Analysis of the Importance of Job Insecurity, Psychological Safety and Job Satisfaction in the CSR-Performance Link

: Existing works on the association between corporate social responsibility (CSR) and performance have investigated how CSR a ﬀ ects external stakeholders (e.g., shareholders, customers, local communities, and environment). However, the e ﬀ ect of CSR on internal stakeholders such as employees who ultimately determine organizational performance is relatively underexplored. Institutional theory suggests that institutional enablers, e.g., CSR practices, inﬂuence macro-level variables, like organizational performance, via micro-level intermediating processes, such as perceptions, attitudes and behaviors of employees. Thus, this paper proposes that internal mechanisms are essential to explain the CSR-performance link. Using a 3-wave time-lagged survey data of 301 workers in South Korean ﬁrms (144 men and 157 women, average age: 40.30), this research examines how psychological safety and job satisfaction of an employee function as sequential mediators in the CSR-performance link. Moreover, the current research also investigates how job insecurity negatively moderates the link between CSR and psychological safety. The results demonstrated that psychological safety and job satisfaction function as sequential internal mediators in the link. In addition, job insecurity negatively moderated the inﬂuence of CSR on psychological safety. The results suggest that an employee’s perceptions and attitudes, such as job insecurity, psychological safety, and job satisfaction, signiﬁcantly inﬂuence the relationship between CSR and performance.


Introduction
Since the 1960s, scholars and practitioners in the business field have become increasingly interested in corporate social responsibility (CSR). Although the purpose of a firm, from a traditional perspective, is to maximize profits, social awareness regarding the negative effects of the pursuit of profit-maximization has been gradually spreading in the academic and business worlds [1,2]. Based on this trend, it was widely recognized that companies should both seek to maximize profits and meet the needs of various stakeholders (e.g., employees, consumers, suppliers, local communities, governments, and the environment) who coexist in its business environment. Based on increased interest, many scholars have investigated the performance implications of CSR activities [1][2][3][4].
Although considerable work has examined the association between CSR and performance, some issues still have not been addressed. First, existing studies on the CSR-performance link reported that the association between the two variables is inconclusive [4][5][6]. For example, some works have shown that CSR positively influences organizational performance. CSR also functions as a strategic resource to increase its competitive advantage by enhancing the firm's reputation, brand value, and attractiveness [4,[7][8][9]. However, other works have argued that CSR is insignificant or has a negative impact on organizational performance due to its inefficient use of organizational resources [3,[10][11][12][13]. Scholars suggest that this inconsistent outcome originates from a lack of research on the moderators/mediators in the CSR-performance link [3,[10][11][12][13]. Therefore, to deal with the mixed outcomes of the influence of CSR on organizational performance, more attention should be paid to the exact mechanisms and contingent factors in the association [9,13,14].
Second, existing works on the CSR-performance link have relatively underexplored the importance of the employee's reactions to CSR activities (i.e., internally-oriented approach). Those studies have mainly taken an externally-oriented approach, which focuses on the impact of CSR on external stakeholders (e.g., shareholders, customers, and communities) as an intermediating mechanism to explain the relationship [5,15,16]. For example, an empirical study [7] showed that brand assets and renown of a firm act as mediators in the CSR-performance link. In addition, another study [9] reported that the organization's renown and customer satisfaction function as mediators in the CSR-performance link. Although this line of work has helped broaden the understanding of the CSR-performance link, there is still a lack of work on how members within an organization as internal stakeholders influence the association within an organization. In fact, organizational members are entities who substantially plan and execute CSR activities [17][18][19], so their attitudes and behaviors towards CSR are important to determine whether CSR practices may succeed or fail in the organization. This is the reason why an internal-oriented perspective would supplement the previous externally-oriented perspective [5,15,16,20].
Third, although some studies on CSR have taken the internally-oriented approach, those have underexplored the role of job characteristics in the process of the employee's reactions to CSR activities [15,16]. According to the review paper of Gond and his colleagues [15], employees' individual differences (e.g., cultural values and moral values), individual attitudes toward CSR (e.g., personal beliefs about CSR importance, CSR-induced attributions of motives), and perceptions about the organization (e.g., perceived organizational support, psychological contract breach) have been considered as boundary conditions for their responses to CSR. However, despite the importance of job characteristics on the perceptions, attitudes, and behaviors of employees [21], studies which investigate the moderating role of job characteristics on the link between CSR and employees' reactions are scarce.
To deal with the above issues, this paper attempts to provide a comprehensive picture of the CSR-performance link by integrating the two perspectives (i.e., externally versus internally). Specifically, the current research investigated the micro-intermediating mechanism in the CSR-performance link by focusing on the sequential intermediating influence of psychological safety and job satisfaction in the CSR-performance link. Psychological safety can be defined as feeling "able to show and employ one's self without fear of negative consequences to self-image, status, or career" (p. 708, [22]). Although existing studies have underexplored the impact of CSR on psychological safety [15,16], relying on pertinent works [23][24][25], this paper suggests that CSR practices may enhance the level of employees' psychological safety.
Based on previous studies [21,22,26], the current research proposed a hypothesis that the psychological safety of an employee enhances their level of job satisfaction. Job satisfaction is defined as an employee's cognitive and emotional evaluation of their job, which has either a positive or negative form [27]. Traditionally, this concept has been considered as one of the most critical factors due to the long-standing view that organizational performance is the result of the employee's satisfaction [17,18,27,28]. By integrating the arguments described above, this paper suggests that the level of psychological safety and job satisfaction may function as sequential mediators in CSR-performance link.
Furthermore, the current research explores the contextual variables that affect the association between CSR and psychological safety. As though the argument that CSR activities enhance the quality of employee psychological safety may be generally acceptable, it cannot be applied in all situations and Sustainability 2020, 12, 3514 3 of 16 environments in the same way. Employees are not influenced by the systems or rules of their firm in a unilateral way. They tend to actively interpret the meaning of CSR activities based on the situations or circumstances around the organization [29,30]. As a result, they are likely to perceive and respond to CSR practices in different ways. In this paper, we investigate the contingent role of job characteristics because those have a significant impact on employee perceptions, attitudes, and behaviors. Among various job characteristics, this study focused on job insecurity, which is one of the most important job characteristics and serves as the fundamental factor of trust and loyalty that employees have in their organization [31][32][33][34]. This concept is considered as one of the most substantial causes of job stress for employees since the employees are likely to experience the high rates of unemployment in the period of economic crisis. Considering that not only the problems pertinent to the job insecurity significantly affects their working and organizational outcomes but also employees may perceive that providing secure jobs for them is one of the most important responsibilities of the firm, the relationship between job insecurity and CSR activities should be examined. This paper suggests that employees are likely to doubt the authenticity of the firm's CSR activities when they feel a sense of job insecurity. In that situation, they perceive that there are hidden questionable purposes in the firm's CSR activities. Such doubt may deteriorate the positive effect of CSR on employee psychological safety.

