The above section 2.1.5 discusses in detail about CSR and media reporting, and in this section, the study exhibits the relationship between CSR and firm performance. This section is comprised of four parts, including the measurement of CSR, CSP and firm performance, Direction of Causality Between CSP and CFP, and Moderators of CSP-CFP link. The measurement of CSR explains why this study uses CSP rather than CSRD to measure CSR. Then, the review of CSP-CFP relationship helps establish an understanding why the positive linkage exists. Then, the review of the direction of causality explains why the results of prior literature are mixed, and how to extend their work. Finally, the review of moderators between CSP-CFP presents why media coverage is a moderating variable.
According to Ullmann (1985), CSR is concerned with both CSR performance (CSP) and CSR disclosure (CSRD). Many prior studies fail to differentiate CSP and CSR disclosure. Therefore, it is important to point out the difference between CSP and CSR disclosure. On the one hand, it can be argued that firms doing “good” should be willing to provide more “green” information (Clarkson et al., 2008; Herbohn, 2014). On the other hand, it can also be argued that poor environmental or social performers may tend to provide more information for gaining legitimacy (Patten, 2002a; Cho and Patten, 2007; Cormier, 2011). Under this circumstance, the firm’s CSR disclosure will not match its actual CSR performance accurately. Because the information asymmetry exists between the firm and stakeholders, the stakeholders may not fully be received the true picture of firm’s CSR activities. After considering the potential manipulation that may exist in firm’s CSR disclosure, this study, therefore, selects CSP as the proxy of CSR. The below two subsections review measurements of CSP and CSRD respectively.
Measurement of CSP
Corporate social performance (CSP) is defined as a business organization’s configuration of principles of social responsibility, processes of social responsiveness, and observable outcomes as they relate to the firm’s societal relationships (Wood, 1991). In this study, CSP refers to firms’ actual CSR performance such as carbon emission, employee CSR training, charitable donation, and community contribution (Al-Tuwaijri et al., 2004; Clarkson et al., 2008; Luo et al., 2015; Bouslah, 2016). Researchers measure the CSP in diverse ways including the use of reputational measure, spending measures, unidimensional, and multidimensional ratings based on some observable social responsibility indicators. Each of these measures has unique strengths and weaknesses. Orlitzky et al. (2003) argue that the diversity of CSR measures largely contributes to the contradictory findings on the nature of the CSR-CFP relationship.
It is possible to calculate some scores on the goodwill associated with the reputation of a firm and use these scores as measures of the CSR for research purposes. Fortune corporate reputation index provides such a calculation based on the reputation perceived by their respondents and publishes company reputation ratings, such as America’s Most Admired Companies (AMAC). As a CSR strategy, defending reputation helps the business corporations to develop legitimacy and gain competitive advantage (Surroca et al., 2010; Sun & Cui, 2014). The challenge with the reputation-based measures is that the respondents’ perception and the resultant corporate reputational ratings are distorted by the firm’s prior financial record (Soana, 2011). Therefore, reputation based metrics are not likely to adequately measure CSR.
CSP could be measured by the level of expenditures such as the voluntary donations and the charitable contributions made by the firm toward improving the social and environmental wellbeing of the stakeholders (Soana, 2011). The voluntary social spending such as donations, advertising expenditures, and training costs may help to bolster the firm’s image, reduce the social pressure against the firm, and ultimately improve the firm’s competitive performance leading to greater profits and stockholders’ wealth (Weshah et al. 2012; Servaes and Tamayo, 2013). Nevertheless, the motive for the expenditure and the prevalence of information asymmetry make the efficacy of the social spending by business managers to be shrouded in uncertainty which increases agency cost (Sun & Cui, 2014).
The unidimensional indicators are focused on only a single aspect of social responsibility practices such as environmental or philanthropic practices in the local communities. However, the unidimensional CSR measures are criticized by a lack of comprehensiveness. To overcome the narrow focus of unidimensional CSR measures, some researchers combine the measure with other measures. For example, Busch & Hoffmann (2011) measure the CSR as carbon intensity which they relate to the firm sales but combine this unidimensional measure with the questionnaire surveys and the sustainability rating index.
Multidimensional ratings (KLD Database and Thomson Reuters Asset 4)
Choosing which data source to use is largely dependent on the research question posed. A one-dimensional indicator has the benefit of relative simplicity and can be appropriate where a single or narrower aspect of CSR is being examined. Nevertheless, data from multidimensional rating agencies has been most frequently used. The main reason is that CSP itself is multidimensional (as indicated by Carroll’s CSR framework) and is comprised of both internal (corporate governance issues) and external (environmental and social impact) factors, which should be considered when evaluating CSP. Based on above discussion, this study focuses on overall social responsibility performance.
