Does capital structure influence the performance of corporate social responsibility? An analysis in companies of the world's largest economies

A estrutura de capital influencia a responsabilidade social corporativa? Uma análise das companhias nas maiores economias do mundo

There is still no consensus in literature about the determinants that lead companies to adopt corporate social responsibility (CSR) practices. From this restlessness, we propose to investigate the effects of capital structure influence on CSR of listed companies related to the world's 10 largest economies in 2019-2020. Prior studies revealed different explanations for the social responsibility practices implementation: it minimizes the firm's risk (Teixeira, Nossa & Funchal, 2011); adds value to companies (Servaes & Tamayo, 2013); reduces informational asymmetry (Cho, Lee & Pfeiffer Jr., 2013); improves corporate reputation (Baraibar-Diez & Sotorrío, 2018).

Notwithstanding that different topic have been researched, most of studies seek to relate CSR to Corporate Financial Performance (CFP); however, the results of these investigations differ from one another (McWilliams & Siegel, 2000; Campbell, 2007; Carroll, 2008; Saeidi et al., 2015; Cantino, Devalle & Fiandrino, 2017; Yang, He, Zhu & Li, 2018).

Certain studies investigate the effects of CSR on CFP. Researchers found a positive and significative results in this correlation. For example, while analyzing Spanish companies, Rodriguez-Fernandes (2016) verified that corporate policies focused on responsible social actions improve positive financial returns, being the organizations that most invest, those with the highest financial returns; and, according to Wang and Sarkis (2017), the CSR-based management renders into greater financial results.

In turn, positive and significant results of the relationship between CSR and CFP are not always confirmed in empirical studies (McWilliams & Siegel, 2000; Saeidi et al., 2015). Nollet, Filis and Mitrokostas (2016), whilst analyzing the companies that compose the S&P 500 index, inferred that even CSR has a positive coefficient in relation to CFP, nevertheless not significant. Chen, Hung, and Wang (2018) found negative association in Chinese companies, denoting that companies with spending on CSR decreased profitability.

Such divergences in results, as already pointed out by McWilliams and Siegel (2000), are not surprising. Part of the problem lies in the econometric models’ specification used and the other part exists in the fact that the theoretical omission CSR creates intangible attributes for the firm, such as reputation, legitimacy, and reliability. Therefore, these other elements suggest that the relationship between CSR and CFP is more complex than is usually addressed in studies and that, actually, it does not occur directly (Saeidi et al., 2015).

In this perspective, a group of researchers argues that CFP would only be a consequence of the association between CSR and corporate capital structure (Teixeira et al., 2011; Almeida & Santos, 2016; Yang et al., 2018; Hamrouni, Boussaada & Toumi, 2019), which we also support in this research. This linkage results from change of thoughts, in which political and community forces seek convergence between financial and social goals (Campbell, 2007; Carroll, 2008; Chen et al., 2018). Thereby, the logic propagated worldwide is that companies should not only maximize the returns and defend shareholders', but also consider all stakeholders (Al-Dah, Dah & Jizi, 2018; Baraibar-Diez & Sotorrío, 2018).

The relevance of stakeholders to companies is noticeable since they validate CSR actions. Though, validation will only occur when such practices meet their expectations of what would be appropriate and acceptable. Given this fact, managers ponder the pressures received by the company for the implementation of responsible social practices (McWilliams, Siegel & Wright, 2006; Hamrouni et al., 2019).

Researchers argue that CSR has, above all, a relationship with the capital structure of companies, since this arrangement represents the materialization of the stakeholders’ influence on organizations (Yang et al., 2018; Hamrouni et al., 2019). This influence results from the operating activities’ financing by debts contracting with its stakeholders. Thus, involvement with socially responsible actions impacts the access to finance by organizations, whether through equity or third parties (Almeida & Santos, 2016; Hamrouni et al., 2019).

Accordingly, the organizations involvement in socially responsible actions would impact on access to funding for organizations, either through shareholders’ equity or debt (Yang et al., 2018; Hamrouni et al., 2019). From shareholders’ equity perspective, current shareholders search for organizations with CSR practices and, thus invest more resources in these companies (Teixeira et al., 2011). As far as debt is concerned, CSR actions lead to greater ease of access to debt contraction, since this behavior increases third-party reliability in the company (Almeida & Santos, 2016; Hamrouni et al., 2019).