CSR and Psychological Safety
The current research proposes that CSR improves the level of psychological safety of employees in an organization. Although there have been few studies directly exploring the association between CSR and employees' psychological safety [15,16,26,35], our argument may be justified based on the characteristics of CSR activities. CSR practices include various kinds of charities and altruistic actions for employees, customers, local communities and the natural environment [36]. CSR activities for employees consist of education/development programs, safety programs and employee-friendly benefits [37], which make employees feel that they are valued and protected by their organization. Thus, CSR activities for employees may "directly" make employees perceive that their organization is a psychologically safe place. This perception makes them manifest their true self, thoughts and feelings without fear of judgment and rejection [25]. This will eventually increase the level of employee psychological safety within the organization [24,38].
In addition, CSRs for customers, communities and the natural environment may also enhance the level of employee psychological safety in an "indirect" manner. CSRs for customers refer to prioritizing customer satisfaction and consumer rights beyond legal requirements and practices [36]. When a company prioritizes customer interests over their own interests, employees may consider the organization to be ethical and trustworthy. These perceptions are likely to boost their psychological safety [23,25]. Moreover, organizations that actively conduct CSR practices for the local community tend to emphasize the importance of altruistic behaviors for various members in the community (e.g., manufacturers, suppliers, government organizations, non-governmental organizations, and the socially disadvantaged). Considering that employees are important members of the community, they are likely to feel that the firm protects them. This would increase their psychological safety [23]. Lastly, when a firm actively carries out CSR activities for the natural environment, employees will perceive that the company as a moral representative positively contributes to the sustainability of the earth [1]. Employees who perceive that the company is ethical are more likely to feel a lower level of uncertainty and anxiety when faced with various challenges in an organization [39]. Then, the level of employees' psychological safety may be enhanced. Relying on the above arguments, this paper suggests the following hypotheses.

