Organizational citizenship behavior (OCB) refers to individual voluntary behaviors that can promote the functioning of an organization (Organ, 1988). Considering that these behaviors are neither mandatory nor officially rewarded, why do employees still conduct OCB? Scholars have identified many antecedents of employee OCB, which can be divided into two main types: (a) organization-level antecedents, such as perceived organizational support (Thompson et al., 2020), leadership style (Abdullahi et al., 2020), and employment security (Liu et al., 2019); and (b) individual-level antecedents, such as prosocial motivation (Arshad et al., 2021) and vitality (Spanouli & Hofmans, 2021). However, whether OCB conducted by supervisors, as a concept that has been receiving recent research attention (Yaffe & Kark, 2011), relates to employee OCB remains to be seen.

In accordance with social learning theory (Bandura, 1986), individuals imitate the behaviors of others, especially significant others (e.g., supervisors). In line with this theoretical perspective, when supervisors, as agents of the organization, conduct OCB, employees will conduct similar behaviors. Therefore, our first aim was to examine the impact of supervisor OCB on employee OCB, which will enrich the literature on the antecedents of employee OCB.

We argued that the influence of supervisor OCB on employee OCB would vary depending on the level of social contact between employees and supervisors. Social learning theory holds that individuals have a greater potential to imitate the behaviors of others (e.g., supervisors) who are close to them and have authority (Bandura, 1986). Leader–member exchange (LMX), as a crucial construct in employee OCB research, describes the overall quality of the exchange relationship between supervisors and employees (Graen & Uhl-Bien, 1995). High-quality LMX entails a high degree of mutual respect, trust, liking, and support (Graen & Uhl-Bien, 1995). That is, in high-quality LMX contexts, socioemotional connections are richer and social distance is lower between employees and supervisors. In consequence, employees in high-quality LMX settings tend to imitate supervisors who conduct OCB. Thus, our second aim was to examine whether the relationship between supervisor OCB and employee OCB is moderated by LMX. We argued that employees with high-quality LMX would think of their supervisors as an example worthy of following, and this would enhance the possibility of employees imitating the example of supervisor OCB. Our findings will increase understanding of the boundary conditions under which supervisor OCB influences employee OCB and provide valuable insight into the interaction between supervisor OCB and the social environment.

The Dimensions of Organizational Citizenship Behavior

Early research suggested that OCB could be divided into five dimensions (Organ, 1988): Altruism, which entails actively providing help to others to complete work tasks; courtesy, which is actions taken at work to avoid conflicts with others; conscientiousness, which relates to involvement in the job exceeding the minimum requirements set by the organization for the role; civic virtue, which involves being constructively involved in organizational life; and sportsmanship, which means tolerating the inconvenience and pressure brought by the job rather than complaining. Building on this, Williams and Anderson (1991) later consolidated the dimensions of OCB into two types: OCB targeted toward individual coworkers (OCBI), which entails discretionary behaviors beneficial to individual coworkers, such as giving advice; and OCB targeted toward the organization (OCBO), which involves discretionary behaviors beneficial to the organization as a whole, such as abiding by the company’s norms.

Although there are differences in the dimensional division of OCB, we believe altruism and courtesy can be classified as OCBI, whereas conscientiousness, civic virtue, and sportsmanship can be classified as OCBO. Organ (1988) pointed out that the higher an individual’s position in the organizational hierarchy, the more important OCB may be. Consistent with this viewpoint, MacKenzie et al. (1999) found that compared to employees (insurance agents), supervisors (insurance managers) who performed OCB were more likely to receive positive performance evaluations. Mackenzie et al. (1999) reported that supervisor OCB and employee OCB include the same three dimensions: helping, civic virtue, and sportsmanship. However, Bachrach et al. (2007) argued that helping and civic virtue are universal and can be regarded as applicable to all members (supervisors or employees) of an organization. Meanwhile, given that our goal was to examine the impact of supervisor OCB on employee OCB, the selected OCB dimensions should apply to supervisors and employees. Therefore, the OCB dimensions we chose to examine were helping (OCBI) and civic virtue (OCBO).

