You may not reap what you sow: How and when ethical leadership promotes subordinates’ online helping behavior

Abstract

Recently, home office and remote working have gained momentum triggering questions of the impact of ethical leadership on helping behavior and its ethical implication for employees and companies. In this study, we propose a moderated mediation model to clarify the mechanism of ethical leadership on online helping behavior, and tested this model using three-wave data from a sample of 481 employees in mainland China. We found that ethical leadership had a positive effect on subordinates’ other-praising moral emotion, and that these moral emotions significantly influenced online helping behavior. Subordinates’ other-praising moral emotion mediated the relationship between ethical leadership and online helping behavior. Moreover, the relationship between ethical leadership and other-praising moral emotion was more significant for employees with high moral identity. Finally, our findings provide new insights into how to motivate employees’ online helping behaviors through managerial practices for organizations.

With the rapid development of the Internet, online, social networks have become an important platform for daily communication. A significant amount of human interaction occurs online, and employees usually turn to the Internet for their colleagues’ help (Bothma et al., 2018; Wang & Wang, 2008). During the period of COVID-19, networks have become the predominant and almost exclusive way for employees to help each other. Unfortunately, communication in this way is usually inefficient as e-mail or WeChat recipients often disregard the received messages unintentionally. Numerous studies have examined online helping behavior from multiple perspectives, such as social loafing, the bystander effect, the diffusion of responsibility and network density (Blair et al., 2005; Lewis et al., 2004; Guéguen et al., 2005; Kuchmaner et al., 2019). These studies, however, have neglected an important fact that whether individuals are willing to show helping behaviors is closely related to their ethical characteristics (Tang et al., 2008; Clercq et al., 2019). Recently, organizational behavior scholar proposed that “helping behaviors targeted at coworkers are intrinsically connected to workplace ethics, and employees perceived as more ethical tend to engage in helping behaviors to a greater extent than those who are perceived as less ethical” (Clercq et al., 2019, p.1169). Therefore, the tendency of employees to go out of their way (e.g., online) to help other members constitutes a value-based phenomenon that relates closely to employee’s own individual ethical beliefs (Deckop et al., 2003; Turnipseed, 2002).

Social learning theory suggests that individuals tend to learn and to emulate their leaders’ behaviors (Men et al., 2020; Wang et al., 2018; Mayer et al., 2012). Ethical leaders’ proactive communication about what is (un-)ethical behavior, coupled with such moral characteristics as caring, honesty and integrity presented in leaders’ daily work-life, gives employees a model of what is (in-)appropriate behavior at work (Men et al., 2020). Accordingly, it is not surprising that employees rely on their leaders for guidance when faced with ethical questions or problems (Brown & Mitchell, 2010). Research supports this contention and shows that leaders with high morality are more inclined to establish high ethical standards, and stimulate subordinates’ ethical behaviors while inhibiting their counterproductive behaviors (Brown & Treviño, 2006; Paterson & Huang 2019; Gerpott et al., 2019; Velez & Neves, 2018; Mo & Shi, 2018; Zhu et al., 2015). Drawing insights from social learning theory, we explore the influence of ethical leadership on employees’ online helping behavior.

In addition, this study also explores the mechanism of ethical leadership on online helping behavior from the perspective of emotion. When leaders show ethical characteristics such as honesty, caring others and integrity (Resick et al., 2011), followers will gratify their leaders’ solicitude and admire their leaders’ honesty and integrity, thereby presenting other-praising emotions (Greenbaum et al., 2019; Haidt, 2003) considered that other-praising emotion has prosocial attribute, that is, other-praising emotion enables to stimulate individuals’ willingness to help others and promote their helping behaviors. Other-praising moral emotion, described as moral emotions arising from others’ exemplary actions (Algoe & Haidt, 2009), is an important intervening mechanism linking leadership and online helping behavior. Some studies stated that other-praising emotions, including compassion, empathy, and gratitude, are the bright side of moral emotions and motivate individuals to be engaged in admirable and respectful deeds in their relationships with others (Tangney et al., 2007; Kim & Johnson, 2013). This in turn involves those behaviors that help others by advancing their well-being (Dasborough et al., 2020; Janoff-Bulman et al., 2009). Therefore, we argue that ethical leadership has implications for the other-praising moral emotion of employees, which associates with online helping behavior.