Psychological Safety and Job Satisfaction
Existing studies reported that the psychological safety of employees enhances their level of job satisfaction [21,22,26]. When an employee feels psychological safety in an organization, he or she may feel protected and valued by the organization. Considering that the quality of the relationship with their colleagues or organization is an important factor that determines their overall job satisfaction (Hackman & Oldham, 1976), those kinds of positive feelings toward the organization are likely to increase the level of job satisfaction [26,40,41].
In addition, previous studies on workplace satisfaction suggested that employees who perceive that they are safe without the feeling of blame or rejection may experience a low degree of stress at work [42,43]. Relying on a psychologically secure base, they are likely to feel less anxiety of failure, and actively seek challenging new and difficult tasks. As a result, they tend to develop a high level of expertise, producing a higher level of work performance. Due to positive experiences, they feel greater efficacy and competence at work, and receive more direct and indirect rewards from the organization [26,40,41]. In other words, psychological safety functions as a fundamental source of job satisfaction. Thus, the current research suggests the following hypotheses.

Hypothesis 2.
Employee psychological safety is positively related to job satisfaction.

Job Satisfaction and Organizational Performance
Job satisfaction is defined as an employee's cognitive and emotional evaluation of their job, which has either a positive or negative form [27]. It is a traditional academic belief that job satisfaction positively affects organizational performance [17,18,27,28]. Strauss [44] proposed a long-standing view that productivity is the result of employee morale or satisfaction in an organization. This perspective tends to function as the implicit basis in the works of organizational behavior.
The main point of Strauss' [44] idea is that attitudes towards something leads to actions related to the object [18,45]. When a person has a favorable attitude towards a particular object, he or she may act positively on that object. Positive attitudes, such as a sense of job satisfaction, are likely to motivate employees to embrace organizational goals and values, prompting them to work hard and achieve its goals. In this way, the quality of behavior and effort of employees increases, which, in turn, directly improves individual performance [18,[45][46][47]. This perspective is also supported by many meta-analysis works [17,28].
Meanwhile, according to social contagion theory [47,48], the employee's positive attitudes and behaviors are likely to be shared among colleagues via social contagion procedures. Through social interactions, employees may share attitudes towards a job, eventually forming a collective-level of job satisfaction. Job satisfaction at the collective level tends to promote helping behaviors and collaboration in the organization [49,50]. The increased cohesion may enhance the quality of interdependent work processes, which eventually facilitate the efficiency and performance of the organization [51,52]. Therefore, this paper expects that employee job satisfaction, when shared collectively, would improve the organizational performance of the organization.

Hypothesis 3.
Employees' job satisfaction is positively related to organizational performance.

Sequential Mediating Role of Psychological Safety and Job Satisfaction in the CSR-Performance Link
Relying on the above arguments about the associations among our research variables (i.e., CSR, psychological safety, job satisfaction, and performance), the current research suggests that the relationship between CSR and performance may be sequentially mediated by the level of employee psychological safety and job satisfaction. The associations can be explained by institutional theory [53]. This theory suggests that micro-variables (e.g., organizational member's perceptions, attitudes, and behaviors) play a mediating role in the association between macro-level variables (e.g., CSR practices and organizational performance). Scott [53] argued that institutional enablers like organizational Sustainability 2020, 12, 3514 5 of 16 practices, structures, or systems in an organization influence macro-level variables by increasing or decreasing the quality of organizational member's perceptions, attitudes, and behaviors. Therefore, this paper expects that CSR activities may positively contribute to organizational performance by improving the level of an employee's psychological safety and job satisfaction.

Hypothesis 4.
Employees' psychological safety and job satisfaction sequentially mediate the CSR-organizational performance link.