These contradictions in CSR literature let research such as ours demonstrate the relevance of studying the reasons that drive responsible social practices, as well as the related effect of capital structure on CSR. Our findings show that capital structure and shareholder’s equity correlates positively with CSR performance, while the relation between debt and CSR performance is negative. Furthermore, the results demonstrate variations in different countries and sectors.

These findings contribute with CSR literature by showing that the higher levels of transparency and disclosure of CSR practices boost investments and reliability from stakeholders, whereas the debt financing decreases due to investments on equity. We trust that this debate on the results can be helpful to other researchers obtain new insights into CSR practices.

Conceived as a global phenomenon, CSR marks a shift in thinking about what interests’ companies should address (Carroll, 2008). It occurs because management focused exclusively on the interests of shareholders presents problems, as other stakeholders of the company that are important to its operation may be overlooked, such as the Government, Suppliers, Customers (Teixeira et al., 2011). As a consequence, the perspective shift from shareholders to stakeholders reinforces CSR's widespread view that all parts of the organization are important and should have their expectations met (Baraibar-Diez & Sotorrío, 2018; Yang et al., 2018; Garcia, Mendes-da-Silva & Orsato, 2019).

The duality between the shareholders and stakeholders’ objectives does not cease to exist. Because of its importance, conscious companies tend to avoid actions that could harm their stakeholders, and if it does, they try to correct the incident and recover the damage (Campbell, 2007), due to the impact that other parties may have on the entity. Nevertheless, firms' goal remains to maximize shareholder wealth, even when facing increasing external pressures to meet social expectations (Al-Dah et al., 2018).

However, as a matter of fact, not all companies are on the same level of behavior and assimilation of responsible social practices. Carroll (2008) pointed out that there are differences in CSR levels from sector to sector, as responsibilities change. In McWilliams, Siegel & Wright (2006) view, CSR demands vary by country, region, and industry. Similarly, Garcia et al. (2019) consider that these differences depend on economy characteristics in which companies are inserted, as well as from the different stages of economic development.

Hamrouni et al. (2019) emphasize that the implementation of CSR policies in organizations involves ethical, legal, sustainability and reputation aspects. Because of this, the top management of companies ponder the implementation of CSR, considering the distinct aspects and consequences that occurs by the adoption of this practice (McWilliams et al., 2006).

Pursuant this line of argument, responsible social practices are considered by companies, choosing the capital structure (Almeida & Santos, 2016). Since the financing of operating activities comes from the contracting of external debts, the organizations take the decision between third parties or shareholders' equity. According to Yang et al. (2018), the capital structure affects decision making about CSR strategies, because the adoption of that corporate practice improves the environment and reduces part of the inherent business risks.

In relation to equity, Ferris, Javakhadze and Rajkovic (2017) argue that this group is constantly looking for investments that prove to be socially responsible. This engagement causes shareholders to put pressure on organizations to take up CSR-oriented practices. By means of CSR practices, the company's value increases, as it demonstrates a commitment to all stakeholders to improve the relationship between internal and external parts of the organization, which would encourage the shareholder to maintain or increase its capital in the entity, and also, would attract new investors (Cho et al., 2013; Borghesi, Houston & Naranjo, 2014).

The third-party capital is composed by debts contracted by the companies for its operational leverage. The CSR practices promote the credit access since reduce information asymmetry issues and insolvency risks (Almeida & Santos, 2016; Chen et al., 2018). Besides this, the ease on credit access also derives from the reputation generated by the CSR. Because of trust increasing between company and its creditors and thus, as a reward for high levels of CSR, there would be this easier access to debt taking (Hamrouni et al., 2019).

Hence, given the above, we formulate the following research hypothesis:

1. H(1): The corporate social responsibility index is high (low) when corporate capital structure increases (decreases).

2. H(1a): The corporate social responsibility index is high (low) when debt increases (decreases).

3. H(1b): The corporate social responsibility index is high (low) when shareholders’ equity increases (decreases).

Listed corporations of the 10 highest GDP countries in 2019, with CSR index represented by the ESG score (environment, social & governance) provided by Thomson Reuters® database, for the 2010-2018 period, represent our sample for the study. According to the list released by the International Monetary Funds (IMF, 2019), the ten countries ranked globally with the highest GDP are: United States of America (US $ 21,439 billion), China (US $ 14,140 billion), Japan ($ 5,154 billion), Germany ($ 3,863 billion), India ($ 2,963 billion), United Kingdom ($ 2,744 billion), France ($ 2,707 billion), Italy ($ 1,989) billion), Brazil ($ 1,847 billion) and Canada ($ 1,731 billion).