In addition, we extend our model of ethical leadership and online helping behavior by identifying a key boundary condition of our presumed causal sequence. From the perspective of employee’ s self-concept, employees’ replies to the questions like “who am I” questions are crucial for the in-depth understanding of prosocial behavior (e.g., online helping behavior) (Gerpott et al., 2019; Thoits, 1992). Scholars considered that ethical identity, as a self-concept in leader-follower relationships, can reflect employees’ recognition of their leaders’ ethical values (Sanders et al., 2018; Gerpott et al., 2019). In other words, employees’ self-attribution regarding ethics can better help to capture their leaders’ ethical signals (Rudolph & Tscharaktschiew, 2014). On the one hand, ethical leaders may make use of visible signals to strengthen employees’ recurrence of moral value such as moral judgements facing the dilemma of business and moral objectives. Employees with high moral identity are able to make swift emotional responses to those visible signals, and apply them into their work (Dasborough et al., 2020). On the other hand, ethical leaders may also influence their followers’ other-praising moral emotions via low-key ethical allusions in their daily work (Eisenbeiss & Knippenberg, 2015). Employees with high moral identity are more inclined to internalize those moral signals (Skubinn & Herzog, 2016; Hogg, 2001), and regard those signals as moral guidance in work. Therefore, we suggest that moral identity may probably act as a boundary condition of the indirect relationship between ethical leadership and online helping behaviors via other-praising moral emotion. Figure 1 depicts the hypothesized research model of this study.

Fig. 1.

Research model

Theoretical background and hypotheses

Ethical leadership and subordinates’ other-praising moral emotion

Ethical leadership has been defined as a set of individual behaviors conforming to ethical norms conducted by leaders in the process of leader-subordinate communication, and leaders generalize these behaviors to their subordinates through a two-way communication, reinforcement and decision-making (Brown et al., 2005; Mo & Shi, 2018; Li et al., 2020). Ethical leaders who show high moral characters and moral standards can nurture followers’ emotional identity (Eisenbeiss & Knippenberg, 2015; Velez & Neves, 2018), and thereby motivate followers’ other-praising emotions. We expect that ethical leadership can promote followers’ other-praising emotions based on the following considerations. First, ethical leaders gain their legitimacy by showing suitable moral behaviors (Resick et al., 2011; Men et al., 2020), such as honesty, justice, and solicitude. They not only focus on ethics, but also perform specific behaviors to highlight ethical characteristics (communication with followers about ethical issues), and encourage followers to express their own opinions freely (Chamberlin et al., 2017; Mo & Shi, 2018; Zhu et al., 2015). In the meanwhile, they set moral standards and award those ethical behaviors (Wang & Xu, 2019; Cheng et al., 2019; Qin et al., 2018). Social learning theory emphasizes that the moral signals from ethical leader impact an individual’s cognition (Men et al., 2020; Wang et al., 2018; Mayer et al., 2012), thereby stimulating their other-praising emotions (Brown & Mitchell, 2010), and thus make followers imitate those moral behaviors. Previous evidence has confirmed that employees can internalize their leaders’ moral standards and norms as their own, which can result in employees’ other-praising moral emotions (Skubinn & Herzog, 2016; Algoe & Haidt, 2009). Second, ethical leaders integrate ethics into decision-making processes, including considering the moral consequences of decision-making and making just decisions (Kuenzi et al., 2019; Gerpott et al., 2019), which is conducive to employees’ other-praising emotions (Eisenbeiss & Knippenberg, 2015; Velez & Neves, 2018, p.185) stated that “the experience of moral emotions can change the way in which followers perceive and process the information conveyed by ethical leaders and thereby determine followers’ behavioral responses to ethical leadership”. Brown (2010) believed that ethical leadership is related to followers’ other-praising moral emotions while unethical leadership is associated with followers’ other-condemning moral emotions. Similarly, some scholars emphasize that emotion has spotlight function, and it enables leaders to center on moral characteristics when making moral decisions, and further inspires employees to show other-praising moral emotions (Aquino et al., 2011; Lindebaum et al., 2017). Hence, we hypothesize:

Hypothesis 1

Ethical leadership will be positively related to subordinates’ other-praising moral emotion.