Reviewing the literature, most recent researchers use KLD Database (Servaes and Tamayo, 2013; Cahan et al., 2015; Rhou et al., 2016). The KLD database is compiled by an independent rating service which focuses on a wide range of firms over a broad spectrum of CSR screens. This database rates company on 13 dimensions of CSR including community, corporate governance, diversity, employee relations, environment, human rights, product quality and safety, alcohol, firearms, gambling, military, nuclear power, and tobacco. Each dimension in the KLD database is summarized as strengths (positive values) and concerns (negative values). A firm is given a score of 0 or 1 across each strength or concern. KLD based CSP index is widely accepted by practitioners and academics as an objective measurement. According to Callan and Thomas (2009), two of the more prominent aggregate measures used in academic studies are the Fortune rating data and the indices provided by KLD, with more recent studies gravitating toward the use of KLD data. Waddock and Graves (1997), for example, believe that the KLD indicators are superior to the Fortune data because the latter are more about a firm’s overall management than its socially responsible decisions. Furthermore, Chand (2006) asserts that a KLD-based index offers more objectivity than a measure based on Fortune’s survey data.
However, KLD database is not without limitation. As argued by Humphrey et al. (2012), it is not appropriate to calculate an overall CSP score as the total number of strengths minus the total number of concerns. If a firm is engaged in business practices that involve hazardous waste, this activity will be scored 1 under concern. This will lower the firm’s overall CSP score. If another firm is not engaged in practices that involve hazardous waste, it will receive 0, and thus hazardous waste will not contribute to the firm’s overall score. Furthermore, Humphrey et al. (2012) mention that there are several issues of KLD’s rating systems. Take hazardous waste production as an example. First, the binary ratings do not distinguish between the levels of hazardous waste production. Second, firms in heavy polluting industries like Oil and Gas have lower KLD score than other firms that have very limited or no disclosure to producing hazardous waste, regardless how well the firm manages its hazardous waste. Third, the number of measures within each of KLD’s dimensions can skew overall CSP scores. For example, in environmental dimension, a firm has a total of 5 strengths and 10 concerns. By definition, the KLD rating system is biased toward a higher concern score for those industries disclose information about their environmental concerns.
Notably, this study uses Thomson Reuters Asset 4 ESG score as a proxy of CSP. Firstly, Thomson Reuters Asset 4 ESG is available in Durham Library (KLD is criticized for not free). Secondly, the Thomson Reuters Asset 4 ESG database is frequently used by academic researchers and professional investors. The proof of this claim is the variety of researchers who have used this database (e.g. Lam et al., 2012; Cheng et al., 2012; Wimmer, 2012; Eccles et al., 2011; Qiu, Shaukat, and Tharyan, 2016). Finally, the ratings from Thomson Reuters Asset 4 are superior to KLD ratings in two aspects: (1) ratings of Asset 4 are Z-scored (normalized), hence there is no skewness issue; (2) ratings of Asset 4 are continuous (between 0 and 100%), which is relatively comparable than the categorical number of KLD (e.g. -2 as more concern, -1 as concern, 0 as neither concern nor strength, 1 as strength, 2 as more strength). The study discussed the details of Thomson Reuters Asset 4 below.
Thomson Reuters Asset 4 database provides environmental, social and governance performance information. As said in its Handbook (Thomson Reuters Asset 4, 2016), “…provides objective, relevant and systematic ESG performance information. A team of over 130 experienced analysts collects 700 ESG evaluation points (see Appendix 2) per firm, which is used to calculate 250 key performance indicators. Each data point goes through a multi-step verification process, including a series of data entry checks, automated quality rules, and historical comparisons. Data sources include stock exchange filings, CSR and annual reports, non-governmental organization websites and various news sources. Primary data used are objective and publicly available.”
There are 18 categories of key performance indicators within four pillars (see Figure 1), namely Economic (3 categories), environmental (3 categories), social (7 categories) and governance (5 categories). Only environmental, social and governance pillars are relevant for the analysis in this thesis. The environmental score as defined by Asset4 “measures a company’s impact on living and non-living natural systems, including the air, land, and water, as well as complete ecosystems. It reflects how well a company uses best management practices to avoid environmental risks and capitalize on environmental opportunities”. It covers “hard” performance indicators (as classified by Clarkson et al., 2008) such as information on energy used, CO2 emissions, water and waste recycled, and spills and pollution controversies. Hence, the aggregate environmental score measures a firm’s environmental performance. The social score as defined by Asset4 “measures a company’s capacity to generate trust and loyalty with its workforce, customers, and society, through its use of best management practices.” It covers issues like employee turnover, accidents, training hours, donations, and health and safety controversies. The social score also includes mostly “hard” performance indicators and hence is an objective measure of the social performance of a firm. The Asset4 ESG data is available from 2004 onwards. Firms with no Asset4 ratings are excluded from the study.
Figure 1 Thomson Reuters Asset 4 ESG performance indicators, categories, pillars and overall performance score
(Adapted from Thomson Reuters Asset 4 handbook 2016)
The Thomson Reuters Asset 4 ESG provides a more comprehensive calculation of the rating scores. Key performance indicators, categories, pillars and overall score are compared by using equally weighted computer calculations of relative company performance, the benchmark being the Asset4 company universe. These ratings are Z-scored and normalized to position the score between 0 and 100%. The Z-score is a relative measure comparing one company with a given benchmark. It expresses the value in units of standard deviation of that value from the mean value of all companies.