We adopted the cut from 2010 due to the availability of the ESG performance index on the Eikon platform (Thomson Reuters database) for publicly traded companies as well as the removal of financial sector companies, resulting in a total of 1,660 companies. From this total of 1,660 we excluded companies with negative equity, totaling in the final sample 1,642 companies represented by 14,370 observations, as per Table 1.

Table 1
Data sample distributed between the 10 highest GDP countries by economic sectors

Source: research data. Note. USA (United States of America), CHI (China), JAP (Japan), GER (Germany), IND (India), UK (United of Kingdom), FRA (France), ITA (Italy), BRA (Brazil) and CAN (Canada); S1 (Communication Services), S2 (Consumer Discretionary), S3 (Consumer Staples), S4 (Energy), S5 (Health Care), S6 (Industrials), S7 (Information Technology), S8 (Materials), S9 (Real Estate) and S10 (Utilities).

The sample of 1,642 companies is composed by ten economic sectors: industrials (331), consumer discretionary (265), materials (201), information technology (171), energy (135), consumer staples (117) , real estate (116), health care (111), communication services (98), and utilities (97), distributed among the ten highest 2019 GDP countries: United States of America (632), China (78), Japan (338) , Germany (60), India (42), United Kingdom (201), France (73), Italy (21), Brazil (27) and Canada (170).

The data on Table 2 evince the highest values of CSR, equity and debt of companies distributed by economic sector on each country. Analyzing by sectors, we note the following the higher performance of CSR on communication services (S1) correspond to French companies, on consumer discretionary (S2) to German companies, on consumer staples (S3) to American companies, on energy (S4) to Canadian companies, on health care (S5) to UK’s companies, on industrials (S6) to German companies, on information technology (S7) to American companies, on materials (S8) to Japanese companies, on real estate (S9) to American companies and on utilities (S10) to Italian companies.

Regarding the higher values of equity, we have on communication services (S1) correspond to Japanese companies, on consumer discretionary (S2) to German companies, on consumer staples (S3) to American companies, on energy (S4) to UK companies, on health care (S5) to American companies, on industrials (S6) to UK companies, on information technology (S7) to Japanese companies, on materials (S8) to Canadian companies, on real estate (S9) to UK companies and on utilities (S10) to Indian companies.

At last, in relation to the higher values of debt, we have on communication services (S1) correspond to American companies, on consumer discretionary (S2) to Brazilian companies, on consumer staples (S3) to American companies, on energy (S4) to UK companies, on health care (S5) to Canadian companies, on industrials (S6) to American companies, on information technology (S7) to American companies, on materials (S8) to UK companies, on real estate (S9) to American companies and on utilities (S10) to Japanese companies.

Table 2
Cross-data between companies of the 10 highest GDP countries and economic sectors

Source: research data. Note. USA (United States of America), CHI (China), JAP (Japan), GER (Germany), IND (India), UK (United of Kingdom), FRA (France), ITA (Italy), BRA (Brazil) and CAN (Canada); S1 (Communication Services), S2 (Consumer Discretionary), S3 (Consumer Staples), S4 (Energy), S5 (Health Care), S6 (Industrials), S7 (Information Technology), S8 (Materials), S9 (Real Estate), S10 (Utilities) and ST (Sector total). Letter (C) correspond to ESG index; letter (E) to Equity; and letter (D) to Debt.

When crossing the data from CSR indexes, equity, and debt values between countries by economic sectors, the sectors that show the highest performance in CSR are information technology (S7) companies of United States, Germany, and India, followed by communication services (S1) French companies and German industrials (S6). Whereas the highest equity values are present in the materials (S8) sector in Canadian companies, followed by German consumer discretionary (S2) companies and UK energy (S4) sector companies. As for debt, the sectors that demonstrate the highest values are Brazilian and American consumer discretionary (S2) companies, followed by UK energy (S4) companies and American real estate (S9).

