An empirical study on the relationship between corporate social performance and financial performance in automotive Industry
Student Number: 3102378 Student Name: Xiaoxi Luo Email address: x.luo.6@student.rug.nl
Supervisor: Dr.J.V. Tinang Nzesseu
Abstract
Corporate social responsibility (CSR) is widely recognized as a multidimensional construct in theory, however, the empirical research regarding the relationship between different dimensions of CSR and corporate financial performance (CFP) is still relatively lacking. This study aims to measure CSR in terms of six stakeholder dimensions including shareholders, creditors, employees, government, suppliers and consumers, and investigate how each dimension affect CFP for companies in the automotive industry. The influence of country on CSR-CFP relationship is also examined in this study. The data of 413 companies in China, Germany, US and Japan was collected and multiple statistical analyses were applied to answer the research questions including cross-sectional linear regression and panel fixed effects model. While all dimensions of CSR are expected to have positive impact based on our theoretical analysis, only some dimensions’ impact are statistically significant. Furthermore, the significant dimensions and their effect sizes are also different depending on the country. These findings give insights to managers regarding which dimensions of CSR activities are most likely to improve the CFP of their companies.
1. Introduction
From the view of classical economics, the primary responsibility of company is to provide qualified products and services to the society in order to obtain profits. However, an enterprise is not only an economic entity but also part of society and it eventually depends on the society to achieve sustainable profits (Crane, McWilliams, Matten, Moon, & Siegel, 2008). Therefore, companies shall not sacrifice the interest of other stakeholders and the whole society while striving for higher profit. Moreover, consumers have increasing concerns regarding social, environmental and ethical issues such as whether products are environmental-friendly and whether companies are employee-friendly. In order to satisfy these demands, companies have to implement a strategy dedicated to relevant issues (Inoue and Lee 2011).
These strategies are usually termed as Corporate Social Responsibility (CSR) or Corporate Social Performance (CSP)1. Its official definition is “A business organization’s configuration
of principles of social responsibility, processes of social responsiveness, and policies, programs, and observable outcomes as they relate to the firm’s societal relationships (p. 693) (Wood 1991)”. CSR has become a standard component of company strategies and many companies even assigned it as the highest priority issue (Choi et al. 2010).
Given the necessity of enterprises to fulfill their social responsibilities, the biggest concern of profit-seeking companies is whether CSR activities have a positive impact on company performance. Among all aspects of company performance, the relationship between CSR and corporate financial performance (CFP) is especially of interest.
In the past few decades there has been abundant academic studies investigating the relationship between CSR and CFP. Margolis and Walsh (2003) reviewed the research literature on the relationship between corporate social responsibility and financial performance from 1972 to 2002. There were 127 empirical research reports and 131 research results. They found that there were 19 studies on this issue in the 1970s, 31 in the 1980s, 71 in the 1990s, and 10 in 2000 to 2002, which suggests that scholars' attention to this issue does not diminish over time. However, the results are greatly divided and inconsistent, and the research stream remains inconclusive. Most studies found a positive correlation between CSP and CFP (Comenll and Shap 1988; McGuire 1988; Preston and O'Bannon 1997; Ruf et al., 2001; Barles et al., 2002) Some other studies instead found correlation between CSP and CFP to be (Vance 1975, HoIman et al., 1985). Finally, there are also research which did not detect any significant correlation between CSR and FP, which implies that CSR performance does not affect the financial performance of companies (Fogler and Nutt 1975; Mcwilliams and Siegel 1997).
Several meta-analyses were conducted in order to extract consensus from existing research and grasp the trend of research. Orlitzky et al. (2003) analyzed the findings of 52 studies and conclude that corporate social responsibility is likely to have a positive impact on financial performance. Margolis and Walsh (2003) also conducted a meta-analysis of 127 studies on the relationship between CSR and CFP. They found that among the 109 studies which used CSR as an explanatory variable and financial performance as the dependent variable, 54 studies concluded that there is a positive correlation between CSR and FP. Only 7 reported a negative relationship, 28 studies found no significant correlation and 20 had mixed results.
Several reasons may explain the conflicting findings, the main reasons include problems regarding the measurement of CSP and the use of multi-industry samples (Godfrey and Hatch 2007; Inoue and Lee 2011; Ruf et al. 2001). First, most studies only use a single indicator or variable for CSR in their model, while CSR may actually have multiple dimensions and the impact of each dimension on CFP is different. For example, Hillman and Keim (2001) divided CSR into two parts, namely measure of stakeholder management and measure of social problems. They found that the management of stakeholders can increase CFP, but the participation of enterprises in social issues reduce CFP. Second, most previous research literature on CSR and CFP used multi-industry sample. However, the specific relationship between CSR and CFP may depend on the specific environment: different industries have different degrees of emphasis on social responsibility and unique social interests based on their internal and external environments, therefore the relationship between CSR and CFP may differ depending on the industry(Inoue and Lee 2011).
In order to address these issues, research on the relationship between CSR and CFP should be more “fine-grained”(Miles Morgan and Covin Jeffrey 2001), such as use CSR variables in multiple dimensions and use sample within a single industry. The latest studies are increasingly taking these specificities into account, such as studying the impact of a single dimension of CSR (Schuler and Cording, 2006), the relationship between CSR and CFP in a single industry (Mooer, 2001) or a single region/country (Orlitzky, 2005). These studies have enriched the research findings and made the research conclusions more specific and detailed.