Measurement of CSRD
CSR disclosure measures the extent a firm reveals its CSR information to the public. Many researchers have conducted surveys to measure CSRD (Webb et al., 2009). However, in several cases, surveys have posed difficulties due to low return rates and inconsistency among survey participants. Worse still, survey questionnaires can only represent the views and perceptions of the respondents, which could be biased (Lange & Washburn, 2012). Researchers have also attempted to use experiments or case studies to measure CSRD (O’Dwyer, 2011). These methods, however, often suffer from a lack of generalizability and are influenced by participant biases. Another way to measure CSRD is through a content analysis of a firm’s sustainability report, annual report, and website (Webb et al., 2009). The content analysis involves counting of words, phrases, clauses or sentences in the publications that relate to the social or environmental themes and using binary values to quantify them (Ganescu, 2012). Presupposing that the social disclosure in the corporate publications is a good proxy for the CSR, many studies adopted this measure to examine the CSR-CFP link. For example, Uadiale and Fagbemi (2012), and Usman and Amran, (2015) use content analysis of corporate disclosures to examine the CSR-CFP relation. It has been argued that no research has attested to the validity of content analysis of published corporate information (Soana, 2011). For this reason, the content analysis methodology lacks theoretical base and offers only limited practical value.
This study reviews the measurement of CSR in two ways including CSP and CSRD. Regarding CSP, the measurement methods such as reputational measure, spending measures, unidimensional, and multidimensional ratings are used by prior researchers. In terms CSRD, the methods such as survey questionnaires, experiments, and the content analysis of firm’s reports (e.g. sustainability report, annual report, and website) are adopted. Each of these measures has unique strengths and weaknesses. Notably, it is argued by a lot of meta-analysis researchers (e.g. Orlitzky et al., 2003), the diversity of CSR measures largely contributes to the contradictory findings on the nature of the CSR-CFP relationship. The prevalent diversity of measures is exacerbated by the multiplicity of possible approaches within each measure. For example, for some studies in which multidimensional social responsibility ratings are used, researchers adopt KLD ratings (e.g. Servaes and Tamayo, 2013; Cahan et al., 2015; Rhou et al., 2016), and Thomson Reuters Asset 4 ESG (e.g. Lam et al., 2012; Cheng et al., 2012; Wimmer, 2012; Eccles et al., 2011; Qiu, Shaukat and Tharyan, 2016). Each of these rating bodies determines its index based on some survey questionnaires as well as measures on several qualitative factors. Finally, it is also worth mentioning that the CSR is measured by CSP based on the ratings of Thomson Reuters Asset 4 ESG, which is superior in both comprehensiveness and comparability.
Corporate financial performance (CFP), unlike the CSR, presents little challenge in research in both conceptual and measurement terms. Stewardship model of business requires every business firm to make a profit and to increase the firm value (Friedman, 1970; Jensen, 2010). Empirical researchers are unanimous in viewing profit or value creation from two perspectives: accounting perspectives and market perspectives, each of these presents its unique challenges (Goss & Roberts, 2011; Wu & Shen, 2013). Although the number of the CSR studies is still inadequate, studies that are based on the accounting measures of performance have more proportionately received due attention, while the market-based performance remains underexamined (Becchetti et al. 2012; Ghoul et al. 2011).
Accounting returns are the primary method of measuring financial performance. The basic idea behind using accounting returns as a measure of financial performance is to focus on how firm earnings respond to different managerial policies. The accounting measures of financial performance proliferate in the recent literature, and they include the earnings per share (Becchetti et al. 2013; Ghoul et al. 2011), the EPS growth and return on equity/assets/sales (Becchetti et al. 2013; Chen & Wang, 2011; Servaes & Tamayo, 2013; Wu & Shen, 2013), and the asset growth (Wu & Shen). The other related measures in the recent literature include the turnover and turnover growth (Arnold & Valentine, 2013), the net-/non-interest income on non-performing loan (Wu & Shen, 2013), the loan contract terms (Goss & Roberts, 2011), the brand equity (Torres et al. 2012), the cost-to-income ratio (Soana, 2011), and the absolute forecast error on EPS (Becchetti et al. 2013). Reviewing the literature, return on equity/assets/sales are most frequently used accounting ratios to measure CFP. For example, Callan and Thomas (2009) use KLD data to examine the relationship between corporate social performance and its financial performance including four different financial performance measures, namely ROA, ROS, ROE and Tobin’s Q. When financial performance is the dependent variable, they find a positive link between financial performance (measured as ROA, ROS or Tobin’s Q) and corporate social performance.
