Who Is the Beneﬁciary of Slack on Corporate Financial Performance and Corporate Philanthropy? Evidence from South Korea

: As stakeholder relations vary depending on ﬁrm characteristics, the associations among corporate ﬁnancial performance (CFP), corporate giving, and corporate social performance (CSP) are complex. In this paper, we contribute to the literature by exploring CFP as a predictor of CSP by differentiating the stakeholder groups that ﬁrms interact with; that is, primary versus secondary stakeholder relations. Our study also extends the existing literature by examining who the beneﬁciaries of corporate philanthropy are, and the role played on the CFP/CSP association. By extracting a sample of 52 ﬁrms and 312 ﬁrm-year observations from the Korea Economic Justice Institute database, we ﬁnd that while CFP has a positive effect only on primary stakeholder relations, corporate philanthropy has a positive impact on both primary and secondary stakeholder relations. Furthermore, we observe an overall inﬂuence of CFP on stakeholder relations when corporate philanthropy is high. Our ﬁndings suggest that differentiating multiple stakeholder groups together with the role played by corporate philanthropy provides a more valuable and meaningful analysis of the antecedents of CSP. Models 4 to 6 present the results of the regression analyses testing the hypothesized moderating role of high corporate giving on the relationship between CFP and CSP (Hypothesis 3). Model 4, using primary stakeholder relations as the dependent variable, showed a signiﬁcant positive effect of return on sales ( β 1 = 9.574; p < 0.001) and a signiﬁcant positive effect of high corporate giving ( β 2 = 0.887; p < 0.05), but not a signiﬁcant interaction. Conversely, in Model 5, where the dependent variable was Secondary Stakeholder relations, we found a signiﬁcant negative effect of returns on sale ( β 1 = − 3.482; p < 0.05) and a signiﬁcant positive interaction effect ( β 3 = 7.226; p < 0.01). A Wald test indicated that the combined effect (i.e., β 1 + β 3 ) on Secondary Stakeholders relations was positive and signiﬁcant ( χ 2 = 5.371, p < 0.05). As a robustness test, we reran the analysis using overall CSP as the dependent variable. Model 6 found that the effect of return on sales on overall CSP was positive ( β 1 = 6.09; p < 0.05) and that high corporate giving moderated the relationship between return on sales and overall CSP ( β 3 = 12.903; p < 0.01). Therefore, Hypothesis 3 is supported. The nature of this interaction is illustrated in Figure 2.


Introduction
The association between corporate financial performance (CFP) and corporate social performance (CSP) has been intensively explored over the last decades [1][2][3]. While the instrumental stakeholder theory suggests that CSP influences CFP, the slack resources hypothesis argues that firms with superior CFP may potentially have greater freedom to allocate resources into social and environmental activities, which, in turn, will improve CSP [4,5]. While the instrumental approach has been the dominant focus of CSP research, research on the antecedents of CSP has received relatively much less research attention [6].
Most empirical studies have found an overall positive CSP/CFP relationship. As pointed out by several literature review papers [7,8] and meta-analyses [9][10][11], a recent meta-analysis [12] provides new insights in terms of the causal direction between CSP and CFP; while there is strong evidence supporting CSP as a predictor of CFP, the direct reverse causal direction arguing CSP as a consequence of CFP has less empirical support [5].
In this study, we posit that the lack of empirical findings for a direct positive causal effect of CFP on CSP may be the result of failing to consider different stakeholder relations as major dimensions of CSP. Prior research suggests that corporate social engagement responds to internal or external stakeholders based on managers' perceptions of their relative importance [13]. Specifically, we contribute to the Sustainability 2019, 11, 252 2 of 12 literature exploring CFP as a predictor of CSP by differentiating the stakeholder groups that firms interact with; that is, primary from secondary beneficiaries of corporate slack resources. CSP actions targeting primary stakeholders, rather than pursuing moral capital, are likely to create beneficial exchanges between the firm and critical stakeholders such as investors, employees, customers, and suppliers. Indeed, CSP activities with these primary stakeholders are essential to the firm's core business [14]. Because these actions are consistent with the firm's profit-making goal, we predict that primary stakeholders will be the main beneficiaries of corporate financial slack resources.
Our study also extends the existing literature by examining who are the beneficiaries of corporate philanthropy. While the positive effect of CSP on CFP has been generally accepted, corporate giving has long been considered as a corporate social duty and has been used as a measure of CSP [15,16]. In most studies, corporate philanthropy is not distinguished from CSP, which compromises their results. Also, research on antecedent conditions for good CSP is scarce. Corporate giving is a discretionary managerial behavior that is not critical for firm survival. Because corporate philanthropy has no direct influence on primary stakeholders' expectations, in this study, we posit that secondary stakeholders will be the main beneficiaries of corporate giving. Finally, we also explore the overall interaction effect of CFP and corporate philanthropy on CSP. Specifically, we expect a positive influence of CFP on overall stakeholder relations when corporate giving is high.
We test our predictions using a sample of 52 firms and 312 firm-year observations from the Korea Economic Justice Institute (KEJI) database. Our findings provide strong support for our hypotheses, allowing us to identify the ideal beneficiary of corporate financial slack resources given stakeholder relations, where CFP and corporate philanthropy were considered to be two distinct antecedents of CSP.
Our results indicate that the roles played by wealth and goodness differ in overall stakeholder relations.
The remainder of this study is organized as follows. First, we propose our hypotheses. Then, we present the empirical method used to test the hypotheses and present our results. The fourth section discusses our findings and concludes.