Apart from the influence of different industries, the countries where companies are located in may also affect the relationship between CSR and CFP: the differences in culture and level of economic development may have a tremendous impact on stakeholders’ attitude and behavior towards CSR (Apostolides and Boden 2017). Most previous studies on CSR-CFP relationships were conducted based on companies in developed countries such as US and European countries (Saeidi et al. 2015); their results may not be able to be generalized to other developing countries. There is hardly any comparative studies focusing on contrasting the CSR-CFP relationships between different countries: although Aras et al. (2010) argue that there is no difference between developed and developing economies, more empirical evidence is needed.
In order to address the aforementioned research gaps, this paper aims to explore the relationship between CSP and CFP in the automotive industry. In order to enable a comparison between countries, we collected data of 413 companies in China, the United States, Japan, and Germany between 2014-2018. The main contribution is twofold: first, investigate the relationship between multiple dimensions of CSR and financial performance in the automotive industry; second, compare the CSR-CFP relationship in different countries; third, provide insights regarding which dimension of CSR activities are more beneficial for CFP.
Section 2 briefly introduces the theoretical framework regarding the multidimensionality of CSR and the relationship between CSR and CFP. Section 3 discusses the methodology for data collection and analysis. Section 4 presents the results of data analysis and the final section concludes with the implications of the results.
2. Theoretical background and research framework
2.1 Theoretical framework for the relationship between CSR and CFP
From the view of neo-classical economics, the implementation of CSR activities will consume resources and increase operating costs, which may lead to a decline in financial performance and violate the management’s responsibility to shareholders. Others argue that management is also responsible for social causes which may not directly benefit the company. Earlier research of corporate social responsibility therefore mostly takes a philanthropic view when discusses the responsibility of the enterprise to the society.
other stakeholders can also place demands upon the companies. Companies must deal with these demands as well, otherwise it may lead to negative consequences including confrontations with stakeholders which may reduce the value for shareholders (Ruf et al. 2001). Wood and Jones (1995) pointed out that stakeholders play at least four roles in corporate social responsibility: (1) stakeholders are the recipients of corporate social behavior and bear the influence of corporate social behavior; (2) Stakeholders evaluate the impact of corporate social behavior; (3) Stakeholders anticipate corporate social responsibility behavior; (4) Stakeholders can influence and promote corporate action based on their interests, expectations and tolerance. In short, stakeholders define the norms of corporate social behavior. Wood (1991) argues that stakeholders are the best answer to the question of “who should be responsible for the business”, because stakeholders determine the company’s code of conduct, set the normative content of corporate social responsibility, and different stakeholders. Therefore, from the perspective of stakeholder theory, CSR can be defined as how a company satisfies the demands of multiple stakeholders.
stable development of the company. These are expected to have a positive impact on company’s financial performance.
Transaction realization: a company is essentially a carrier of a series of explicit and implicit contracts with its various stakeholders. There are implicit contracts between the enterprise and various stakeholders. These contracts stipulate the mutual power and responsibility of each group. Each stakeholder provides resources for the company, including the shareholder's equity, creditor's investment, and employees. Human resources, market capital of consumers and suppliers, public environmental resources of the government and the community. These resources constitute the basis for the company’s development. The company must also fulfill its corresponding responsibilities to all stakeholders: the relevant business decisions of the enterprise need to consider stakeholders’ interest and give them corresponding compensation. Therefore, the company and the various stakeholders are essentially a transaction relationship. The company fulfills the CSR in order to improve the efficiency of the transaction. If the company does not pay attention to the fulfillment of social responsibility, it will be hard to continue obtaining resources and a good external environment from all stakeholders. It is more likely to bear the relevant risks and its financial performance will also be negatively affected. Value creation: By fulfilling CSR, enterprises can improve their image and reputation, increase labor efficiency, enhance their relationship with the external environment, reduce the operating costs and risks of enterprises, and create higher value. Via fulfilling stakeholder demands, companies can also gain competitive advantages by developing required skills which other competitors cannot imitate (Ruf et al. 2001). The fulfillment of CSR by enterprises not only guarantees the interests of all stakeholders but also increases their own value while contributing to the whole society. Therefore, the implementation of CSR by enterprises is a win-win operation mechanism of “self-interest and altruism”, which can achieve long-term financial operation benefits for enterprises and realize the goal of maximizing profits.
All the above mechanisms contend that improving CSP is expected to result in higher financial performance. Therefore, the following hypothesis is proposed:
Hypothesis 1 ： There is a positive correlation between a company’s corporate social performance and financial performance.
2.2 Multi-dimensionality of CSR and selection of dimensions
on CFP. The Hillman and Keim (2001) study mentioned in the introduction demonstrate that stakeholder management and social issue participation have opposite impact on CFP. Barnett and Salomen (2006) find that CSR addressing community relations increase CFP while CSR in the dimension of environmental and employee relations decrease CFP. Therefore, using a single measurement for CSR in CSR-CFP relationship studies can also mask the diverse relationships between various dimensions of CSR and CFP.
Given the necessity of using multi-dimension measurement for CSR, we have to decide which specific dimensions to include in our study. One of the biggest contributions of the introduction of stakeholder theory to CSR studies is that it clarifies the definition of corporate social responsibility, thus enabling more accurate measurement of CSR and studying CSR quantitatively. Measuring corporate social responsibility from the perspective of stakeholder theory has become the mainstream method in latest research. Based on stakeholder theory, CSR activities involve a wide range of voluntary activities regarding relations with multiple stakeholders (Benabou and Tirole 2010; Choi et al. 2010).
According to the decomposition of CSR, we revise Hypothesis 1 as follows:
Hypothesis 1: There is a positive correlation between each dimension of corporate social performance and financial performance.