On the one hand, it can be argued that accounting numbers follow strict sets of accounting rules, are validated by the independent external auditors, and are contained in the published financial statements, they are expected to be of high quality and subject to minimal manipulation (Jiao, 2010). On the other hand, because accounting indices are backward looking and are based on convention and corporate choice, they can be biased, incomparable, and open to manipulation (Gregory et al. 2014). This is typified by the spate of corporate scandals that have been recorded in recent history which were characterized by manipulation by the corporate managers often in tacit collusion with their auditors.
Market-based performance measures focus less on accounting numbers; hence they are less susceptible to managerial subjectivity, manipulation or opportunism (Hajiha & Sarfaraz, 2013). Market-based measurement can be classified into two broad types: (1) measures based on share value, and (2) measures based on the cost of capital. Each of these measures can be separately related to the CSR measures for testing possible association.
Share values have been used extensively to measure the firm financial performance (Tafti et al. 2012). The most commonly adopted measures are based on share prices, such as (1) stock price (Baird et al. 2012), (2) price-earnings ratio (Soana, 2011), (3) MTB ratio (Deng et al., 2013; Soana, 2011), and (4) Tobin’s q (Jo & Harjoto, 2011; Servaes & Tamayo, 2013; Cahan et al., 2015; Rhou et al., 2016). Notably, the use of Tobin’s q is more prominent among researchers. Market returns have also been used by researchers to measure of financial performance in the current literature. This involves the calculation of excess or abnormal market returns (Becchetti et al. 2012; Deng et al. 2013). Only very few scholars used the volatility of market returns (Deng et al., 2013) or adopted asset pricing approach in their empirical models (Becchetti et al. 2012). Further research can be conducted by analyzing the link between CSR and volatility or asset pricing approaches to enrich this branch of literature.
Cost of capital
According to finance theory, there are two ways to be profitable: maximizing returns and minimizing financing cost. There is inadequate research into the relation between CSR and the financing cost dimension of corporate financial performance (Goss & Roberts, 2011). Campbell et al. (2012) observed that the relationship between the weighted average cost of capital (WACC) and internal financial resources is positive and significant. For CSR research purposes, the cost of capital can be operationalized as the cost of debt, the cost of equity, and WACC. Reviewing the current literature, the studies of Cajias et al. (2014), Ghoul et al. (2011), Gregory et al. (2014), and Hajiha & Sarfaraz (2013) fall into this area. Ghoul et al. (2011) examined the relationship between the CSR composite scores and the cost of equity; they found a positive interaction between the two variables. In their research, Campbell et al. (2012) examined the cost of debt, the cost of equity and the WACC separately against their non-CSR related independent variables.
As discussed by Gregory et al. (2014), market-based measures are determined by the external and independent evaluation of the firm performance, reflecting the investors’ perception and expectations of the future performance of the firm. Market-based measures are also not without limitations. Because market-based measures utilize capital market parameters such as security prices to evaluate firm financial performance, their focus is only on the financial stakeholders, while non-financial stakeholders who are also affected by corporate social conduct and misconduct are ignored (McWilliams et al. 2006). Worse still, according to the efficient markets hypothesis, when information that might affect future cash flows of a firm becomes available, it immediately will be reflected in its current share price. The implication of this is that even if CSR does lead to improved financial performance, as soon as the market becomes aware of any change in a firm’s CSR rating, it will immediately alter price per share to reflect that information. Alexander and Buchholz (1978) claim that only new information regarding a firm’s social responsibility will have any effect on the firm’s financial performance. Thus, if the perception of a firm’s social responsibility changed in 1975 and a naive researcher examined only the period 1977-1979, then he or she probably would conclude that CSR and financial performance are unrelated. To employ investor returns measures of financial performance properly, the researcher must conduct an event study (e.g. Wang et al., 2011; Sabbaghi et al., 2013). Failure to do so could lead the researcher incorrectly to the conclusion that there is no relationship between CSR and financial performance, even if one exists.
This study reviews the measurements of CFP and classifies them into accounting-based measurements and market-based measurements (including share-value approach and cost of capital approach). Regarding accounting-based measurements, ROA, ROE, and ROS are most popular measures of CFP. On the one hand, accounting ratios are expected to be of high quality and subject to minimal manipulation. On the other hand, they can also be biased, incomparable, and open to manipulation because of corporate choice. Regarding market-based measurements, Tobin’s Q is the prominent measures of CFP. Market-based measures reflect the investors’ perception and expectations of the future performance of the firm. However, non-financial stakeholders who are also affected by corporate social conduct and misconduct are ignored. Worse still, based on efficient markets hypothesis, if the market becomes aware of any change in a firm’s CSR rating, it will immediately alter price per share to reflect that information. Therefore, only new information regarding a firm’s social responsibility will have any effect on the firm’s financial performance. Finally, it is claimed by some meta-analysis researchers (e.g. Orlitzky et al., 2003; Peloza, 2009), that the diversity of CFP measures contributes to the contradictory findings on the nature of the CSR-CFP relationship. In this study, both accounting-based (e.g. ROA, ROE, ROS) and market-based (e.g. Tobin’s Q) measures are used. Specifically, ROA is the proxy of CFP in this study, and the other three measures (ROE, ROS, and Tobin’s Q) are used in robustness tests.