CSP and Stakeholder Relations
Corporate social responsibility, conceptualized as the role played by corporations in society, implies that firms should not always pursue profit maximization. The firm's stakeholders have a responsibility to hold the firm accountable for its actions [17]. The social responsibility of any given firm encompasses 'the economic, legal, ethical, and discretionary expectations that society has of business organizations at a given point in time' [15] (p. 500).
CSP may be viewed as the outcome of a firm in terms of social and environmental efforts. Rowley and Berman [18] (p. 398) defined CSP as 'social outcomes of firm behaviors'. Scott [19] (p. 143) stated that 'organizations produce outputs over whose characteristics they typically exercise considerable control, but outcomes represent the joint product of organizational performance and environmental response. In this sense, environments directly influence outcomes'. Outcomes must be followed by an analysis of the net benefits of and stakeholders' satisfaction with organizational actions [7]. The importance of these stakeholders in CSP must, therefore, be considered.
A stakeholder in an organization is defined as 'any group or individual who can affect or is affected by the achievement of the organization's objectives' [20] (p. 46). Clarkson [21] and Godfrey et al. [14] argued that corporations are influenced by, or have an influence on, two types of stakeholders: primary (direct) stakeholders and secondary (indirect) stakeholders. Primary stakeholders, such as shareholders and investors, employees, customers, and suppliers, are essential to core business operations. Without their participation, a firm cannot be able to continue in existence. On the contrary, secondary stakeholder groups, such as the community, general public, and the media, are not critical for the survival of the firm. While secondary stakeholders have legitimate claims on the firm and can eventually influence the firm's primary stakeholders, they do not have the power to enforce those claims. Nevertheless, secondary stakeholders 'have the capacity to mobilize public opinion in favor of, or in opposition to, a company's performance' [21] (p. 108). However, research on this issue is scarce. One exception is the work of Godfrey et al. [14]. These authors found that corporate efforts targeting secondary stakeholders can act as an 'insurance-like' benefit by tempering negative judgments and sanctions toward firms after public scrutiny. Conversely, they found that commitment towards primary stakeholders does not necessarily yield such benefits.

The Effect of CFP on CSP
Slack resources theory points out that firms with available financial slack resources from superior CFP may decide to allocate those financial resources on 'doing good by doing well', which, in turn, may improve their CSP [22]. While some empirical research provides support for this argument [2,3,14,23], a recent meta-analysis of 42 studies indicated that a high level of CFP is not significantly associated with firms' subsequent CSP [12]. However, several studies showed a positive association. For example, Cochran and Wood [24] (1984) provided theory and evidence that a significant correlation between financial and social performance does exist. Waddock and Graves [22] found that CFP fosters a financial slack that can be eventually reinvested into CSP. Similarly, Rodgers et al. [3] found that a firm's CFP positively affects its CSP commitment and social perception in the subsequent period. More recently, Surroca et al. [5] provided evidence suggesting a virtuous cycle between CSR and CFP as mediated by intangible assets, such as innovation, reputation, human capital, and culture. However, none of the previous research has considered exploring CFP as a predictor of CSP by differentiating the stakeholder groups that firms interact with; that is, primary versus secondary stakeholder relations.
The aforementioned mixed results in the CFP/CSP literature can be explained from the viewpoint of stakeholder theory. Because previous studies tend to examine CSP in general, the roles of various stakeholders in CSP were not specified. Primary stakeholders may have different interests from those of secondary stakeholders. Both may influence the performance and survival of for-profit organizations. However, good financial performance is critical for firms to survive. Therefore, relations between firms and their primary stakeholders are influenced by the firms' CFP. Thus, we suggest the following hypothesis: Hypothesis 1. CFP will have a positive effect on CSP in terms of primary (direct) stakeholder relations.