Moderating Role of Job Insecurity in CSR-Psychological Safety Link
Although our suggestion that CSR improves the level of an employee's psychological safety seems reasonable, it is somewhat naïve to argue that this relationship may always exist in all situations and contexts within an organization. Employees do not simply accept and adhere to the systems, norms, or behaviors of their organization-rather, they actively pursue meaning in their work experiences [30]. Thus, the employees may interpret the meaning and true intent of the firm's moral activities (i.e., CSR) based on their perceptions, experiences, and values. In this sense-making process, they may respond differently to moral behaviors [29]. In other words, some factors may function as a moderator in the CSR-psychological safety link.
In specific, this paper investigated the contingent role of job characteristics because the variable has been known to significantly affect organizational members' perceptions, attitudes, and behaviors [21]. Among various job characteristics, this study focuses on job insecurity, which is one of the most important job characteristics and serves as the fundamental factor of trust in their organization [31][32][33][34]. Job insecurity is defined as "worker's perception or concern about potential involuntary job loss" (p. 770, [32]). Many scholars have reported that job insecurity is one of the most harmful job stressors in an organization. For instance, job insecurity is closely associated with a low level of mental and physical well-being, job satisfaction, organizational commitment, organizational trust, creativity, and citizenship behavior [33,34,43,[54][55][56][57][58][59][60][61][62].
More specifically, the current research suggests that the positive effect of CSR on an employee's psychological safety would be moderated by the level of their job insecurity. Considering that it is one of the most fundamental responsibilities for firms to provide employees with jobs and ensure their stability at work [1][2][3], employees who feel a high level of job insecurity are likely to perceive that the organization does not fulfill its essential responsibility. When the firm attempts to achieve good things for external stakeholders without prioritizing the essential interest of internal stakeholders (i.e., job security), employees are more likely to be suspicious of the authenticity of the CSR activities, recognizing that the company's CSR activities have hidden questionable purposes. These doubts will reduce the positive impact of CSR on employee psychological safety. On the other hand, if employees feel a high level of job stability, they likely believe that the firm's moral commitment to helping external stakeholders is an authentic behavior, which originates in its inherent moral values. Then, the employees would trust in the genuine intention of moral activities. Eventually, the positive impact of CSR on their psychological safety would be amplified.
By relying on the arguments, it is possible to infer that the influence of CSR on the psychological safety of employees would depend on the level of job insecurity they perceive. Job insecurity plays a role of a critical contingent or contextual factor that determines the effectiveness of CSR activities in an organization. Thus, this paper suggests the following hypothesis. The framework of this research model is presented in Figure 1. By relying on the arguments, it is possible to infer that the influence of CSR on the psychological safety of employees would depend on the level of job insecurity they perceive. Job insecurity plays a role of a critical contingent or contextual factor that determines the effectiveness of CSR activities in an organization. Thus, this paper suggests the following hypothesis. The framework of this research model is presented in Figure 1.
Hypothesis 5: Job insecurity negatively moderates the association between CSR and psychological safety.

Participants and Procedure
To test our hypotheses, an online survey was performed by a big research company in Korea. Data for this research were gathered from South Korean employees. To deal with the issue of sampling bias, the research company selected respondents by utilizing a random sampling method. By collecting data at three different time points and from different sources (i.e., employees and directors of human resource departments in each firm), this paper attempted to complement the limitations of the cross-sectional research design. At each point, the survey was conducted with a lag of one month. At the first time point, 512 workers participants in our survey, 378 employees at time point two, and lastly 335 workers responded to our survey. The survey system was opened for three days to provide enough time for the participants to respond to it. During the time periods, the participants accessed the system anytime when they want. Then, we excluded incomplete and missing responses from the raw dataset. As a result, data from 301 responses were finally used for the final analysis (response rate: 58.79%). The characteristics of the sample are presented in Table 1.

Measures
Research variables were measured using a five-point Likert scale (with scores ranging from 1 = strongly disagree to 5 = strongly agree).