Those who have suggested a negative linkage between social responsibility and financial performance have argued that high responsibility results in additional costs that put a firm at an economic disadvantage compared to other, less socially responsible firms (Aupperle et al., 1985; Vance, 1975). Another dominant critique of business social involvement originates from the principal–agent paradigm, which suggests that the purpose of the firm is primarily for the profit of the stockholders. As said in Friedman’s book (Capitalism and Freedom, 1970), there is one and only one social responsibility of business – to use its resources and engage in activities designed to increase its profits so long as it stays within the rules of the game. Moreover, Friedman (1970) supposed that in responding to calls for socially responsible practices, executives take money and resources that otherwise go to stockholders and dedicate those resources to enhance their personal benefits. According to agency theory, the agents (managers) and the principals (the shareholders) always show a conflict of interests and objectives. Therefore, managers may act in their best interests, but at the expense of the firm’s owners. As a result, the costs involved in agency relationships may be high and damage corporate value.
Though some studies have argued that there is a negative association between CSR and financial performance (Friedman, 1970; Griffin & Mahon, 1997; Waddock & Graves, 1997; Harrison & Freeman, 1999; McWilliams & Siegel, 2000), the majority of prior research demonstrates that CSR and firm performance are positively associated (Porter & Kramer, 2002; Saiia et al., 2003; Brammer & Millington, 2005; Godfrey, 2005; Orlitzky et al., 2003; Roman et al., 1999). Waddock & Graves (1997), based on stakeholder analysis, propose that a tension exists between the firm’s explicit costs (e.g., payments to bondholders) and its implicit costs to other stakeholders (e.g., product quality costs, environmental costs). From the stakeholder theory, they predict that a firm that attempts to lower its implicit costs by socially irresponsible actions will, as a result, incur higher explicit costs, resulting in a competitive disadvantage. A compatible view is that the actual costs of CSP are minimal and the benefits potentially great. For example, an enlightened employee relations policy may have a very low cost but can result in substantial gains in morale and productivity, actually yielding a competitive advantage in comparison to less responsible firms. Firms that are reported in lists of “best companies to work for,” for instance, may find it easier to recruit top quality employees, possibly resulting in increases in productivity at relatively low cost (e.g., Greening & Turban, 2000). Because the above arguments are fundamentally normative, it is necessary to emphasize on the instrumental perspective, especially in a business case.
In this line of thinking, scholars rely on instrumental stakeholder theory and suggest that firms view their stakeholders as part of an environment that should be managed to assure revenues, profits, and, ultimately, returns to shareholders (Berman, Wicks, Kotha, & Jones, 1999). Instrumental stakeholder theory is an instrumental view of stakeholder theory. The theory suggests that firm’s business and strategy that are managed in the name of all stakeholders tend to maximize profits for shareholders. As Berman, Wicks, Kotha, & Jones (1999) said, firm ought to pay attention to stakeholders, and such attention to stakeholder concerns may help a firm avoid decisions that might prompt stakeholders to undercut or thwart its objectives. It has been well documented that employees show greater commitment to a firm that has a good public image in supplying human capital (Dutton, Dukerich, & Harquail, 1994). Moreover, such firms are often perceived as an attractive employer by job seekers (Backhaus, Stone, & Heiner, 2002; Greening & Turban, 2000). Customers may respond to the positive social performance by increasing their demand for the firm’s products or services, or by paying premium prices (Bhattacharya & Sen, 2003). Furthermore, some investors, particularly certain institutional ones, are more willing to invest in firms known for pursuing CSR (Barnett & Salomon, 2006; Graves & Waddock, 1994; Johnson & Greening, 1999).
Others have augmented stakeholder theory with aspects of RBV (Salancik & Pfeffer, 1978). Drawing upon the RBV, researchers claim that by developing close relationships with primary stakeholders, a firm can develop certain intangible resources such as innovation (Klassen & Whybark, 1999), human resources (Russo & Harrison, 2005), and organizational culture (Howard-Grenville, Hoffman, & Wirtenberg, 2003), which enable the most efficient and competitive use of the firm’s assets and help it acquire a competitive advantage over its rivals (Surroca et al., 2010). CSR can function as a means for firms to secure the acquisition of critical resources controlled by stakeholders and help firms to reduce the risk of losing resources they already control (Barnett & Salomon, 2006; Brammer & Millington, 2004; Godfrey, 2005). So far, the discussion suggests that the relationship between CSR and financial performance is positive from instrumental stakeholder theory and RBV.