Supervisors’ and Employees’ Organizational Citizenship Behavior

The Moderating Role of Leader–Member Exchange in Predicting Employee Organizational Citizenship Behavior

Method

Participants

We conducted a two-wave (separated by a 1-month interval) online survey among employees from organizations in China. We obtained 264 paired surveys by matching cell phone numbers or social media ID numbers that are unmatchable with their actual accounts, which were left by participants who completed both waves of the survey. Of these surveys, 239 were valid (rate of return = 90.53%). The participants spanned different types of work units, including 136 (56.9%) from public institutions, 48 (20.1%) from private enterprises, 23 (9.6%) from state-owned enterprises, seven (2.9%) from foreign enterprises, four (1.7%) from government departments, three (1.3%) from joint ventures, and 18 (7.5%) marked “other.” Among the respondents, 115 (48.1%) were men and 124 (51.9%) were women. There were 34 (14.2%) participants aged 25 years or under, 156 (65.3%) participants aged 26 to 30 years, 39 (16.3%) participants aged 31 to 35 years, and 10 (4.2%) participants aged over 35 years. Regarding level of education, 11 (4.6%) participants had a college degree or below, 171 (71.5%) had a bachelor’s degree, and 57 (23.8%) had a master’s degree or above. In terms of supervisor–employee tenure, 121 (50.6%) had been working together for 1 year or less, 75 (31.4%) for 1 to 3 years, 28 (11.7%) for 3 to 5 years, and 15 (6.3%) for 5 years or more.

Procedure

Data collection occurred in April and May 2019. A hyperlink to the online survey launched on the WJX survey platform was sent to the participants through WeChat social media groups. Interested people clicked the hyperlink to enter the survey. To encourage participation and obtain valid data, we emphasized three points: First, our survey was anonymous and confidential, and the data would be used only for academic research purposes. Second, participants could draw a random red envelope to earn at least CNY 1 (USD 0.14) after each survey. Participants who completed both survey waves were rewarded with at least CNY 10 (USD 1.40). Finally, we informed participants that after completion of each survey, their questionnaire would be checked within 48 hours. If their data did not meet the quality requirement, they would not obtain the red envelope.

Two different online surveys were sent to the participants, separated by a 1-month interval. Participants were invited to rate their own OCB, their supervisor’s OCB, LMX, and demographic variables at Time 1 and we obtained 423 questionnaires. Participants who completed the first survey were invited to evaluate their OCB at Time 2 and we obtained 376 questionnaires. Of these, 239 surveys were matched to provide valid paired data using cell phone numbers or WeChat ID numbers left by participants who completed the two-wave online survey. All procedures conducted in this study were based on the 1964 Declaration of Helsinki and the ethical standards of the British Psychological Society.

Measures

Supervisor Organizational Citizenship Behavior

Leader–Member Exchange

Employee Organizational Citizenship Behavior

Control Variables

Results

Correlation Analysis

Table 1 depicts the means, standard deviations, and correlation matrix among the study variables. Supervisor OCB (Time 1 [T1]) was positively related to employee OCB (Time 2 [T2]; r = .51, p < .01). LMX (T1) was positively related to supervisor OCB (T1; r = .66, p < .01) and employee OCB (T2; r = .37, p < .01). In terms of control variables, gender was negatively related to employee OCB (T2; r = .20, p < .01). Age (r = .24, p < .01), supervisor–employee tenure (r = .18, p < .01), and employee OCB (T1; r = .67, p < .01) were positively related to employee OCB (T2). Educational level was not related to employee OCB (T2; r = .11, ns).

Confirmatory Factor Analyses and Common Method Bias

We conducted a series of confirmatory factor analyses to test for common method bias as well as to assess discriminant validity among the three variables of supervisor OCB, LMX, and employee OCB. The results indicated that among all the models we examined, the hypothesized three-factor model fitted the data best, chi square (χ²) = 45.04, degrees of freedom (df) = 24, χ²/df = 1.88, p < .001; root mean square error of approximation (RMSEA) = .06, comparative fit index (CFI) = .98, standardized root mean square residual (SRMR) = .03. Chi-square difference tests revealed the three-factor model was superior to the alternative models, including a two-factor model combining supervisor OCB and LMX, Δχ²(2) = 49.56, p < .001; χ² = 94.60, df = 26, χ²/df = 3.64, p < .001; RMSEA = .11, CFI = .93, SRMR = .05, and a one-factor model combining supervisor OCB, LMX, and employee OCB, Δχ²(3) = 187.87, p < .001; χ² = 232.91, df = 27, χ²/df = 8.63, p < .001; RMSEA = .18, CFI = .80, SRMR = .08. Our results suggest that the three variables had adequate discriminant validity and common method bias was not a significant concern (Podsakoff et al., 2003).

Hypothesis Testing

In all hypothesis tests we controlled for gender, age, educational level, supervisor–employee tenure, and employee OCB (T1). Hierarchical regression analyses were utilized to test our hypotheses, both of which were supported (see Table 2). Specifically, supervisor OCB was significantly related to employee OCB (β = .14, p = .043; Model 1), supporting Hypothesis 1. Further, the positive relationship between supervisor OCB and employee OCB was moderated by LMX (β = .14, p = .002; Model 3); thus, Hypothesis 2 was supported.