CSR (Time Point 1, Collected from Members in an Organization)
CSR activities were measured using the 12 items of Turker's CSR scale [36]. This measure takes the stakeholder approach that includes survey items about social responsibility towards a variety of stakeholders. It consists of four domains, including CSR for the environment, community, employee, and customer. To measure CSR activities for the environment, three items, including "our company participates in activities which aim to protect and improve the quality of the natural environment" were used. Second, CSR for the community was measured by three items, such as "our company contributes to campaigns and projects that promote the well-being of the society". Third, CSR for the employee was measured by three items, such as "the management of our company is primarily concerned with employees' needs and wants". Fourth, this paper utilized three items to measure CSR for the customer, including "our company respects consumer rights beyond the legal requirements". The correlations among the subcomponents of CSR were relatively high (e.g., 0.64, 0.63, 0.58, etc.), this paper utilizes the composite variable of the subcomponents to measure CSR. The items were gathered from employees at the first time point. The value of Cronbach's alpha was 0.90.

Job Insecurity (Time Point 1, Collected from Employees)
This paper used four items from the job security scale of previous work [63]. The scale includes items like "If my current organization were facing economic problems, my job would be the first to Sustainability 2020, 12, 3514 8 of 16 go," "I will not be able to keep my present job as long as I wish," "My job is not a secure one," and "My job will not be there although I want it." The value of Cronbach's alpha in this research was 0.89.

Psychological Safety (Time Point 2, Collected from Members in an Organization)
The current research utilized four items of the psychological safety scale, relying on existing research [23]. Sample items were "It is safe to take a risk in this organization" and "I am able to bring up problems and tough issues in this organization." The Cronbach's alpha value was 0.75.

Job Satisfaction (Time Point 2, Gathered from Members in an Organization)
This paper used three items of job satisfaction scale, relying on extant work [64]. Sample items were "Generally, I am satisfied with my work", "I am satisfied with what I am doing at work", and "I am satisfied with relationships at work". The Cronbach's alpha value was 0.89.

Organizational Performance (Time Point 3, Gathered from the Directors of Human Resource Departments in Each Company)
The directors of human resource departments in each company were asked to evaluate the level of organizational performance with four items. Cronbach's alpha value was = 0.92. Sample items were from an existing study [29], including "our company is more efficient and productive than our competitors," "our management performance is superior to our competitors," and "our financial performance is excellent in comparison to our competitors." By collecting data from multiple sources, we attempted to decrease the potential negative consequences of the common method bias.

Control Variables (Time Point 2, Gathered from Members in an Organization)
Considering that existing studies have reported that the size and industrial characteristics of a firm significantly affect organizational performance [65,66], this paper included control variables for organizational performance in our research model. In addition, employees' tenure (in months), position, and the level of education were included to control for job satisfaction [18,28]. The control variables were collected at time point two.

Analytical Approach
The relationships among our research variables were checked through a Pearson correlation analysis. To compute the internal consistency of the variables, we utilized Cronbach's alpha values. Based on the previous work [67], we used a two-step approach that consists of the measurement and the structural model. Then, to check the validity of the measurement model, we performed a Confirmatory Factor Analysis (CFA). A moderated mediation model analysis with SEM was performed to test the structural model with the AMOS 21 program. This paper utilized maximum likelihood (ML) estimator to conduct the SEM. In addition, to check whether our mediation hypothesis was supported, this paper implemented a bootstrapping analysis. The current research utilized the 95% bias-corrected confidence interval (CI) to test the effect of mean indirect mediation. When the CI does not have 0, we can interpret that the indirect effect was statistically meaningful with 0.05 level.
To evaluate whether the model fit is appropriate, this paper considered several goodness-of-fit indices such as the comparative fit index (CFI), the Tucker-Lewis index (TLI), and the root mean square error of approximation (RMSEA). Previous work suggested that an adequate fit is indicated by CFI and TLI values greater than 0.90 and an RMSEA less than or equal to 0.06 [68]. Finally, a bootstrapping analysis was implemented to test whether the indirect effect was significant [69].

Descriptive Statistics
Correlations among variables are provided in Table 2. Our research variables such as CSR, psychological safety, job satisfaction, organizational performance were significantly correlated.