Three views on the direction of causality between CSR and CFP have been tested empirically: (a) the view that CSP positively influences CFP, (b) the view that CFP positively influences CSR, and (c) the view defining a recursive relationship between both constructs (Surroca et al., 2010). The first research stream, related to instrumental stakeholder theory or RBV mentioned earlier, suggests that prior CSP positively influences subsequent CFP. The second strand of literature is supported by slack resources theory, which proposes that better financial profitability potentially results in the availability of slack resources that may increase a firm’s ability to invest in socially responsible domains such as community and society, employee relations or environment (Waddock and Graves, 1997). In contrast, firms that are in financial trouble may have little freedom to invest in CSR activities such as philanthropy. Some of the empirical evidence, particularly McGuire, Schneeweis, and Branch (1990) and Godfrey et al. (2009), provides support for the slack resources theory.
These two previous streams of research were reconciled by Surroca et al. (2010), who argued that causation may run in both directions. That is, favorable environmental and social performances lead to a surplus of available funds (instrumental stakeholder theory) which is reallocated, in part, to the different stakeholders (slack resources theory). There may then be a simultaneous and interactive positive relation between CSP and CFP, forming a virtuous circle (Waddock and Graves, 1997; Lo et al., 2008; Qiu, 2016). Orlitzky et al. (2003) argue that financially successful companies can afford to spend more money on social issues, but CSR also helps them become financially successful. The meta-analysis of Orlitzky et al. (2003) also supported this bidirectional causality, providing evidence that “both instrumental stakeholder theory and slack resources descriptions are accurate.” This study reviews the following papers.
Makni et al. (2008) use simple Granger causal model to examine the causal relationship between corporate social performance and financial performance. Their empirical analyses are based on a sample of 179 publicly listed Canadian firms covering 2004 and 2005. Social performance data is collected from Canadian Social Investment Database, and financial performance is measured by using ROA, ROE and market returns. They find no significant relationship between the composite measure of CSR performance and corporate financial performance (except for market returns). However, using individual measures of CSP, they find a robust and significant adverse impact on the environmental aspect of CSR performance on all financial performance measures. They state that from a short-term perspective, this is consistent with trade-off theory and negative synergy hypothesis. The trade-off theory supposes a negative impact of CSR on financial performance, supports that socially responsible behavior will net few economic benefits but its numerous costs will reduce profits and shareholder wealth. The negative synergy hypothesis supposes that higher levels of CSR lead to decreased financial performance, which in turn limits the socially responsible investments.
In the same way as Makni et al. (2008), Nelling and Webb (2009) adopt Granger causality models to examine the relation between CSR performance and financial results using KLD data. Their findings suggest that strong stock market performance leads to greater firm investment in aspects of CSR program related to employee relation, but social responsibility activities do not affect financial performance. The findings support trade-off hypothesis, which is different from the work of Makni et al. (2008). As McWilliams and Siegel (2001) discussed, the reason for different findings arises from omitted variables, such as R&D, and advertising expenditure.
Distinct from both Makni et al. (2008), Nelling and Webb (2009), Surroca et al. (2010) examine the effects of a firm’s intangible resources in mediating the relationship between corporate responsibility and financial performance. They hypothesize that previous empirical findings of a positive correlation between social and financial performance may be spurious because the researchers failed to account for the mediating effects of intangible resources. After using the Sustainalytics Platform rating to measure CSR variables, Surroca et al. (2010) conclude that no direct relationship between corporate responsibility and financial performance is found, and there is merely an indirect connection that relies on the mediating effect of a firm’s intangible resources. Their findings are based on the use of Sustainalytics database comprising 599 companies from 28 countries. The Work of Surroca et al. (2010) is important and unique because it helps better understand the underlying mechanisms linking CSP with financial outcomes.
Emerging evidence presented in some studies (e.g., Clarkson et al., 2011 and Arora and Dharwadkar, 2011) suggests that more profitable firms (i.e., those having sufficient financial resources/slacks) are more likely to engage in CSR activities. Clarkson et al. (2011) adopt the resource-based view of the firm theory and find that positive changes in firms’ financial resources in the prior periods lead to significant improvements in firms’ relative environmental performance in the subsequent periods. Furthermore, they find that significant improvements in environmental performance in previous periods can result in improvements in financial performance (ROA) in the following years after controlling for the influence of Granger causality. The result of 3SLS test shows that there is endogeneity between environmental performance and financial performance (Q ratio).
The recent work of Qiu et al. (2016) investigates the link between firms’ environmental and social disclosures and their financial performance, as well as establish the direction of causality between the two. Using Bloomberg ESG to measure CSR disclosure and ROA to measure CSP, results regarding market value analysis show that overall CSR disclosure, in particular, social disclosures matter to investors. Specifically, it finds that more social (environmental) disclosure in prior year reflects better social (environmental) performance as captured by higher employee productivity (more carbon eco-efficiency) in the current year. Notably, Qiu et al. (2016) focus on CSR disclosure rather than CSP, although he finds the evidence of causality between CSR and CFP.
To sum up, while some prior work (Callan and Thomas, 2009) assumes the causality to run from corporate social performance to corporate financial performance, other studies find mixed results. Furthermore, while the causality between CSP and CFP has been tested, only few of the studies (Clarkson et al., 2011; Surroca et al., 2010) find such “virtuous circle.” Last but not least, there seems to be a lack of understanding of the underlying mechanisms linking CSP with financial outcomes. As discussed by Aguinis et al. (2012), the academe seems to know quite a bit about the reasons why organizations engage in CSR and what happens as a result. In short, this knowledge gap refers to the need to conduct further research that can help us understand the conditions under which these results are more or less likely to be observed.