Corporate Philanthropy and CSP
Corporate philanthropy entails the 'gifts given by corporations to social and charitable causes, such as support for education, culture, or the arts; for minorities or health care; or for relief funds for victims of natural disasters' [25] (p. 1161). Research on corporate philanthropy has mainly focused on two areas [26]. One research stream explores the motivation behind the corporate monetary donation. The results of these studies show that profit maximization and utility maximization motivate firms to make monetary donations [27]. Other studies have explored the relationship between corporate giving and CFP. However, the results of these studies are inconsistent. While good management theory posits that corporate philanthropy results in good financial performance [12,28,29], slack resource theory suggests that a good financial performer can make more frequent monetary donations [30,31].
In the majority of these studies, corporate philanthropy is not distinguished from CSP. However, corporate giving may be an antecedent of CSP from the viewpoint of stakeholder relations; it is not an outcome, but rather an output. As stated by Scott [19], an output can be controlled, but the outcome is a joint result from organizational performance and environment reaction. That is, the level of corporate giving can be controlled by a managerial decision. However, CSP is an outcome measure. Therefore, distinguishing between corporate giving and CSP (stakeholder relations) is necessary.
In addition, corporate giving has a dual effect on CFP. While corporate giving results in a good reputation (i.e., an intangible resource), it also involves a financial cost. This dual effect is the cause of the inconsistent results of the empirical research on the association between corporate giving and CFP.
As corporate giving is a discretionary managerial behavior, it is not critical for firm survival. Therefore, in terms of stakeholder relations, it has no direct influence on primary stakeholders' expectations, but only an indirect influence, in that it results in good relations with secondary stakeholders. However, active involvement in corporate giving signals to shareholders that a company has a large pool of financial slack [25,32]. According to agency theory, managers (the agent) may be motivated to act in their self-interests rather than those of the firm's owners and other shareholders [33]. Shareholders may, therefore, suspect managers of engaging in opportunistic decision-making not related to corporate giving. Hence, corporate giving may not enhance primary stakeholder relations. Because CFP is associated only with primary stakeholder relations, the results of studies using overall stakeholder relations as an indicator of CSP do not provide the whole picture. While their results imply that primary stakeholders are the only beneficiaries of CFP, the role of secondary stakeholders is ignored (see Figure 1). This reasoning leads us to propose our second hypothesis: Hypothesis 2. Corporate giving will have a positive effect on CSP in terms of secondary (indirect) stakeholder relations. of the inconsistent results of the empirical research on the association between corporate giving and CFP.
As corporate giving is a discretionary managerial behavior, it is not critical for firm survival. Therefore, in terms of stakeholder relations, it has no direct influence on primary stakeholders' expectations, but only an indirect influence, in that it results in good relations with secondary stakeholders. However, active involvement in corporate giving signals to shareholders that a company has a large pool of financial slack [25,32]. According to agency theory, managers (the agent) may be motivated to act in their self-interests rather than those of the firm's owners and other shareholders [33]. Shareholders may, therefore, suspect managers of engaging in opportunistic decision-making not related to corporate giving. Hence, corporate giving may not enhance primary stakeholder relations. Because CFP is associated only with primary stakeholder relations, the results of studies using overall stakeholder relations as an indicator of CSP do not provide the whole picture. While their results imply that primary stakeholders are the only beneficiaries of CFP, the role of secondary stakeholders is ignored (see Figure 1). This reasoning leads us to propose our second hypothesis:

Interaction of CFP and Corporate Philanthropy
In a recent study, Wang and Choi [34] argued that establishing good stakeholder relations depends not only on outstanding CSP, but also on the consistency of that performance. Both of these factors have significant financial implications. However, the conditions under which consistently high CSP can be maintained among stakeholders have not been clearly identified.
In this study, we argue that high CFP enhances primary stakeholder relations. Also, the higher the firm's commitment to corporate giving, the better the secondary stakeholder relations. Figure 1 provides a visual representation of these concepts, which is fundamental to our research model.
These two arguments infer an interaction between CFP and corporate giving, and that this interaction affects stakeholder relations, as shown in Figure 1. Thus, both CFP and corporate giving are essential for maintaining a consistently high CSP among stakeholders. On the basis of this logic, we develop a third hypothesis: Hypothesis 3. CFP will have a positive effect on overall stakeholder relations when corporate giving is high.

Interaction of CFP and Corporate Philanthropy
In a recent study, Wang and Choi [34] argued that establishing good stakeholder relations depends not only on outstanding CSP, but also on the consistency of that performance. Both of these factors have significant financial implications. However, the conditions under which consistently high CSP can be maintained among stakeholders have not been clearly identified.
In this study, we argue that high CFP enhances primary stakeholder relations. Also, the higher the firm's commitment to corporate giving, the better the secondary stakeholder relations. Figure 1 provides a visual representation of these concepts, which is fundamental to our research model.
These two arguments infer an interaction between CFP and corporate giving, and that this interaction affects stakeholder relations, as shown in Figure 1. Thus, both CFP and corporate giving are essential for maintaining a consistently high CSP among stakeholders. On the basis of this logic, we develop a third hypothesis: Hypothesis 3. CFP will have a positive effect on overall stakeholder relations when corporate giving is high.

Sample
To test our hypotheses, panel data on CSP were collected from the Korean Economic Justice Institute (KEJI) index for the years 2005 to 2010. The sample included firms that were consecutively nominated on the KEJI index for six years. The sample included 52 firms for which all variables were available. The financial data and corporate giving data of all firms selected appeared in the KisValue database, except for banks/financial services firms because of the systematic operating and regulatory differences in these industries.
To investigate the impact of CFP and corporate giving on CSP, in terms of primary and secondary stakeholder relations (Hypotheses 1 and 2, respectively), we propose the following model: To investigate the moderating effect of high corporate giving on the relation between CFP and CSP (Hypothesis 3), we use the following model:

Dependent Variables: Primary Stakeholders, Secondary Stakeholders, and Overall CSP
We used the KEJI index to measure CSP in three different ways. The KEJI index is composed of seven evaluation criteria: soundness, justice, economic development contribution, employee satisfaction, customer satisfaction, community service contribution, and environmental protection. To measure CSP in direct stakeholders (Primary Stakeholders), we aggregated the scores provided by KEJI on soundness, justice, economic development contribution, employee satisfaction, and customer satisfaction. For the measurement of CSP in indirect stakeholders (Secondary Stakeholders), we aggregated the KEJI scores on community service contribution and environmental protection. Finally, we created a general CSP measurement (Overall CSP) by aggregating all KEJI scores.

Independent Variables
CFP. We used a profitability measure, return on sales, that is, net income divided by sales, as the measure of CFP. Return on equity could be another proxy for CFP. However, this alternative of profitability is considered more complex because changes in return on equity can be the result of changes in return on sales, asset turnover, and gearing [35,36]. Following Palepu et al. [37], we employed return on sales because a change in return on equity is typically the result of a change in profit margins.
Corporate giving. To test Hypotheses 1 and 2, corporate giving is measured as the ratio of monetary donations to sales. According to the Korean Corporate Community Relations White Book, the ratios of monetary donations to total expenses of social contributions in Korea were 56.7%, 54.9%, 53%, and 54.9% for the years 2006, 2007, 2008, and 2009, respectively. Therefore, monetary donation was chosen as the most appropriate proxy to measure corporate giving in Korea. Hypothesis 3 predicts that CFP will have a positive effect on overall CSP when corporate giving is high. To test this hypothesis, we created a dummy variable, named high corporate giving, that takes a value of 1 when the firm's ratio of monetary donations to sales is higher than the average for the full sample, and 0 otherwise.