2.3 The influence of countries
The relationship between CSR and CFP is expected to depend on the location of the company: the economic and social environment of different countries will affect the scope and depth of corporate social responsibility, which in turn influences the ability of companies to commit to their CSR and the relationship between CSR and CFP. Possible moderating mechanisms include level of economic development (Arsoy et al. 2012), cultural differences such as the distinct legal tradition (Liang and Renneboog 2017) and institutional environment (Xie et al. 2017) in different countries. Therefore, we propose the following hypothesis:
Hypothesis 2：The country in which a company is located moderates the relationship between CSR and CFP. More specifically, the relationship between CSR and CFP vary in different countries.
3. Research methodology
3.1 Measurement indicator design and operationalization 3.1.1 Measurement of Corporate Social Responsibility
In section 2 we have already specified the six dimensions of CSR measurement based on stakeholder theory. This section introduces the indicators we used to measure each dimension. According to our theoretical framework and research questions, our measurement approach has to meet the following requirements: first, it has to include multiple dimensions which allows a comprehensive measurement of CSP in several aspects; second, the data collection required by the measurement approach cannot be too cost- or effort-consuming since this research project is rather limited in terms of duration and budget. Apart from these two requirements, measurement approach can also be evaluated based on its flexibility (whether researchers are free to choose the dimensions and measurement indicators) and level of bias (whether the measurement approach is subject to multiple types of bias such as selection bias, reporting bias, etc.).
mentioned above. Then we will present the details regarding the selection and construction of indicators for CSR in our study.
Third-party CSR indices: this is the most frequently used way of CSR measurement in
academic studies regarding CSR. These CSR indices are issued by independent specialized third-party rating agencies. It can be further categorized into two classes, namely reputation indices and social audits. Reputation indices are derived by surveying business professional and students regarding companies’ CSR-related reputation (Orlitzky et al. 2003). An example is Fortune magazine reputation index. Some studies use it as a surrogate for companies’ CSR activities. Another category is social audit of CSP. These indices are assessment based on companies actual CSP behavior and activities. Major indices include CEP index (Bragdon & Marlin, 1972; Folger & Nutt, 1975; Spicer, 1978), KLD index (Waddock, Graves, 1997; McWilliams, Siegel). , 2000) Different indices mostly have similar themes, but vary in terms of specific dimensions and indicators, number of dimensions available, selection of geographic locations and industries (Galant and Cadez 2017; Igalens 2005). There is not yet consensus regarding which index is “the best” but KLD is the most commonly used.
CSR indices are usually multi-dimensional: take KLD index as an example, it includes a separate score for each dimension and also a composite score for the general CSR performance. According to different research designs, researchers can choose different combinations of dimensions in their study. Regarding data availability, it minimizes data collection effort if you have access to an index database; however, most indices are not publicly accessible. Moreover, most major indices are focused on US and European companies; therefore they are not applicable if the subject of study includes developing countries. One of third-party indices’ major shortcoming is that it is subject to several types of bias: first, the scores for CSP are compiles by private rating agencies which have their own interest and agenda; second, it also suffers from selection bias: the inclusion of companies in third-party index databased is not random and not independent of the CSR related decisions of these companies. This omission results in a restricted sample and can harm the accuracy of results of the analysis (Shahzad and Sharfman 2017). Another weakness of third-party indices is that it is not so flexible: researchers can only select the dimensions and companies which are available in the index (although some indices have a rather comprehensive coverage of dimensions); moreover, the scores may not be based on the latest scientific methods and the specific procedure of scoring is not fully transparent.
Content analysis:
can be used in the subsequent analyses (Galant and Cadez 2017). When it is applied to measure CSR, the source of the qualitative information mainly consists of various types of reports or documents that the company has published, including the company's annual reports, letters to shareholders, and other corporate disclosures (Orlitzky et al. 2003). Researchers generally determine the specific items measuring CSR based on the focus of the specific study, identify corresponding data from various reports or documents, and assign a certain score for each dimension. There are various ways to recode qualitative information into quantitative indicators: 1) counting words or sentences on the specific CSR issue and assign binary variables if this issue is mentioned (Aras et al. 2010); 2) instead of binary scores, a more advanced way of coding is to assign interval scores which is similar to Likert scales (Chen et al. 2015).
Many previous studies used content analysis to generate CSR measurements. Bowman and Haire (1975) conducted a study of the 1973 annual report of 82 US food companies. The analysis analyzes the length of the annual report used to report corporate social responsibility and corporate social activities as a measure of corporate social participation. Ingram (1978) analyzed the inclusion of Fortune 500 and completed between 1971 and 1975. The annual report of 287 companies that conduct market transaction data divides the social responsibility information disclosed in the annual report into two categories: monetary information and non-monetary information, as well as environmental, fair management, employees, community activities and products.
The key advantage of content analysis is its flexibility: researchers can specify any number of CSR dimensions and collect data corresponding to these dimensions. The coding process is also entirely under researchers’ control. However, this flexibility also leads to the biggest disadvantage of content analysis, namely its subjectivity. The freedom of researchers in each step of the research process renders the final data subject to individual bias.
In addition, a point worth mentioning is that the data source of content analysis is mainly the disclosure of CSR activities instead of the actual CSR performance of companies. This is actually an entirely different concept from CSP which may not be a good proxy for actual CSR activities: the level of disclosure and actual performance can be different (e.g. high level of performance but low level of disclosure); moreover, their impact on CSP can also be different (Balabanis et al. 1998). This leads to another major drawback: reporting bias. First, disclosure of CSR activities is mostly voluntary and many companies may simply fail to report the CSR activities they conducted; second, companies may also attempt to create a more favorable image via biased reporting.
Questionnaire: when the companies/industries/countries researchers wish to study are not
relevant dimensions and items included in the questionnaire depending on the research question. Therefore, it is also highly flexible and usually covers multiple dimensions. Its disadvantage is response bias: first, companies which place more importance on CSR and behave better in terms of CSR are more likely to respond to CSR related questionnaires; second, respondents may be inclined to give answers which are more socially desirable than their actual behavior. Therefore it is harder to ensure the quality of data (Zhou et al. 2016).
