing effect of CSR is more pronounced during the period of losing confidence by the insurance effect of CSR.

Alternative measure of CSR performance

The quantification and precision of a firm's involvement in CSR is a crucial aspect that the researchers and investors considering a firm's CSR performance must pay attention to. Apart from donations or the establishment of charitable foundations and related philanthropic activities, how to quantitatively measures its social responsibility contributions and ensures the accuracy of measures become important consideration. Financial statements primarily disclose financial information, and the disclosure of non-financial information often relies on a firm's willingness to provide it. Even in the case of CSR reports or sustainability reports, there is still a relative lack of examples where firms disclose the actual amount of their investments in various aspects of CSR engagement. Existing studies on CSR performance evaluation often utilize assessment criteria developed by impartial and objective third-party institution to assess and rank firm’s social responsibility performance. These CSR rating agencies include KLD, FTSE 4GOOD Index, and others (Wu & Shen, 2013; Kim et al., 2014; Lins et al., 2017).

In this study, until now the measurement of CSR performance is quantitative which follows the inclusion criteria of constituent stocks of Shanghai Stock Exchange Social Responsibility Index. To increase the robustness of the empirical result, this study employs an alternative approach to measure the CSR performance of sample companies, specifically by examining whether a particular company is included in the prestigious list of CSR performance published by rating agencies. In Taiwan, a highly reputable business magazine, the Common Wealth,³ conducts an annual assessment of the CSR performance of publicly traded firms in Taiwan. Each year, the evaluation criteria and scope may undergo slight adjustments to align with the global dynamics of CSR issues. According to Chang (2011), since 2007, the Common Wealth has conducted an annual survey and ranking of "Corporate Citizens" drawing references from international indicators and evaluation methods such as the United Nations principles, OECD guidelines, and the Dow Jones Sustainability Index. The assessment focuses on four dimensions: corporate governance, corporate commitments, social engagement, and environmental protection. The goal is to select the best corporate citizens among publicly traded firms in Taiwan. Corporate governance primarily measures the independence of the board of directors and the transparency. Corporate commitments include commitments to consumers, employee development and welfare, and investment in innovation and research and development. Social engagement evaluates the firm's social contributions and influence, while environmental protection examines the firm's efforts in environmental conservation and energy efficiency.

In the annual selection process for "Corporate Citizens" by the Common Wealth, it begins by screening profitable companies among over 2,000 publicly traded companies (including those listed on the stock exchange, over-the-counter market, and emerging stock market) regulated by the Financial Supervisory Commission for three consecutive years. Then, more than a hundred institutional analysts, accountants, and experts from the business, government, and academics who have long been concerned with CSR, evaluate and score the firm' performance in the four dimensions mentioned above. After weighting, each company receives a total score. The Common Wealth names the top 50 companies as the "Best Corporate Citizens TOP50," with 30 belonging to the category of "Large Enterprises" with annual revenue exceeding 10 billion, 10 in the category of "Medium-Sized Enterprises" with annual revenue below 10 billion, and 10 "Foreign Companies".

Another reputable magazine within the same business group as the Common Wealth is the Global Views Monthly, which has also received multiple media awards.⁴ The Global Views Monthly focuses on major international and cross-strait trends, important figures, and significant events as its primary content direction. Similarly, it conducts surveys and rankings of CSR performance for listed firms on the Taiwan Stock Exchange and Taipei Exchange Market (the predecessor was the Over-the-Counter Securities Markets), and announces the results on an annual basis.

The Global Views Monthly has been conducting an annual "CSR Survey" in Taiwan since 2005. It refers to the scoring and weighting criteria of OEKOM, a German CSR research institution. The survey assesses social performance, environmental performance, and financial information,⁵ and examines various company-related information and news exposure, including: (1) questionnaire responses and negative news reports; (2) external audits from organizations such as the Environmental Protection Administration, the Ministry of Labor, consumer protection associations, and NGOs; (3) elimination of companies involved in significant labor disputes, environmental pollution cases, major consumer disputes, or cases where business owners are restricted from leaving the country due to litigation; (4) elimination of companies with consecutive three-year operating losses. Companies that pass the various audits and achieve higher scores are awarded the "Corporate Social Responsibility Award" annually.

Based on annual name-list of winners of the Common Wealth "Best Corporate Citizen" and the Global Views Monthly "Corporate Social Responsibility Award", two CSR performance measures are constructed.⁶ First, the current CSR performance (csrdummy), when a firm wins either of the above two awards in a specific year, csrdummy is equal to 1, and 0 otherwise. Second, the cumulative CSR performance (csrcumu), which is defined as the cumulative years that a firm has won either one of the above two awards. For examine, if a firm has won either one of the above two awards for two years since the data starting year, then the value of csrcumu is equal to 2 in the third year.

Table 8 reports the regression result of the effects of CSR performance (measured by csrdummy and csrcumu) on firm's capital structure variables. In model (1) ~ (8), the estimated coefficients on csrdummy and csrcumu are both negative and both reach a less-than 10% statistical significance level, indicating that firms that are selected by the "Best Corporate Citizen" of the Common Wealth and the Global Views Monthly’s "Corporate Social Responsibility Award" tend to have lower leverage ratios and debt ratios, supporting the hypothesis that firm’s better CSR performance helps to decrease firm’s tendency of debt financing.

Table 8 Regression result of the effects of CSR on the capital structure–alternative proxies of CSR performance

Conclusion and suggestion

This study fills the research gap by examining the effects of firm’s CSR performance on capital structure decision, i.e., the use of debt versus equity financing. Based on the data of 1,590 listed non-financial-industry firms on the TWSE and TPEx from 2007 ~ 2020, by controlling firm’s size, market to book value, fixed assets ratio, depreciation ratio, R&D ratio, institutional investors’ shareholding, directors’ shareholding and directors’ shareholding pledge ratio, empirical evidence shows that firm with better CSR performance tends to have lower leverage ratio and debt ratio and thus use less debt financing. The evidence also shows that greater CSR performance is associated with inter-temporal reduce of debt use. Firm’s taking more care of stakeholders’ interests tends to be more conservative in financing decision and enjoys lower equity cost, thus tends to issue equity securities to raise required funds, thereby reducing the leverage ratio and debt ratio year by year. Furthermore, due to the core agency problem, the above effect is relatively weak in family-controlled firms. During the financial tsunami period, due to the insurance effect of CSR, the effect of firm’s CSR performance on reducing the debt use is more pronounced. Overall, the principal outcome supports the hypothesis that firm’s commitment on CSR helps to reduce the tendency of debt financing.

The implication of the study has three folds. First, for investors, firms with better CSR performance tend to use less debt and then incur less financial risk, and the investors who choose to invest in firm with better CSR performance face lower investment risk. Second, for the management, firm’s self-disciplined in stakeholders’ interest protection and avoiding investment or involvement in controversial industries fosters firm to be more conservative in financing decision and incurs less litigation and bankruptcy risk. Third, the government authorities may continue to strengthen the regulation and encouragement of firm’s CSR engagement in order to promote the financial stability of firms and financial markets soundness.⁷