Measurement Model
To test the adequacy of the measurement model, this paper conducted confirmatory factor analyses (CFA). We checked the discriminant validity of the four variables (i.e., CSR, job insecurity, psychological safety and job safety) with CFA of all 23 items. Our four-factor model had a very good fit to the data (χ2 (df = 80) = 161.86; CFI = 0.964; TLI = 0.953; RMSEA = 0.058; 90% CI = [0.045, 0.071]). Then, by performing a series of chi-square difference tests, this research sequentially compared our 4-factor model to alternative ones (i.e., 3, 2, and single factor model). The results demonstrated that the 4-factor model was best when compared to all the alternative models, indicating that the four variables were distinct.

Structural Model
To test our hypotheses using structural equation modeling, we made a "moderated mediation model" that includes a mediation structure that led to CSR → psychological safety → job satisfaction → organizational performance with a moderation structure that proposes that job insecurity negatively moderates the association between CSR and psychological safety. This paper computed interaction terms by multiplying CSR and job insecurity with two variables centered on their means to decrease the multicollinearity [70]. Then, the current research tested the multicollinearity bias in CSR and job insecurity. Using SPSS, we computed the variance inflation factor (VIF) and tolerance to test if CSR and job insecurity have the problem [70]. The VIF values for CSR and job insecurity were both 1.02 and 1.02, which were lower than 10. In addition, the tolerance values were both 0.98 and 0.98, which were above 0.2. The results mean that the CSR, and job insecurity did not have the bias.

The Results of Mediation Analysis
After that, this paper tested the adequacy of our mediation model by performing structural equation modeling analyses. To get the best model, the current research compared model fit indices between our hypothetical model (i.e., full mediation model) and a nested model as an alternative one (i.e., partial mediation model) by performing chi-square difference tests. Our results showed that the model fit of our hypothetical model (i.e., Model 1) were sufficient (χ 2 = 355.07 (df = 187); CFI = 0.938; TLI = 0.923; RMSEA = 0.055). Then, we compared the hypothetical model with an alternative one (i.e., Model 2). The model is identical to the hypothetical model, but having an additional one path from CSR to performance). The alternative model had good fit to the data (χ2 = 339.57 (df = 186); CFI = 0.943; TLI = 0.930; RMSEA = 0.052), the result of our chi-square difference test indicated that Model 2 had a better fit than Model 1 (∆χ2 [1] = 15.5, p < 0.001), meaning that the alternative model was better, supporting for the partial mediation.
Among the control variables (tenure, position, education level for controlling job satisfaction, firm size and industry type for controlling organizational performance), none of them were significant. With controlling for indirect effects via psychological safety and job satisfaction, CSR did significantly explain organizational performance. CSR was positively related to psychological safety (β = 0.45, p < 0.001). Both the psychological safety-job satisfaction link (β = 0.83, p < 0.001) and the job satisfaction-organizational performance link were significant (β = 0.34, p < 0.001). The results of path coefficients are presented in Figure 2.

The Result of Moderation Analysis
To check whether the moderating role of job insecurity on the association between CSR and psychological safety was significant, the current research built an interaction term between CSR and job insecurity by mean-centering and multiplying. According to the previous work [70], when the path from the interaction term to psychological safety is statistically meaningful, it indicates that job insecurity moderates the CSR-psychological safety link [70]. The value of the interaction term (β = −0.20, p < 0.01) demonstrated that the job insecurity negatively moderates the association. In other words, the association between CSR and psychological safety was weaker when the degree of job insecurity was high than the degree of it was low. The moderation effect is presented in Figure 3.

Bootstrapping
To check whether Hypothesis 4 was supported, which suggested that psychological safety and

The Result of Moderation Analysis
To check whether the moderating role of job insecurity on the association between CSR and psychological safety was significant, the current research built an interaction term between CSR and job insecurity by mean-centering and multiplying. According to the previous work [70], when the path from the interaction term to psychological safety is statistically meaningful, it indicates that job insecurity moderates the CSR-psychological safety link [70]. The value of the interaction term (β = −0.20, p < 0.01) demonstrated that the job insecurity negatively moderates the association. In other words, the association between CSR and psychological safety was weaker when the degree of job insecurity was high than the degree of it was low. The moderation effect is presented in Figure 3.