Control Variables
Firm age was included as a control variable to control for demographic effects of firms. Sales variable was also used to control for the size effect of firms. Other dummy variables used in previous studies [38] were also included, such as membership in the manufacturing industry, Chaebol/Korean conglomerates, or the Federation of Korean Industries (FKI). Manufacturing firms, compared with services companies, are more complex in nature; pollute more; and, as a result, tend to engage more in CSP activities related to environment and community. Therefore, we expect our manufacturing variable to be negatively (positively) associated with CSP in direct (indirect) stakeholders. A distinguishing characteristic of the Korean system is the existence of Chaebols, which are business conglomerates equivalent to family businesses. These businesses have been highly criticized by their lack of transparency, low accountability of managers, and little separation of ownership and control. Because Korean Chaebols often adopt CSP in order to restore their damaged reliability and reputation [38], we expect a positive association between this Chaebols dummy variable and CSP. The FKI has been very active from 2003 in promoting CSR or ethical departments, as well as the adoption of codes of conduct, at firms' members. Therefore, we expect a general positive association between FKI and CSP. Table 1 shows the descriptive statistics and correlation analysis for the independent and dependent variables. The mean CSP on Primary Stakeholders relations (mean = 38.44) is significantly greater than that on Secondary Stakeholders relations (mean = 10.30, t = 189.60.90, p < 0.001), indicating that our sample of Korean firms tends to engage more in direct rather than indirect stakeholder relations. Primary and secondary stakeholder relations were not significantly correlated (r = −0.003). Consistent with hypotheses 1 and 2, we found that return on sales was positive and significantly correlated with Primary Stakeholders relations (r = 0.305, p < 0.01), but not with Secondary Stakeholders relations (r = 0.013), and that donation ratio to sales was positively correlated with both primary and secondary stakeholder relations (r = 0.274 and 0.235, respectively; p-values < 0.01). Return on sales and donation ratio to sales were also positively correlated (r = 0.290, p < 0.01), indicating that higher CFP was associated with higher corporate giving. As some significant correlations were observed, the variance inflation factor was calculated to check for the possibility of multicollinearity. In this test, most values were <2. The maximum value was only 1.946. Because no values exceeded 10, multicollinearity does not seem to be a problem. Table 2 provides the results of the panel generalized least squares (GLS) analysis for testing our hypotheses. Variances of errors were assumed to be heteroscedastic and with autocorrelation. The results of the likelihood ratio test supported the assumption of heteroscedasticity in our models (all p-values < 0.001). The results of the Wooldridge test showed an autocorrelation in the panel data (all p-values < 0.05).