According to Grewatsch et al. (2015), the financial value of CSR is directly contingent upon the ability to influence stakeholders and their perception of the firm’s CSR activities. A firm’s CSR involvement may only be beneficial if it gains legitimacy and reward in the stakeholders’ eyes. Clear communication and reliable information create awareness and allow stakeholders to assess the firm’s CSR performance (Jayachandran et al. 2013). Due to information asymmetry and uncertainty between different stakeholders (Van der Laan et al. 2008), firms need to work on their CSR reputation and communication, as well as symbolic management. Through advertising intensity (Servaes and Tamayo 2013), media coverage (Rhou et al., 2016), high qualitative CSR reports (Schreck 2011), and consistent good treatment of different stakeholders over time (Wang and Choi 2013), firms can reduce the information gap, so that stakeholders find out more about the firm’s CSR engagement and reward it, which enhances the benefits of CSR. This study finds following papers that discussed the moderators of CSP-CFP link through review of the literature.
Schuler and Cording (2006) present a consumer behavioral model linking corporate social performance and corporate financial performance. The conceptual model examines the roles of information intensity and moral values in linking CSP with consumer purchase behavior. They predict that information intensity will influence the consumer’s brand attitude and expect consumer moral values to have a primary effect on purchase intentions, as well as to interact with information intensity in predicting purchase intentions. They conclude that information about a firm’s CSP influences the decisions of a stakeholder to engage in either supportive or deleterious behavior that ultimately affects the firm’s financial performance. Notably, although the study of Schuler and Cording (2006) is conceptual, it provides empirical implications for scholars to test the CSP-CFP behavioral model.
Barnett (2007) conceptually argue that CSR influences CFP through improving a firm’s relationship with relevant stakeholder groups. By developing the construct of stakeholder influence capacity, his study explains why the effects of corporate social responsibility on corporate financial performance vary across firms and time. The conceptual work provides a set of propositions to aid future research on the contingencies that produce variable financial returns to investment in corporate social responsibility.
Du et al. (2010) point out that stakeholders’ low awareness of and unfavorable attributions towards companies’ CSR activities undermine companies’ attempts to maximize business benefits from their CSR activities, which highlights a need for companies to communicate CSR more effectively to stakeholders. Du et al. (2010) present a conceptual framework of CSR communication, from message content and communication channels to company and stakeholder-specific factors that influence the effectiveness of CSR communication. His conceptual framework helps us understand different communication channels and contingency factors that may influence the effectiveness of CSR communication.
Servaes et al. (2013) rely on Barnett’s (2007) insight that the impact of CSR on firm performance depends on the ability of CSR to influence stakeholders in the firm. They focus on one of the key stakeholders, consumers, and suggest that a necessary condition for CSR to modify consumer behavior and, hence, affect firm performance, is consumer awareness of firm CSR activities. By employing the KLD Stats database over the period 1991–2005, and combine it with financial statement data obtained from Compustat, they find that firms with high consumer awareness, as proxied by advertising spending, can enhance firm performance by increasing CSR. However, firms with high public awareness are also penalized more when there are CSR concerns. Moreover, for firms with low public awareness, the impact of CSR activities on firm value is either insignificant or negative. The finding of an interaction between advertising intensity and CSR activities is consistent with theoretical work suggesting that without awareness customers are unable to reward CSR involvement (Sen and Bhattacharya 2001, McWilliams and Siegel 2001).
Rhou et al. (2016) investigate the moderating role of CSR awareness, measured by CSR media coverage, on the relationship between CSR and corporate financial performance (CFP) in the restaurant context. Based on the work of Servaes et al. (2013), they argue that the use of advertising expenditure as a proxy for CSR awareness may not be fully appropriate, as one could argue that advertising matters more in capturing positive rather than negative CSR awareness. By matching data observations for CSR data, media coverage, and financial data for each restaurant company from 1992 to 2002, they find that positive CSR activities add financial value to restaurant companies only if the companies effectively publicize their CSR involvement. Even with positive CSR activities if the companies fail to communicate with stakeholders, financial benefits do not occur. Regarding negative CSR awareness, they find that the effect of negative CSR on CFP becomes negative as negative CSR awareness increases, which means negative CSR activities significantly and adversely affect firm performance as the media expose restaurant companies’ socially- irresponsible activities to the public.