The Result of Moderation Analysis
To check whether the moderating role of job insecurity on the association between CSR and psychological safety was significant, the current research built an interaction term between CSR and job insecurity by mean-centering and multiplying. According to the previous work [70], when the path from the interaction term to psychological safety is statistically meaningful, it indicates that job insecurity moderates the CSR-psychological safety link [70]. The value of the interaction term (β = −0.20, p < 0.01) demonstrated that the job insecurity negatively moderates the association. In other words, the association between CSR and psychological safety was weaker when the degree of job insecurity was high than the degree of it was low. The moderation effect is presented in Figure 3.

Bootstrapping
To check whether Hypothesis 4 was supported, which suggested that psychological safety and job satisfaction function as sequential mediators in CSR-performance link, a bootstrapping analysis was implemented with a sample of [71]. The value of confidence interval from CSR to performance

Bootstrapping
To check whether Hypothesis 4 was supported, which suggested that psychological safety and job satisfaction function as sequential mediators in CSR-performance link, a bootstrapping analysis was implemented with a sample of [71]. The value of confidence interval from CSR to performance through psychological safety and job satisfaction did not have 0 (95% CI = [0.07, 0.23]). It means that the indirect intermediating effect of psychological safety and job satisfaction in the CSR-performance link was statistically meaningful, supporting Hypothesis 4. In addition, the current research described the result of direct, indirect, and total effects of the CSR-performance link in Table 3. Table 3. Direct, indirect, and total effects of final research model.

Model Direct Effects Indirect Effects Total Effects
CSR -> Organizational Performance 0.282 0.127 0.409 All values are standardized.

Discussion
In this paper, this paper tried to demonstrate that CSR would eventually improve organizational performance by improving the employee's psychological safety and job satisfaction. In addition, the current research attempted to verify the hypothesis that an employee's perception of job insecurity may function as a moderating factor in the CSR-psychological safety link. To empirically support the hypotheses, the current research analyzed data collected from 301 people over three points. The results of this paper showed that all hypotheses were supported.

Theoretical Implications
We believe that this research has some theoretical contributions to extant literature. First, the current research revealed that CSR improves organizational performance by enhancing employees' perceptions and attitudes (i.e., psychological safety and job satisfaction). Although there are many previous works on the CSR-performance link, there has not been a clear conclusion about the association [3,[10][11][12][13]. To clarify the relationship, the current research explored the intermediating mechanisms and the contingent factor in the link. Our results show that CSR activities enhance organizational performance by boosting the level of the employee's psychological safety and job satisfaction. Also, job insecurity functions as a contingent factor between CSR and psychological safety.
Second, this paper attempted to provide a comprehensive picture of the CSR-performance link by combining macro-and micro-perspectives. Existing studies tend to primarily focus on an externally-oriented approach at the macro-level [5,15,16,20]. To supplement the limitations of extant work, the current research examined the association between CSR and performance by focusing on internal intermediating processes at the micro-level. Specifically, based on institutional theory [53], we proposed that the impact of macro-level phenomena such as CSR activities and organizational performance) may be developed via micro-level processes including employee perception and attitude (i.e., psychological safety and job satisfaction). These findings showed that the association between CSR and performance is sequentially mediated by psychological safety and job satisfaction, supporting our hypotheses. Beyond a traditional externally-oriented approach, this paper demonstrated the essential role of internal processes between macro-level variables. The efforts of this research may positively contribute to CSR literature.
Third, this work emphasizes the importance of 'authenticity', which is reflected by the level of job insecurity by empirically demonstrating that an employee' job insecurity negatively moderates the CSR-psychological safety link. This indicates that employees may not simply believe in the firm's goodwill when their job is not secure. No matter how rarely a company performs moral activities, the positive influence of CSR on employees' perceptions and attitudes (i.e., psychological safety and job satisfaction) would be diminished when there did not exist enough authenticity in such moral actions. Whether such actions are implemented with authenticity is critical.