Results
Primary Stakeholders was the dependent variable employed in Model 1 to test Hypothesis 1. In Model 2, which examined Hypothesis 2, Secondary Stakeholders relations was employed as the dependent variable. Models 1 and 2 showed that return on sales of the previous year positively influenced only Primary Stakeholders relations (β 1 = 8.741; p < 0.001; model 1), thus supporting Hypothesis 1. These models also demonstrated that the ratio of donation to sales of the previous year positively influenced both Primary Stakeholders relations (β 2 = 66.692; p < 0.001; Model 1) and Secondary Stakeholders relations (β 2 = 60.375; p < 0.001; Model 2). In Model 3, we performed a robustness analyses using a single independent measure of CSP and found that both return on sales and ratio of donation to sales of the previous year positively influenced overall CSP (β 1 = 7.698 and β 2 = 167.232; p < 0.01 and p < 0.001, respectively; Model 3). Hence, Hypothesis 2 is also supported. , otherwise 0; return on sales = net income over sales revenues; donation ratio to sales = total expenses of social contributions over sales revenues; primary stakeholder = sum of the scores on the Korea Economic Justice Institute (KEJI) index for the following evaluation criteria: soundness, justice, economic development contribution, employee satisfaction, and customer satisfaction; secondary stakeholder = sum of the scores on the KEJI index for the following evaluation criteria: community service contribution, and environmental protection; overall corporate social performance (CSP) = sum of the scores on the KEJI index for the following evaluation criteria: soundness, justice, economic development contribution, employee satisfaction, customer satisfaction, community service contribution, and environmental protection; * p < 0.05, ** p < 0.01 (two-tailed). Data obtained for the years 2005 to 2010.  Definitions of variables: firm age = number of years of incorporation of the company; sales = sales revenues expressed in 10 billion KRW; manufacturing = a dummy variable, taking value 1 if the firm belongs to manufacturing industry, otherwise 0; Chaebol = a dummy, taking value 1 if the firm belongs to a Chaebol conglomerate, otherwise 0; member of FKI = a dummy, taking value 1 if the firm belongs to Federation of Korean Industries, otherwise 0; return on sales = net income over sales revenues; donation ratio to sales = total expenses of social contributions over sales revenues; primary stakeholder = sum of the scores on the KEJI index for the following evaluation criteria: soundness, justice, economic development contribution, employee satisfaction, and customer satisfaction; secondary stakeholder = sum of the scores on the KEJI index for the following evaluation criteria: community service contribution, and environmental protection; overall CSP = sum of the scores on the KEJI index for the following evaluation criteria: soundness, justice, economic development contribution, employee satisfaction, customer satisfaction, community service contribution, and environmental protection; high donation ratio to sales (dummy) = a dummy, taking value 1 if the firm's donation ratio to sales is higher than the average, otherwise 0; * p < 0.05, ** p < 0.01, *** p < 0.001. Number of observations = 312; number of groups = 52.
Sustainability 2019, 11, 252 9 of 12 Models 4 to 6 present the results of the regression analyses testing the hypothesized moderating role of high corporate giving on the relationship between CFP and CSP (Hypothesis 3). Model 4, using primary stakeholder relations as the dependent variable, showed a significant positive effect of return on sales (β 1 = 9.574; p < 0.001) and a significant positive effect of high corporate giving (β 2 = 0.887; p < 0.05), but not a significant interaction. Conversely, in Model 5, where the dependent variable was Secondary Stakeholder relations, we found a significant negative effect of returns on sale (β 1 = −3.482; p < 0.05) and a significant positive interaction effect (β 3 = 7.226; p < 0.01). A Wald test indicated that the combined effect (i.e., β 1 + β 3 ) on Secondary Stakeholders relations was positive and significant (χ 2 = 5.371, p < 0.05). As a robustness test, we reran the analysis using overall CSP as the dependent variable. Model 6 found that the effect of return on sales on overall CSP was positive (β 1 = 6.09; p < 0.05) and that high corporate giving moderated the relationship between return on sales and overall CSP (β 3 = 12.903; p < 0.01). Therefore, Hypothesis 3 is supported. The nature of this interaction is illustrated in Figure 2. These models also demonstrated that the ratio of donation to sales of the previous year positively influenced both Primary Stakeholders relations (β2 = 66.692; p < 0.001; Model 1) and Secondary Stakeholders relations (β2 = 60.375; p < 0.001; Model 2). In Model 3, we performed a robustness analyses using a single independent measure of CSP and found that both return on sales and ratio of donation to sales of the previous year positively influenced overall CSP (β1 = 7.698 and β2 = 167.232; p < 0.01 and p < 0.001, respectively; Model 3). Hence, Hypothesis 2 is also supported. Models 4 to 6 present the results of the regression analyses testing the hypothesized moderating role of high corporate giving on the relationship between CFP and CSP (Hypothesis 3). Model 4, using primary stakeholder relations as the dependent variable, showed a significant positive effect of return on sales (β1 = 9.574; p < 0.001) and a significant positive effect of high corporate giving (β2 = 0.887; p < 0.05), but not a significant interaction. Conversely, in Model 5, where the dependent variable was Secondary Stakeholder relations, we found a significant negative effect of returns on sale (β1 = −3.482; p < 0.05) and a significant positive interaction effect (β3 = 7.226; p < 0.01). A Wald test indicated that the combined effect (i.e., β1 + β3) on Secondary Stakeholders relations was positive and significant (χ 2 = 5.371, p < 0.05). As a robustness test, we reran the analysis using overall CSP as the dependent variable. Model 6 found that the effect of return on sales on overall CSP was positive (β1 = 6.09; p < 0.05) and that high corporate giving moderated the relationship between return on sales and overall CSP (β3 = 12.903; <0.01). Therefore, Hypothesis 3 is supported. The nature of this interaction is illustrated in Figure 2. The results for control variables in Table 2 are generally consistent with previous studies [3, 5,38]. We found that the coefficients on firm age (sales) are negative (positive) and significant, with the exception of the Primary Stakeholders relations regressions (Models 1 and 4). The coefficients on Chaebol were positive and significant for all regressions. As expected, we found the coefficient on manufacturing to be negatively associated with Primary Stakeholders relations (Models 1 and 4), but positively associated with Secondary Stakeholders relations and overall CSP (Models 2, 3, 5, and 6). Conversely, the coefficient on FKI membership was negatively associated with Primary Stakeholders relations (Models 1 and 4), but positively associated with Secondary Stakeholders relations (Models 2 and 5). The results for control variables in Table 2 are generally consistent with previous studies [3, 5,38]. We found that the coefficients on firm age (sales) are negative (positive) and significant, with the exception of the Primary Stakeholders relations regressions (Models 1 and 4). The coefficients on Chaebol were positive and significant for all regressions. As expected, we found the coefficient on manufacturing to be negatively associated with Primary Stakeholders relations (Models 1 and 4), but positively associated with Secondary Stakeholders relations and overall CSP (Models 2, 3, 5, and 6). Conversely, the coefficient on FKI membership was negatively associated with Primary Stakeholders relations (Models 1 and 4), but positively associated with Secondary Stakeholders relations (Models 2 and 5).
In summary, Table 2 shows that, after controlling for firms' operating characteristics, we found evidence supporting the following: (1) CFP had a positive effect on CSP in terms of primary stakeholder relations and overall CSP; (2) corporate giving was positively associated with subsequent CSP in terms of both primary and secondary stakeholder relations; and (3) a positive moderating effect of CFP on CSP when corporate giving was high.