In particular, we also note that among the values of the CSR indexes, equity and debt, Chinese companies of the ten sectors presented did not reveal high levels in these variables, despite China being the second place in the ranking of the ten countries with the highest GDP in the world in 2019.

According to Table 3, the median ESG score is 56.52, which represents the central proxy measure used as RSC. Considering that the ESG index has a scale from 0 to 100, we understand the median value 56.52 of the sample is in the data’s middle, that is, at least half of the observations are less than or equal to it, and at least half are greater or equal to it. Therefore, despite the amplitude of 90.63 in the ESG score observations, we can note that the CSR practices of companies in sample have not yet been widely implemented. In other words, these companies have not yet fully developed these actions.

Although there have been changes in thinking resulting from the spread of CSR in recent years, as commented by some researchers (Carroll, 2008; Baraibar-Diez & Sotorrío, 2018; Yang et al., 2018; Garcia et al., 2019), we believe that this behavior stage occurs, since we are in a process of change, it has not completed entirely. Thus, we expect that in the coming years the CSR practices median of companies around the world may increase as the responsible social actions are more assimilate in the corporate environment.

Table 3
Descriptive statistics of variables

Source: research data.

By the coefficient of variation, we notice that the dispersion of data variation concerning book value per share, ROE, free cash flow, market value in relation to its mean is higher than the others. Likewise, the dispersion of debt, age and ROA data is high, as it corresponds to a variation greater than 70%. Regarding the ESG score, capital structure, equity and tangibility data, the dispersion occurs between 23% to 44%, while the size data variability is low, presenting a dispersion of 7% in relation to its average. Therefore, we can infer that the capital structure and size data set are homogeneous with each other, while the other variables present a heterogeneous data set.

Following the model’s estimates presented in equations (1), (1a), (1b) and the following procedures for the short panel, it performs the first application using the POLS estimator. Observed by the tests applied after regression, there was not the presence of collinearity (VIF 1.56), however the residues were identified heteroscedasticity. Thus, it again uses the POLS method with robust clustered standard errors for heteroscedasticity correction.

However, considering that the dependent variable, the ESG score, presents continuity in time, i.e., past values explain present and future values, it is necessary to find a method that solves the issue of endogeneity. Consequently, we adopt the instrumental variables in the model, operating in the capital structure variable the lag of one and two periods in the dynamic specification. As an efficient proposal to solve, or at least reduce endogeneity problems, are used estimators known as GMM. The equations (1), (1a), (1b) results are provided in Table 4.

The GMM 2SLS estimator were applied in the regression model of equations (1), (1a) and (1b). The estimation results shown in Table 1, represented by its coefficients and significance values, differ between capital structure, debt and equity correlated to CSR (ESG score).

Table 4
Estimation results on GMM 2SLS

Source: research data. Note. ***, **, * Indicate statistical significance at the 1 percent, 5 percent, and 10 percent levels, respectively USA - United States of America, CHI - China, JAP - Japan, GER - Germany, IND - India, UK - United of Kingdom, FRA - France, ITA - Italy, BRA - Brazil and CAN - Canada.

6.3 Hypothesis (1b): The effect of shareholder’s equity on CSR

Following the results between CSR and Capital Structure, the findings between CSR and Equity also reveal a positive and statistically significant relationship, either for the control variables: size, age, market value and tangibility of the company. The outcomes of FCO and ROA are significant, respectively p > 0.000 and p > 0.004, whereas its parameters are negative.

Thought-provoking in our study are the parameters’ results of debt and equity financing. The first one is negatively correlated to CSR practices and the second one has a positive correlation. Analyzing its coefficients and signals (-4,060; +3.013), we can infer that, in order to companies have a consolidated reputation, the internal and external pressures act intensively to obtain an optimal score of environmental, social and governance (CSR practices). Thus, the effects of equity financing are positive on CSR practices due to reduction of risks, informational asymmetry, and costs of equity (Teixeira et al., 2011; Cho et al., 2013), provoking a decrease on debt financing, even if a high score of CSR foments an access to credit. Relevant and higher levels of transparency and disclosure of CSR practices boost investments and reliability from stakeholders, while the debt financing decreases due to equity growth investments.