One-dimensional constructs: some studies choose to measure CSR with indicators from a
single dimension. This approach is more common when the focus is the environmental and human rights (philanthropy) dimensions of CSR (Galant and Cadez 2017). If CSR is only measured with one dimension then data availability is obviously higher than multi-dimensional measurements, but it is theoretically problematic according to our theoretical framework which clearly demonstrates the multi-dimensionality of CSR.
Table 1 lists the features of each measurement approach in terms of the four criteria mentioned above.
Table 1. Summary of different measurement approaches of CSR
Method
Multi-dimensional
Flexibility Data availability Free of bias Third-party
CSR indices
√ × (Methods are
not transparent)
Easy to collect but not free, not available in all countries × (selection bias) Content analysis √ √ √ × (subjective, reporting bias)
Questionnaire √ √ Can be costly and
time-consuming × (response bias, selection bias) One-dimensional measures × √ √ √
collect data via questionnaire. At present, empirical research on CSR of Chinese companies is also mainly based on the content analysis method of corporate social responsibility information disclosure.
Our approach of content analysis differs with most academic studies in English language in one aspect: instead of collecting qualitative information and coding them into quantitative indicators, we decided to use stakeholder-related financial indicators as proxy for each stakeholder dimension of CSR. This is a popular method in Chinese language CSR studies (Jia et al., 2003; Ren and Zhao 2009; Yang and Yang 2016). Wen (2005) adopted 16 financial indicators from the aspects of products, Social impact and employees and constructed a CSR evaluation index system. Li & Li (2005) established a set of CSR indicator systems consisting of financial indicators from five aspects: labor rights, human rights protection, social responsibility management, business ethics and social welfare behavior.
Regarding the construction of CSR evaluation indicators, this paper adheres to the principles of operability, hierarchy, effectiveness and relative completeness. For each stakeholder dimension we select several financial indicators to represent the social responsibility of the company to the specific stakeholder. The specific evaluation indicators and calculation equations for each indicator are shown in Table 1.
Table 2 Corporate Social Responsibility Measurement Indicators Stakeholder
dimension
Indicators
Shareholder Dividend per share
Asset per share = Total shareholders' equity / number of common shares
Creditor
Asset current ratio = Current assets / current liabilities Cash ratio = Cash and cash equivalents / current liabilities Shareholder equity ratio = (1- Total liabilities / total assets)
Supplier Accounts payable turnover rate = Operating cost / accounts payable Cash accounts payable ratio = Cash / accounts payable
Consumer Operating cost ratio = Operating cost / operating income
Employee Welfare growth rate = Employee income growth rate / operating income growth rate
Income growth rate = (Total salary in the current year / Total salary in the previous year) / Total salary in the previous year
Government Tax rate = Full tax / total assets
The selection of specific indicators is based on previous literature which used financial indicators to measure CSR (Dong and Chen 2014; Jiang and Zhu 2011; Mao et al., 2012; Liu 2012).
Shareholder CSR: Shareholders are the owners of the company. The fulfillment of corporate responsibility to shareholders is mainly reflected in improving the efficiency of the use of corporate assets, increasing corporate profits, increasing the value of shareholders' invested capital and creating maximum benefits for shareholders. Actively fulfilling the responsibilities to shareholders can enhance the trust of shareholders in the company, continue to strengthen investment, and further promote the further development of the company. The indicators selected for measuring corporate social responsibility to shareholders in this paper include dividend per share and assets per share.
Creditor CSR: Creditors lend capital to the enterprise, expect to obtain a certain interest income, and recover the principal when the loan expires. The reputation and credit of companies are extremely important for creditors decision making, which is directly influences the company’s easiness of obtaining credit in the future. Only by paying interest in time and repaying debts on time can companies maintain a good image and reputation and long-term cooperation with creditors, thereby ensuring the stability of financing channels and financing costs and increasing corporate profits. Indicators such as the shareholders' equity ratio, asset current ratios and cash ratios are used for this dimension.
Consumer CSR: Selling products to consumers is an important part of the company's profit. This article selects operating cost ratio to reflect the car company's investment in product quality which in turn can serve as a proxy for consumer relations.
Supplier CSR: The supplier's supply chain is an important part of the company's production. It has a very important role in the company's production and operation. The supplier's main concern is that the company can comply with the supply contract and repay on time. Only by fulfilling the social responsibility towards suppliers can we establish a good long-term cooperative relationship with them and achieve mutual benefits and win-win results. This paper selects the accounts payable turnover rate and cash accounts payable ratio to measure the supplier dimension of CSR.
enterprises, and ensure the long-term stable development of enterprises. Otherwise, the satisfaction and enthusiasm of employees will decline, which will affect the product quality and operational efficiency of the company. This paper selects income growth rate and welfare growth rate to measure the dimension of employee relations.
Government CSR: government provides public resources and guarantees for enterprises. Enterprises should also fulfill their responsibilities in response to the government. The responsibility of enterprises to the government is mainly manifested in lawful operation, taxation in accordance with the law, creating a favorable market atmosphere, and appropriately and moderately relieving employment pressure for the country. Only by fulfilling government CSR can we establish a good long-term cooperative relationship with the government and obtain a favorable external development environment to enhance the efficiency of the enterprise. This article uses the asset tax rate and taxes rate as the indicators for the company's CSP in the government dimension.
For all indicators above, higher values represent better CSR performance in the corresponding dimension.