Discussion and Conclusions
This study investigated the discriminatory benefit of CFP and corporate philanthropy from the viewpoint of stakeholder relations. The underlying question concerned stakeholder relations in companies that differed in terms of financial and social performance. Although the variables return on sales and monetary donation were both related to the availability of financial slack resources, the beneficiaries (stakeholders) of these slack resources differed.
In support of Hypothesis 1, the analysis showed that return on sales (CFP) fulfilled only the interests of primary stakeholders, not those of secondary stakeholders. Thus, primary stakeholders were the only beneficiaries of CFP. Hypothesis 2 posited that corporate giving would enhance secondary stakeholder relations rather than primary stakeholder relations. This expectation was related to the agency problem. Principal stakeholders may perceive monetary donations as a kind of agency cost. This perception may reduce their satisfaction with managerial decisions to make a monetary donation. However, the results of the analysis are slightly different from our expectation. Our findings show that the interests of both primary and secondary stakeholders are fulfilled as a result of corporate giving. According to Clarkson [21], primary stakeholder groups are typically comprised of not only shareholders and investors, but also employees, customers, and suppliers. Hence, many primary stakeholders may not hold the same principles as the manager, and thus may not recognize the agency problem of corporate giving. Thus, no reduction in the positive effect of corporate giving on primary stakeholder relations is observed. Primary and secondary stakeholder groups must be distinguished for this argument to be valid.
The positive interaction of CFP and high corporate giving on overall stakeholder relations (Hypothesis 3) implies the conditions required for consistently high CSP among stakeholders. CFP positively influenced overall stakeholder relations only when corporate giving was high. Thus, firms must demonstrate good financial and social performance to gain social support.
Because the effect of CFP on CSP (multiple stakeholder relations) is limited, investigating the relationship between CFP and CSP is no longer meaningful. Classifying CSP into multiple stakeholder groups provides a more valuable and useful analysis. Strategies to meet stakeholder expectations can then be explored. This is the shortcut to becoming both a good company and a financially healthy company.
Our study has limitations. First, we analyzed firms being rated by the KEJI, an independent party, but stakeholders can also rely on a firm's self-reported CSP through websites, annual reports, standalone CSP reports, and so on. We admit that self-reported CSP information may have an effect on our results. Second, we explored the case of Korean firms, so we recommend being cautious about the generalization of our results.
Future research can examine whether intangible assets, such as intellectual capital, innovation, and so on, can also moderate the association between CFP and different stakeholder relations through corporate giving, as primary and secondary stakeholders may also be influenced by intangibles channels in a different manner. Finally, future research can also re-explore the virtuous cycle between CFP and CSP [3,5], where corporate giving might play a critical role in this potential bidirectional association.