Practical Implications
Our findings may provide practical implications as follows. First, corporate executives or practitioners need to understand that CSR activities are not just passive and defensive acts of spending 'money' to fulfill their social obligations. From the perspective of employees, if their firm makes decisions that comprehensively consider the interests of various stakeholders, they are likely to experience greater psychological safety and job satisfaction, which eventually facilitate organizational performance. In other words, CSR practices can be understood as a kind of 'investment' to improve the performance of the firm [4,5,9,15,16]. This argument may be reasonable considering that several previous works have demonstrated that CSR activities contribute to enhancing organizational performance by improving the perceptions, attitudes, and behaviors of employees [5,15,16].
Second, the current study provides top management teams and managers of an organization with insight that they need to monitor employee levels of psychological safety and job satisfaction to check whether CSR activities have positive impacts on the employees. The results of the current research demonstrate that CSR increases organizational performance by enhancing the level of the employee's psychological safety and job satisfaction. This means that the levels of their psychological safety and job satisfaction would function as criteria to determine whether there are positive effects of corporate moral behaviors (i.e., CSR practices) in an organization.
Third, this paper may help corporate leaders to understand that all members of the organization do not always respond to the CSR activities in the same way. In other words, there are individual differences among employees. No matter how well a company performs its social responsibilities, if an employee of the firm feels a considerable level of job instability, he or she would not experience enough psychological safety through the moral behavior of the company. On the other hand, when an employee perceives a low level of job insecurity, the positive effects of CSR activities on the employee's psychological safety may be amplified due to their positive response to the moral practices.

Limitations and Suggestions for Future Studies
Despite the many theoretical and practical contributions described above, the current study has several limitations that should be properly supplemented in future studies. First, while the core values pursued by CSR practices can be accepted in both Western and Eastern societies [20,72], there are cultural differences in the employee's interpretation of CSR activities. In South Korea, where there has been rapid economic growth during the past few centuries, workers in Korean companies are likely to be relatively less sensitive to moral behaviors compared with the Western workers [20,73]. However, this study could not entirely consider the possibility of cultural impacts, as it only utilizes data from Korean companies. Therefore, we must carefully interpret the results of this study to describe CSR associated phenomena in different cultures [20,74,75].
Second, the current study could not utilize objective measures for CSR activities and organizational performance. Due to practical restrictions related to data collection, this paper could not include objective methods of measuring CSR and organizational performance (e.g., objective CSR index or content analysis of firms' annual report, and financial performance measures such as sales, operating profit, and net profit). Although existing works have suggested that the employee's subjective evaluations of certain phenomena (e.g., perceived CSR or perceived organizational performance) tend to reflect objective phenomena as precisely as the objective measures since employee perceptions are likely to critically build the realities from which employees draw their judgements [5,15,16,76], it is likely that the use of the objective CSR index produces different results with subjective measures. Therefore, future research should deal with this issue by utilizing objective CSR and organizational performance scales.
Third, this paper only utilized single-level data to test its hypotheses. Considering that not only both CSR practices and organizational performance are macro-level phenomena, but also the employee perceptions and attitudes are micro-level phenomena [1,2], this paper is limited due to its single-level approach. Instead, a multi-level analysis is required to integrate the externally-and internally-oriented perspectives. By using a multi-level approach, this paper may adequately investigate the dynamics of interactions between macro-level variables and micro-level ones [5]. Future studies should consider (and address) this issue.
Lastly, this paper could not take into account the complex nature of CSR activities. The concept of CSR not only consists of many sub-dimensions, but also is related with various stakeholders such as employees, customers, local communities, and environment [5,15,16]. Therefore, considering it as a uniform concept may have distorted the reality pertinent to CSR phenomena. For example, some recent studies have separated the concept into two different dimensions (i.e., internal CSR vs. external CSR) [16,37]. Future studies should investigate the differential effect of the two different dimensions of CSR on employees' perceptions, attitudes, and behaviors.

Conclusions
Despite its limitations, we believe that this paper provides theoretical and empirical contributions to CSR literature by revealing both the internal mediating process and the contingent factor to explain the CSR-performance link. The study demonstrated the important role of employees' perceptions (i.e., job insecurity and psychological safety) and attitude (i.e., job satisfaction) as critical mediators and moderator to describe the link. Beyond the traditional externally-oriented approach, the current research showed the essential role of internal processes between macro-level variables. Moreover, this paper integrates job insecurity research into CSR literature by emphasizing the importance of 'authenticity', which is reflected by the degree of job insecurity. Overall, we hope these efforts will be a helpful contribution to the CSR literature.