The results of a simple slope analysis demonstrated that the relationship between supervisor OCB and employee OCB was not significant in the case of low-quality LMX (β = .07, p = .290) but was positive and significant in the case of high-quality LMX (β = .30, p < .001; see Figure 1).

Discussion

By analyzing the two-wave data of 239 participants from different organizations in China, we found that the relationship between supervisor OCB and employee OCB is positive and significant, and LMX positively moderates this relationship.

Theoretical Implications

Our study makes three major contributions. First, our results provide a more comprehensive understanding of the antecedents of employee OCB than was previously available. Prior studies explored the antecedents of employee OCB, including perceived organizational support (Thompson et al., 2020), leadership style (Abdullahi et al., 2020), prosocial motivation (Arshad et al., 2021), and vitality (Spanouli & Hofmans, 2021). However, no studies had examined the impact of discretionary behaviors of supervisors, such as supervisor OCB, on employee OCB. Compared to other antecedents of employee OCB, supervisor OCB is more specific and direct, which may help to enhance predictive validity and practical relevance. Drawing on social learning theory, we found that supervisor OCB can facilitate employee OCB, and the impact is stronger under the condition of high-quality LMX than under low-quality LMX.

Second, this study improves understanding of the boundary conditions that can change the impact of supervisor OCB. Past research found that leader proximity and group belief that the supervisor is a highly worthy role model enhance the impact of supervisor OCB on group-level OCB (Yaffe & Kark, 2011). Extending past studies, we found that the impact of supervisor OCB on employee OCB is significant when LMX is high, but not when LMX is low.

Finally, this study extends the literature on employee OCB by examining the role of LMX and its impact on individual-level behavior. Understanding of employee OCB is incomplete if we do not consider the social and psychological environment in which such behavior operates (Lai et al., 2013). We found that in the case of high-quality LMX, supervisor OCB is more likely to drive employees to conduct similar OCB. In contrast, supervisor OCB is not significantly related to employee OCB under the condition of low-quality LMX. Hence, supervisor OCB does not operate in a vacuum and its impact can be changed by the social environment in which employees exist (e.g., LMX).

Practical Implications

Our research indicates that employees imitate supervisor OCB, but the degree of imitation varies significantly depending on LMX quality. We recommend focusing on three aspects of human resource management practice to promote employee OCB: First, through leadership training and development projects, the example effect of supervisor OCB can be emphasized and supervisors can be encouraged to grasp this chance to perform more OCB. Cultivating the awareness and ability of supervisors to build high-quality exchange relationships with employees will enhance the positive impact of supervisor OCB on employee OCB. Second, when selecting supervisors, organizations should consider the tendency of candidates to conduct OCB (Bolino & Turnley, 2003) and select supervisors who are more likely to conduct this behavior, thus enhancing their example effect on employees.

Finally, organizations should establish a long-term commitment with supervisors and give them job autonomy. In this way, supervisors will be more motivated to establish a long-term exchange relationship with the organization. Consequently, they may conduct more OCB (Morrison, 1996), thus increasing the probability that employees will emulate this behavior.

Limitations and Future Research Directions

Our research is subject to limitations. Our findings may have been affected by common method bias because the key variables (i.e., supervisor OCB, LMX, and employee OCB) in this study were all assessed by employees themselves (Podsakoff et al., 2003). To mitigate this concern, two-wave (1 month apart) data were collected and the baseline level of the outcome variable (i.e., employee OCB) was controlled for. Through a series of confirmatory factor analyses, we found that there was adequate discriminant validity among the key variables, indicating that common method variance was not a significant concern (Podsakoff et al., 2003). Nevertheless, future researchers could collect data from multiple sources to further test the proposed hypotheses.

Furthermore, we did not investigate mediators that may explain the relationship between supervisor OCB and employee OCB. Supervisor OCB, as a variety of discretionary behavior conducted by supervisors to benefit others and the organization, may make employees believe that supervisors are good people who are willing to contribute and have a positive attitude toward the organization, as well as increasing their perception that the organization has a promising future, thereby enhancing employees’ organizational commitment. Meanwhile, past research has found that there is a positive relationship between employee organizational commitment and employee OCB (Williams & Anderson, 1991). Hence, future researchers could explore whether employee organizational commitment mediates the relationship between supervisor OCB and employee OCB.

In conclusion, we hope that our findings will encourage scholars to investigate the model effects of supervisors and to synchronously explore further moderators that may affect the relationship of supervisor OCB with other outcomes of interest.

References

This study was supported by the Humanities and Social Sciences Research Project (SK2021A0505) of Anhui Universities.

Congyong Shang, School of Economics and Management, Hefei Normal University, No. 1688, FuRong Road, Shushan District, Hefei, Anhui 230601, People’s Republic of China. Email: [email protected]