While the CSP-CFP relationship has been examined extensively, a clear relationship is not established due to the presence of moderators that may influence the outcome (Orlitzky et al., 2003). This study examines the role of media coverage as one such moderator. Based on above literature, CSR by itself is insufficient to impact CFP, because the strategic impact of CSR on firm performance relies not only on the firm’s efforts in CSR but also on stakeholders’ perceptions of those efforts. Hence, even though companies attempt to maximize advantages from their CSR strategies, low media coverage (first-level) and low media favorability (second-level) of CSR initiatives undermine CSR effects on firm outcomes (See Du et al., 2010; Servaes et al., 2013; Rhou et al., 2016). As a result, this study suggests that the effect of CSR on CFP depends on the media coverage, a brand-new moderator in the CSR literature.
The relationship between CSP and CFP
Most recent empirical and theoretical studies on CSP-CFP linkage indicate that they are positively associated. For example, Orlitzky et al. (2003) conduct a comprehensive meta-analysis study and find a positive relationship between CSP and CFP. Their results are based on 52 studies with a 33,878-sample size over a 30-year span. From the instrumental stakeholder theory, Orlitzky et al. (2003) claim that managers reap financial benefits by meeting the needs of stakeholders. In this line of thinking, firms view their stakeholders as part of an environment that must be managed to assure revenues, profits, and, ultimately, returns to shareholders. Attention to stakeholder concerns may help a firm avoid decisions that might prompt stakeholders to undercut or thwart its objectives (Berman, Wicks, Kotha, & Jones, 1999). Due to reciprocal benefits of the relationship between stakeholders and the organization, it is reasonable to predict that firm’s CSR performance is positively associated with its financial benefits.
Admittedly, high corporate social investments result in additional costs that put a firm at an economic disadvantage compared to other less socially responsible firms (Vance, 1975). It can also be argued that although engaging CSR activities entail additional costs, the company could gain more economic benefits in the long term (Qiu et al., 2016; Cormier et al., 2011). This line of thinking consider that the socially responsible company can gain competitive advantages by developing strong relationships with employees, customers, communities, and governments (Waddock and Graves, 1997; Sweeney et al., 2013; Saeidi et al., 2015). Specifically, firm’s CSR is a signal to indicate its products and services (Jayachandran et al., 2013; Servaes et al., 2013), and to attract socially active investors (Jones, 1995; Brammer and Millington, 2008; Cormier et al., 2011 & 2015). Moreover, employees show greater commitment to these superior CSR firms (Dutton, Dukerich, & Harquail, 1994). Greening & Turban (2000) point out that the excellent CSR firms are often perceived as attractive employers by job seekers. Finally, customers may respond to the positive social performance by increasing their demand for the firm’s products or services, or by paying premium prices (Bhattacharya & Sen, 2003). Based on above discussion, this study, therefore, proposes that:
Hypothesis 1: Firm’s CSP is positively related with its CFP.
The moderating role of media coverage between CSP and CFP
To better understand the effect of CSR on CFP, the current study draws attention to an important moderating variable: media coverage. Whether firms do well by doing good is a primary question in CSR literature (McWilliams et al., 2006). However, for the strategic objectives of CSR activities to be achieved (i.e., stakeholders compensate good companies and penalize bad companies), stakeholders must be cognizant of these activities (Servaes and Tamayo, 2013). It can be argued that the media plays a major role in helping stakeholders better understand the firm’s CSR. For example, McWilliams and Siegel (2001) stress that media coverage heightens stakeholder awareness of CSR behaviors of the firm. Thus, media coverage plays a central role in the process of recording and transmitting CSR information from firm to the public. By relating advertising to CSR, McWilliams and Siegel (2000, 2001) suggested that CSR-related advertising and media coverage may increase consumer awareness of CSR, which, in turn, increases the demand for socially responsible behavior and the returns to engaging in such behavior. Deephouse (2000, 2003) go one step further and claim that media coverage, especially media reputation, is a strategic resource for the firm. Specifically, Deephouse (2000) provides theoretical and empirical support for media reputation which leads to competitive advantage and superior financial performance. In the later work of Deephouse (2003), he provides evidence that the media information plays an indispensable role in the CSR-CFP relationship. Recently, Cahan et al. (2015) point out that good CSR firms rely on the media in building its reputation, consistent with Fombrun and Shanley’s (1990) reputation-building model. Their results suggest that a superior media image may be a missing link in prior studies that find insignificant or mixed results for the effect of CSR performance on firm performance.
From above discussion, even though media coverage is an omitted variable, the influences of media coverage is complicated. Specifically, Carroll and McCombs (2003) propose that “Firstly, the amount of news coverage that a firm receives in the news media is positively related to the public’s awareness of the firm issues. Secondly, the more positive that media coverage is for a particular attribute, the more positively will members of the public perceive that attribute”. The first proposition is about media effects on attention, which companies people are aware of likely to have an opinion about. The second proposition summarizes media effects on the affective images that people have of companies. In this way of thinking, the study proposes two corresponding hypotheses that include the media attention of the CSR topic (first-level), and the tone of the media CSR topic (second-level).
Hypothesis 2a: The relationship between CSP and CFP is more pronounced when firm’s CSR receives more media attention.
Hypothesis 2b: The relationship between CSP and CFP is more pronounced when firm’s CSR receives more media favorability.