Subordinates’ other-praising moral emotion as a mediator

Brown and Mitchell (2010, p.591) summarized three influences of ethical leadership on employee behaviors: identity/identification, fit/congruence and emotion. The first two influences belong to cognitive category while the third one relates to emotional category. Previous research explored the impact of ethical leadership on their followers from the perspective of cognitive category, for instance, psychological ownership (Avey & Palanski, 2012), moral efficacy (Lee et al., 2017), and trust (Xu et al., 2016). However, moral emotions play an equally important role in explaining ethical behavior as does moral cognition (Dasborough et al., 2020). Other-praising moral emotion, described as moral emotions arising from others’ exemplary actions (Algoe & Haidt, 2009), is the bright side of moral emotions. It can motivate individuals to be engaged in admirable and respectful deeds in their relationships with others (Tangney et al., 2007; Kim & Johnson, 2013), and involves those behaviors that help others by advancing their well-being (Dasborough et al., 2020; Janoff-Bulman et al., 2009). Thus, we expect that other-praising moral emotion can bridge the linkage of ethical leadership and online helping behavior.

Previous research has indicated that other-praising emotion has prosocial attribute, and it enables to motivate individuals’ willingness of help others, and further conduct helping behaviors (Haidt, 2003; Eisenbeiss & Knippenberg, 2015; Rudolph & Tscharaktschiew, 2014; Dasborough et al., 2020, p.437) referred to other-praising moral emotions as “prescriptive moral emotions”. Janoff-Bulman et al. (2009, p.523) stated that “prescriptive morality involves those behaviors that help others by relieving their suffering or advancing their well-being”. These emotions have prosocial action tendencies (e.g., online helping behavior) and play central roles in positive interactions (Atkins & Parker, 2012; Dutton et al., 2014; Haidt, 2003; Lim & Desteno, 2016). Social learning theory stresses that subordinates can learn from their leaders’ ethical behaviors, and have high affective recognition on the morality of their behaviors and decisions, thereby presenting ethical behaviors and decisions (Brown & Trevino, 2006; Wang et al., 2018). More specifically, when leaders show ethical characteristics such as honesty, caring others and integrity (Resick et al., 2011), followers will gratify their leaders’ solicitude and admire their leaders’ honesty and integrity, thereby presenting other-praising emotions (e.g., empathy and gratitude) (Greenbaum et al., 2019). Employees with strong empathy are inclined to help others through electronic communication instruments or other ways (Barsade & O’Neill, 2014; Dasborough et al., 2020), and employees with gratitude have a tendency to repay others kindness by helping those in need (Bartlett & DeSteno, 2006; McCullough et al., 2001) considered that gratitude served as a moral re-enforcer function, and it could drive the recipients to conduct moral behaviors in return for the present grace. In a similar vein, prior research suggested that gratitude was able to induce employees’ helping behaviors (Wood et al., 2011), and the relationship between them was closely associated with employees’ central position in the social network (Chang et al., 2012). This logical analysis leads us to hypothesize:

Hypothesis 2

Subordinates’ other-praising moral emotion positively relates to online helping behavior.

Hypothesis 3

Subordinates’ other-praising moral emotion mediates the relationship between ethical leadership and online helping behavior.

Moral identity as a moderator

Moral identity, as an individual’s self-judgement, reflects the recognition and acceptance of a variety of norms in organizational moral system, and is defined as a relatively stable self-concept formed by a series of moral characteristics (Gerpott et al., 2019; Van Quaquebeke et al., 2019). Referring to social identity theory, moral identity includes the following three processes: (1) Internalization process. Kalshoven’s et al. (2016) and Mayer’s et al. (2009) illustrated that leaders’ ethical behaviors can sway employees’ realization about the importance of moral values, and shape explicit ethical implications, standards and supervisions. As a result, employees view themselves as moral persons and internalize the ethical information. (2) Identification progress. Employees will alternatively internalize leaders’ information as their own ethical norms while taking proactive measures to rectify or improve those ethical standards not conforming to their own ethical norms (Van Quaquebeke et al., 2019). (3) Comparison process. During this process, employees will continue moral comparison against their leaders and form organizational identification after establishing ethical norms (Demirtas et al., 2017; Tseng & Wu, 2017). Similarly, Weaver and Agle (2002) found that individuals with high moral identity have a tendency to believe that moral values are crucial to define their own personal identity, and therefore often exhibit more pro-organizational behaviors.