Since we used multiple indicators for most dimensions, factor analysis is used to extract a common factor from the indicators under the same dimension. This factor is supposed to represent the CSR of the corresponding dimension. The steps of factor analysis are: normalize the original data, establish the correlation coefficient matrix of the variable; calculate the eigenvalue and variance contribution rate of the correlation coefficient matrix; establish the factor load matrix, and implement the maximum rotation of the factor load matrix to make the factor variable more Interpretative; calculate factor scores. We used STATA to carry out the factor analysis. If not all the indicators can load on a single factor, then both the factor and the independent indicators will be used in the subsequent analysis as independent variables. 3.1.2 Measurement of corporate financial performance
have some drawbacks: accounting-based measures use historical data which is easily manipulated by managers; as for market-based measures, the market value of a firm is determined by stock market participants who make decision relying on their perceptions of future returns and risks which may not be an accurate measure of actual company performance either. The selection of measurement for CFP may also have an impact on the result of analysis since previous research found that CSP is more highly correlated with accounting-based measures than market-based measures (Orlitzky et al. 2003).
Moore (2001) pointed out that the data availability of accounting-based measures of CFP are better than the market-based measures, especially with ROA which reflects the profitability of the total assets of the enterprise. Higher ROA represents better asset utilization effect, and implies stronger activity ability of the whole enterprise and higher management ability, which will bring good business performance to the enterprise Since market based measures can also be influenced by other factors and measure not only the financial outcome of the company (Griffin and Mahon 1997); therefore, this paper uses ROA as the indicator to for a company's financial performance.
3.1.3 Design of control variable
Apart from CSR, CFP can also be correlated with other external factors, such as the industry, the size of the enterprise, the growth ability of the enterprise, investment in R&D and other characteristics of the enterprise (Mcwilliams and Siegel 2000). When studying the relationship between the two, it is necessary to control these factors in order to make the results more reliable. Most previous literature add company size as the control variable. Large-scale enterprises have sufficient capital and strength to better assume social responsibilities, and are more vulnerable to the public's attention, which in turn brings positive economic benefits to the enterprise. This suggests that the size of the firm has a positive impact on social responsibility (Mcwilliam & Siegel, 2000). The size of the firm is operationalized as the natural logarithm of the total assets of the sample company.
3.2 Data collection
4. Results
4.1 Descriptive statistics
Table 3. Descriptive statistics of variables
Dimension CSR indicators 2014 2015 2016 2017 2018 Std. Dev.
Shareholder
Dividend per share 17.12 17.23 18.27 18.80 21.16 27.64
Asset per share 1192.82 1189.20 1131.91 1152.85 1102.85 1859.1
Creditor
Asset current ratio 1.92 1.88 1.96 1.94 1.94 1.31
Cash ratio 0.62 0.61 0.63 0.61 0.65 0.88
Shareholder equity ratio 0.77 0.76 0.78 0.77 0.78 0.16
Supplier
Accounts payable turnover rate 7.69 7.79 7.84 7.49 7.69 5.01
Cash accounts payable ratio 3.45 1.66 5.37 2.40 1.74 6.34
Consumer Operating cost ratio 37.43 43.74 24.02 73.37 11.60 116.87
Employee
Welfare growth rate 4.08 13.73 4.85 -0.76 1.55 37.53
Income growth rate 0.22 0.18 0.14 0.22 0.39 0.34
Government
Tax rate 0.24 0.32 0.32 0.09 0.19 0.97
18
4.2 Factor analysis
Since we selected multiple indicators for each dimension of CSR, we conducted factor analysis to see whether a common factor can be extracted from the indicators under the same dimension. If so, the common factor score can be used as the proxy measure for the corresponding CSR dimension. All indicators are standardized before factor analysis. Only factors with eigenvalues greater than 1 are retained.
Table 4 Factors for each dimension Factor dimension Variable name Indicator Factor loading Shareholder CSR
CSRsha Dividend per share 0.9551
Asset per share 0.9551
Creditor CSR CSRcre Asset current ratio 0.9414
Cash ratio 0.9066
Shareholder equity ratio 0.6272
Supplier CSR CSRsup Accounts payable turnover rate 0.7710
Cash accounts payable ratio 0.7710
Employee CSR CSRemp Welfare growth rate 0.7315
Income growth rate -0.7337
Government CSR
CSRgov Tax rate 0.7128
Asset tax rate 0.7128
Consumer CSR CSRcon Operating cost ratio /
4.3 Regression analysis
4.3.1 Results of OLS regression
We first start with a cross-section regression model:
!"#!" = %#+ ∑ %$ $!()!"$+ ∑ %% %!*+,-*.!% + /!" (1) where !"#!" denotes corporate financial performance (ROA) of company i in year t, !()!"$denotes corporate social responsibility performance in dimension k of company
i in year t, !*+,-*.!% refers to the control variables (company size in our study) of company i in year t, and /!" is a time-varying error term. This model assumes that the cross-section sample of each year are independent random samples. The estimation of this model is called pooled ordinary least squares (OLS). Before we run the OLS regression, a multicollinearity test was first performed. It is generally believed that if the maximum VIF exceeds 10, there will be multiple collinearity that will seriously affect the least squares estimate. Table 5 shows the maximum VIF is 1.29 which means that there is no serious multicollinearity among the independent variables.