For all ten countries, the results reveal statistical significance at the 1% level, while for economic sectors only energy, health care, information technology, materials and real estate sectors are statistically significant. Furthermore, the model developed by Allouche and Laroche (2005), when performing a meta-analysis in studies concerning the effects of CSR on CFP, demonstrate the relevance of identifying economic sectors and countries. Similarly, McWilliams et al. (2006) reveal the importance of the qualification of the country, region, and the business branch of companies in models with the CSR theme.

For comparison reasons between the effects of equity and debt financing, in Figure 1 and Figure 2, the graph shows the energy sector analysis for companies in China and Italy.

Figure 1
CSR index and Debt of China and Italy for energy sector
Source: research data.

Figure 2
CSR index and Equity of China and Italy for energy sector
Source: research data.

In short, when observing both regression lines, we verify that to achieve a certain CSR index, Chinese energy companies need a larger equity financing than Italian companies as well its debt financing reduces more than Italian companies.

The graphs of Figure 3 and Figure 4 represent the economic sector of consumer discretionary and real estate distributed across the ten countries of this study. Each point indicated on the chart line shows the equity and debt financing correspondence for that sector.

Figure 3
Consumer discretionary by countries on relevant Equity and Debt
Source: research data.

Figure 4
Real Estate by countries on relevant Equity and Debt
Source: research data.

The graphs presented show variation of consumer discretionary and real estate sectors for each country. We highlight that for China to achieve a positive CSR index, the contribution needed for investment in equity of its public enterprises is higher compared to other countries.

We verified that there is no evidence that conduct us to reject our hypothesis (1) and hypothesis (1b), and thereby, our findings reveal that capital structure and shareholder’s equity correlates positively with CSR performance. However, our hypothesis (1a) related to the effects of debt financing on CSR, we have to reject it once our findings show a negative correlation. It is important to highlight that the variations found between countries, especially companies from nations with higher GDP, need a larger capital structure than smaller ones to obtain a positive CSR index. We believe this contrast is due to the companies' size, because of these ones are larger than those presents in countries of lowest GDP. As a consequence, the companies of the highest countries GDP need a larger capital structure to finance all of their operating activities, plus CSR actions.

In addition, other results deserve attention. First, we highlight that the financial performance variables of the model presented negative coefficients, which indicates that the expenses incurred to increase CSR performance does not generate returns. We make this observation due to the short-time period analyzed in this study. We also emphasize that CSR affects the company's reputation. This, in turn, is directly related to financial performance. However, it should be noted that corporate reputation derives from a process that does not occur overnight. Larger reason for inferring that CFP occurs in the long term. Thereby, we encourage further research to verify such relationships in a larger time frame.

Second, it concerns the divergence between the parameter signals of firm's market value and the book value per share as well as its significance, present only in market value. These results are not surprising, since the different purpose of each variable. The first reveals the value attributed by the market to the organization, and in counterpart, the second that displays the value based on the company's accounting. We understand that the positive and statistically significant findings in the relationship between market value and CSR index is due to the fact that CSR has an intangible asset in its constitution: the corporate reputation. Indeed, the market ponders CSR actions and judges whether they should be accepted. The result converges into the acceptance of these practices which generate a firm's value, justifying the positive and significant association.

In the meantime, the results have a close connection with the ESG score used as a proxy for CSR, which we consider a limitation of the study. Despite being a methodological choice, there are several indicators of CSR with different elements and weighting, thus, the results could differ one from another. In order to consolidate the results presented in this research, we encourage further research with different CSR indicators.

Beyond that, we would like to provoke a reflection. Are companies with leveraged capital structure more sustainable or just better disseminating CSR strategies? As with most CSR indexes, the ESG score is also based on disclosures of social, environmental and governance aspects. Though, what is publicized by companies actually performed by companies? Or is it just a window dressing to meet the demands and lessen the external pressures of society? We believe that the answers for these questions can be found in research that look inside companies.

We hope to contribute to the literature from the results obtained and the comments made. We seek to foster new discussions on the issues surrounding the adoption of socially responsible practices by companies from different corners of the world. We understand that the debate does not end in this paper, and it needs new voices and evidence, as we live in a society in constant change and, increasingly, new factors have influenced organizations.