Table 5 Variance Inflation Factor
Variable VIF 1/VIF
20
Table 6 Results of OLS regression
VARIABLES China Developed
countries CSRsha 173.120*** -0.342** (35.282) (0.138) CSRcre 0.677*** 1.207*** (0.170) (0.185) CSRsup 1.055*** -0.080 (0.297) (0.123) CSRemp -1.054*** 0.031 (0.229) (0.123) CSRgov 1.418*** -0.105 (0.399) (0.108) CSRcon 0.000 -0.001 (0.000) (0.001) size -0.217 0.078 (0.178) (0.068) Constant 127.758*** 3.244*** (25.383) (1.226) Observations 402 601 R-squared 0.201 0.081
Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1
Since all CSR variables are standardized factor score, the estimated coefficients can be compared directly. We can see that the shareholder dimension of CSR has the biggest influence on CFP while the other four dimensions have similar level but much smaller impact on CFP. This seems to suggest that shareholder is indeed the most important stakeholder and the relations with shareholder is most closely related to CFP.
The sample of developed countries include companies from Germany, US and Japan. We put these countries together because they have similar level of economic development in contrast to China. The regression results based on the sample of developed countries is quite different from China as expected: only the dimension of shareholder and creditor CSR have significant correlation with CFP, and the impact of shareholder CSR is negative which is opposite to that in China.
4.3.2 Intermezzo: endogeneity
A problem of OLS estimation is that it assumes that the error term is uncorrelated with all independent variables. However, this assumption may not stand in our case, since there are company-specific unobserved factors which can affect both CSR and CFP. When these unobserved factors are not included in the regression, it leads to the problem of endogeneity and can bias the estimates of regression parameters. This type of bias is termed as omission variable bias.
22
overwhelming effort in variable selection and data collection. Therefore, we will only use fixed-effect model to deal with endogeneity and leave the instrumental variable method for future research.
4.3.3 The results of the fixed effects model
!"#!" = %! + 0! + ∑ %$ $!()!"$+ ∑ %% %!*+,-*.!"%+ /!" (2)
Compared to equation 1, we add a company specific intercept %! in equation 2 in order to control for the unobserved characteristics of each company.
Table 7 Results of model with company fixed effects
VARIABLES China Developed
countries CSRsha 28.625 -0.086 (43.718) (0.536) CSRcre -0.108 1.239*** (0.190) (0.414) CSRsup 2.955*** 0.060 (0.697) (0.216) CSRemp -0.768*** -0.010 (0.179) (0.110) CSRgov 0.898*** -0.061 (0.308) (0.103) CSRcon 0.000 -0.002 (0.000) (0.002) size -1.241** 0.042 (0.556) (1.195) Constant 46.032 3.797 (33.183) (21.673) Observations 402 601 R-squared 0.779 0.656
First, we look at the result for Chinese companies. After adding the ’company fixed effect the results are different from the OLS regression implying that omission variable bias does exist. First, the effects size of all dimensions of CSR on financial performance is less than that without fixed effects. Second, the shareholder and creditor dimensions of CSR lose significance. This is especially worth noticing since the results of OLS regression suggest that shareholder dimension has the biggest impact on CFP. Previous literature also has similar findings that the relationship between CSP and CFP becomes insignificant after controlling for fixed effects (Garcia-Castro et al. 2010). Third, the impact of employee CSR become positive. As for the developed countries, the results also changed compared to that of OLS regression: shareholder dimension also lost significance.
These results offer mixed support for our hypotheses. Hypothesis 1 is partly supported since some dimensions of CSR do have significant correlation with CFP regardless of countries. However, some dimensions have a negative impact on CFP and the specific dimensions which are significant depend on the location of companies. As for hypothesis 2, our results offer support since there is significant difference regarding CSR-CFP relationship between China and developed countries both in terms of which dimensions are significant, their direction of impact and the effect sizes.
5. Conclusion and managerial implications
This section will further discuss the conclusions of this paper and put forward relevant practical implications based on the research results. Finally, the research limitations of this paper will be discussed.
5.1 Conclusions
24
the size of effects depend on countries. In China, the CSR dimensions of government, employees and government have a significant positive correlation with CFP. This indicates that the relations with the government, employees and shareholders are closely related to the financial performance of the company.
The regression results also show that in the developed countries (Germany, US and Japan) only the creditor CSR is positively correlated to financial performance, which implies that the more an enterprise can pay interest on time and return the debts due, the higher the credibility of the enterprise, and the better the company can solve the problem of capital shortage, which helps the company to improve its performance. 5.2 Implications of results
First, to enhance the market's response mechanism to social responsibility: In order to enable companies to more actively implement CSR and publicly release relevant information, the market's response mechanism to corporate CSR shall be strengthened. Therefore, the following suggestions can be made from the perspectives of different stakeholders: first, strengthen market investors' awareness of CSR reporting and paying attention to CSR, CSR fulfillment of enterprises is an important factor for investors to make investment decisions. Second, promote media to publish relevant CSR information more fairly. The mainstream media should exercise its supervisory duties, such as reporting negative news related to CSR. It should also give a positive report to food companies that better perform CSR, so that consumers can rebuild their confidence in companies, broaden their understanding of CSR, and not only stay on the quality and safety responsibility of products.
capital and resources, resources should be allocated appropriately according to different types of stakeholders.
Third, strengthen the construction of corporate social responsibility legal system: With the improvement of the market environment, the government and the community are paying more and more attention to the implementation of CSR. However, if we only pressure the enterprise from soft constraints such as public opinion, enterprises will lack effective pressure in the process of fulfilling CSR. Therefore, the government should step up efforts to formulate corresponding laws and regulations to urge corporate social responsibility behavior. This is especially relevant in developing countries such as China which does not yet have many CSR regulations, and the existing regulations fail to fully consider all stakeholders and are not perfect in acting as a guide for company behavior. Based on this, the government should strengthen the legislative work of CSR in the automotive industry, build a sound CSR legal system, and at the same time increase law enforcement and impose mandatory CSR behavior.
Finally, establish a corporate governance model of mutual checks and balances: the long-term sustainable development of an enterprise is inseparable from the various stakeholder groups outside the enterprise. For the long-term stable development of the enterprise, it must meet the needs of various stakeholders and take the initiative. The traditional corporate governance model belongs to the “shareholder supremacy”, which is likely to cause myopic behavior and thus harm the interests of other stakeholders. 5.3 Limitations and future research suggestions
This study has the following limitations:
26
measurement indicators for CSR and investigate whether the choice of measurement approach lead to different empirical findings (Barnett and Salomen 2006).
2.The model we used is rather simple: we only tested linear effect between CSR and CFP. However, there were studies which found non-linear relationship between CSR and CFP (Barnett and Salomen 2006; Shan et al. 2019). A related research recommendation is to test other non-linear specifications in future studies.
3.We did not test the opposite direction of causality, while some studies detect that CFP can also affect CSR performances. This double direction of causality can also lead to endogeneity in cross-sectional regression. Some remedies for this problem include estimating simultaneous equations (Choi et al. 2010; Yin et al. 2014) and adopting a lagged structure to ensure the direction of causality flow (Apostolides and Boden 2017; Shahzad and Sharfman 2017; Yin et al. 2014)
28 References
Apostolides, N., Boden, R.: Corporate environmental responsibility and financial performance: Does bidirectional causality work? Empirical evidence from the manufacturing industry. Soc. Responsib. J. 13, 53–62 (2017)
Aras, G., Aybars, A., Kutlu, O.: Managing corporate performance: Investigating the relationship between corporate social responsibility and financial performance in emerging markets. Int. J. Product. Perform. Manag. 59, 229–254 (2010).Arsoy, A.P., Arabacı, Ö., Çiftçioğlu, A.: Corporate Social Responsibility and Financial Performance Relationship: The Case Of Turkey. J. Account. Financ. 159–177 (2012). doi:10.5923/j.ijfa.20170606.02
Balabanis, G., Phillips, H.C., Lyall, J.: Corporate social responsibility and economic performance in the top British companies : are they linked ? Eur. Bus. Rev. 98, 25–44 (1998)
Barnett, M., Salomen, R.: Beyond Dichotomy: The Curvilinear Relationship Between Social Responsibility And Financial Performance. Strateg. Manag. J. 27, 1101–1122 (2006).
Benabou, R., Tirole, J.: Individual and corporate social responsibility. Economica. 77, 1–19 (2010).
Brammer S, Brooks C, Pavelin S Corporate Social Performance and Stock Returns: UK Evidence from Disaggregate Measures. Financial Management, 2006, 35(3): 97-116 Chatterji, A.K., Levine, D.I. & Toffel, M.W., 2009. How well do social ratings actually measure corporate social responsibility? Journal of Economics and Management Strategy, 18, 125- 169.
Chen, L., Feldmann, A., Tang, O.: The relationship between disclosures of corporate social performance and financial performance: Evidences from GRI reports in manufacturing industry. Int. J. Prod. Econ. 170, 445–456 (2015).
Choi, J.-S., Kwak, Y.-M., Choe, C.: Corporate Social Responsibility and Corporate Financial Performance: Evidence from Korea Corporate Social Responsibility and Corporate Financial Performance: Evidence from Korea. (2010)
Comell Bradford, and Shapiro, Alan C. Corporate Stakeholders and Corporate Finance. Financial Management, 1988, 16(1):5-14
Siegel (Eds.), The Oxford handbook of corporate social responsibility (pp. 3–18). New York, NY: Oxford University Press.
Dong,S.,Chen,M. Research on the Impact of Corporate Social Responsibility on Financial Sustainable Growth Based on GEM, Corporate Data. Friends of Accounting, 2014, 20:7-12
El Ghoul, S., Guedhami, O., Kim, H. & Park, K., 2018. Corporate environmental responsibility and the cost of capital: International evidence. Journal of Business Ethics, 149, 335-361.
Fernando, C.S., Sharfman, M.P. & Uysal, V.B., 2017. Corporate environmental policy and shareholder value: Following the smart money. Journal of Financial and Quantitative Analysis, 52, 2023-2051.
Ferrell, A., Liang, H. & Renneboog, L., 2016. Socially responsible firms. Journal of Financial Economics, 122, 585-606.
Frigant, V.: CSR in the auto industry: do the first-tier suppliers have stakeholders? Int. J. Automot. Technol. Manag. 9, 377 (2009).
Galant, A., Cadez, S.: Corporate social responsibility and financial performance relationship : a review of measurement approaches. Econ. Res. Istraživanja. 30, 676– 693 (2017).
Garcia-Castro, R., Ariño, M.A., Canela, M.A.: Does social performance really lead to financial performance? Accounting for endogeneity. J. Bus. Ethics. 92, 107–126 (2010). Godfrey, P.C., Hatch, N.W.: Researching corporate social responsibility: An agenda for the 21st century. J. Bus. Ethics. 70, 87–98 (2007). doi:10.1007/s10551-006-9080-y Griffin, J.J., Mahon, J.F.: The corporate social performance and corporate financial performance debate: Twenty-five years of incomparable research. Bus. Soc. 36, 5–31 (1997).
Heidarzadeh Hanzaee, K., & Sadeghian, M. (2014). The impact of corporate social responsibility on customer satisfaction and corporate reputation in automotive industry. Evidence from Iran. Journal of Islamic Marketing, 5(1), 125-143.
Heinkel, R., Kraus, A. & Zechner, J., 2001. The effect of green investment on corporate behavior, Journal of Financial and Quantitative Analysis, 36, 431-449.
H.Russell Fogler, Fred Nutt: A Note on Social Responsibility and Stock Valuation.The Academy of Management Journal. 18, 155-160
30
Inoue, Y., Lee, S.: Effects of different dimensions of corporate social responsibility on corporate financial performance in tourism-related industries. Tour. Manag. 32, 790– 804 (2011).
Jiang, R., Zhu, J., Construction of Evaluation Index System of Social Responsibility of Listed Companies. Accountant Friends, 2011, 19:61-63
Kim, K.H., Kim, M.C., Qian, C.: Effects of Corporate Social Responsibility on Corporate Financial Performance: A Competitive-Action Perspective. J. Manage. 44, 1097–1118 (2018).
Liang, H., Renneboog, L.: On the Foundations of Corporate Social Responsibility. J. Finance. 72, 853–910 (2017).
Lins, K. V., Servaes, H., Tamayo, A.: Social Capital, Trust, and Firm Performance: The Value of Corporate Social Responsibility during the Financial Crisis. J. Finance. 72, 1785–1824 (2017).
Liu,W.,Relationship between social responsibility and financial performance of food processing manufacturing enterprises[D]. Dongbei University of Finance and Economics, 2012
Loureiro, S.M.C., Dias, I.M., Reijnders, L.: The effect of corporate social responsibility on consumer satisfaction and perceived value : the case of the automobile industry sector in Portugal. J. Clean. Prod. 37, 172–178 (2012).
Makni R, Francoeur C, Bellavance F Causality Between Corporate Social Performance and Financial Performance: Evidence from Canadian Firms. Journal of Business Ethics, 2009, 89(3): 409-422
Mao, L.,Wang, Z.,Wang,L.,Institutional Investors' Shareholding Preference, Screening Strategy and Corporate Social Performance[J].Management Science,2012,03:21-33 Margolis, J. and Walsh, J. (2003) Misery Loves Companies: Rethinking Social Initiatives by Business. Administrative Science Quarterly, 48, 268-305.
McGuire, J B, Sundgren, A and Schneeweis, T. Corporate Social Responsibility and Firm Financial Performance. Academy of Management Joumal, 1988, 31(4): 854-872 Mcwilliams, A., Siegel, D.: Corporate Social Responsibility And Financial Performance : Correlation Or Misspecification ? Strateg. Manag. J. 21, 603–609 (2000) Miles Morgan, P., Covin Jeffrey, G.: Corporate Social and Financial Performance: An Investigation in the U.K. Supermarket Industry. J. Bus. Ethics. 299–315 (2001)
Lanka. Journal of Emerging Trends in Economics and Management Sciences, 2013, 4(3): 371-376
L., Zhao, J.: An Empirical Study of Relationship Between Corporate Social Responsibility and Financial Performance. J. Chongqing Jiaotong Univ. 9, 60–65 (2009)Nelling, E., Webb, E.: Corporate social responsibility and financial performance: The “virtuous circle” revisited. Rev. Quant. Financ. Account. 32, 197–209 (2009). Orlitzky, M.: Corporate Social Performance and Financial Performance: A Research Synthesis. Oxford Handb. Corp. Soc. Responsib. 1–26 (2008).
Orlitzky, M., Schmidt, F.L., Rynes, S.L.: Corporate social and financial performance: A meta-analysis. Organ. Stud. 24, 403–441 (2003).
Perujo, A., Ciuffo, B. (2010) The introduction of electric vehicles in the private fleet: Potential impact on the electric supply system and on the environment. A case study for the Province of Milan, Italy. Energy Policy, Vol. 38, pp. 4549-4561.
Peter R, Mullen M. R. Some Evidence of the Cumulative Effects of Corporate Social Responsibility on Financial Performance. Journal of Global Business Issues, 2009, (3): 1-14
Ren, Sarkis, J., Gonzalez-Torre, P., Adenso-Diaz, B. (2010) Stakeholder pressure and the adoption of environmental practices: The mediating effect of training. Journal of Operations Management, Vol. 28, No. 2, pp. 163-176.
Ruf, B.M., Muralidhar, K., Brown, R., Janney, J., Paul, K.: An Empirical Investigation of the Relationship Between Change in Corporate Social Performance and Financial Performance: A Stakeholder Theory Perspective. J. Bus. ethics. 32, 143–156 (2001) Saeidi, S.P., Sofian, S., Saeidi, P., Saeidi, S.P., Saaeidi, S.A.: How does corporate social responsibility contribute to firm financial performance? The mediating role of competitive advantage, reputation, and customer satisfaction. J. Bus. Res. 68, 341–350 (2015).
Shahzad, A.M., Sharfman, M.P.: Corporate Social Performance and Financial Performance: Sample-Selection Issues. Bus. Soc. 56, 889–918 (2017).
Shan, M., Lin, Y., Cheng, F.: Nonlinear relationship between corporate social responsibility and financial performance. Friends of Account. 11, 45-51 (2019)
Waddock S, Graves S (1997) The corporate social performance-financial performance link. Strategic Manage J, 18:303–319
32
Wu, M.W., Shen, C.H.: Corporate social responsibility in the banking industry: Motives and financial performance. J. Bank. Financ. 37, 3529–3547 (2013).
Xie, X., Jia, Y., Meng, X., Li, C.: Corporate social responsibility, customer satisfaction, and financial performance: The moderating effect of the institutional environment in two transition economies. J. Clean. Prod. 150, 26–39 (2017).
Yang, W., Yang, S.: An Empirical Study on the Relationship between Corporate Social Responsibility and FInancial Performance under the Chinese Context. Chinese J. Manag. Sci. 24, 143–150 (2016)Yin, K., Liu, X., Chen, H.: Study on the Relationship between Corporate Social Responsibility and Financial Performance from the Endogenous Perspective: Evidence from Chinese Listed Companies. China Soft Sci. Mag. 6, 98–108 (2014)