Researches show that subordinates’ moral identity can activate other-praising moral emotion (Eisenbeiss & Knippenberg, 2015; Wright et al., 2017). In other words, other-praising moral emotions may vary in contexts with different levels of moral identity. Drawing on social identity theory, employees’ internalization, identity and comparison process of moral characteristics constitute the considerations of whether employees regard those moral characteristics as their own behavioral motive (Skubinn & Herzog, 2016; Hogg, 2001). Specifically, when employees have a strong moral identity, they are more inclined to internalize leaders’ moral characteristics as their own behavioral motivation in the workplace (Gerpott et al., 2019; Kuenzi et al., 2020). This can improve employees’ own moral standards and requirements, and thus lead employees to generate other-praising moral emotions in response to the surrounding moral events (e.g., the help or suffering of others) (Dasborough et al., 2020; Eisenbeiss & Knippenberg, 2015; Brown & Mitchell, 2010). Numerous studies have shown that other-praising moral emotions can be helpful for employees’ helping behaviors (Kim et al., 2018; Dasborough et al., 2020; Eisenbeiss & Knippenberg, 2015; Miller et al., 2012). When employees have weak moral identity, they are more likely to consider leaders’ moral behaviors as a management approach, and thus cannot resonate with these moral behaviors. This impairs employees’ other-praising emotions, and employees are unlikely to show helping behaviors. Gerpott et al. (2019) considered moral identity as employees’ moral expectations or beliefs. They found that employees with high moral identity can firmly internalize their leaders’ moral standards as their own work norms, and practice these norms in every aspect of their work. Besides, research on moral emotions has indicated that other-praising moral emotions are derived from the interaction of individual’s cognition and moral events such as leaders’ moral behaviors (Dasborough et al., 2020; Eisenbeiss & Knippenberg, 2015). Particularly, in a positive social cognition context, leaders’ moral behaviors are more likely to accelerate employees’ internalization process of moral beliefs, and thus become the moral guidance of employees’ behaviors (Hogg, 2001; Skubinn & Herzog, 2016). This discussion leads to our final hypothesis:

Hypothesis 4

The relationship between ethical leadership and subordinates’ other-praising moral emotion will be moderated by moral identity, such that the relationship between ethical leadership and other-praising moral emotion will be stronger for employees with high moral identity than for those with low moral identity, thereby promoting online helping behavior.

Methods

Sample and procedures

Data were collected from R&D and production departments of large manufacturing enterprises in Beijing, Shanghai and Shandong, China. These listed enterprises are the leading ones in their industries with outstanding technologies and market advantage, and thus they are extremely representative manufacturing enterprises. One of the authors is a senior manager, and responsible for the communication with HRM departments of the targeted enterprises regarding the survey time and site, which guarantees more employees to participate in our survey. We handed out our questionnaires on site and only responses collected on the spot were used in our data analyses. Due to the concealed characteristics of online helping behavior, and the information in electronic communication tools may involve the personal privacy of employees, it is not easy for the superiors to observe online helping behavior. Therefore, in this study, employees were responsible to assess online helping behavior. Researchers only considered response which possessed following two criteria: (1) all items were answered; (2) replies passed polygraph items in our questionnaires. If responses cannot meet any criteria, we will ask the participants to refill our questionnaires. After communicating with the Human Resource department heads of the targeted enterprises, we randomly assigned 1,255 employees with a unique number to match the coded questionnaires.

To reduce potential common method bias, we collected data in three phases at monthly intervals (Li et al., 2021). In phase 1, the participants reported ethical leadership and demographic information. In phase 2, the participants assessed their own other-praising moral emotion and moral identity. After approximately one month, in phase 3, we asked them about online helping behavior. Participation was voluntary, and we guaranteed responses be confidential and exclusively used in this study. We gathered data in three phases at monthly intervals based on the following considerations: (1) ethical leadership is a rather stable behavioral tendency; (2) other-praising moral emotions evoked by ethical leadership would be stable, or consistent across different time-points as well; (3) measuring the main variables of this study with data from different time-points not only can reduce the possibility of the CMV, but also can strengthen the causal inferences. In addition, we introduced a common method factor into our model, and found that the goodness of model fit cannot improve (Williams & McGonagle, 2016). Therefore, there is no serious common method bias in our study.

In phase 1, we gathered 1,062 questionnaires, having a response rate of 84.62%. In phase 2, we asked the participants who had finished the first questionnaires and provided consent to complete the second questionnaires. In this phase, 847 participants submitted their responses, a response rate of 79.76%. Around one month later, in phase 3, of 847 surveys were sent out, and 532 surveys were returned, making a response rate of 62.81%. After a series of strict screening procedures, such as removing the questionnaires with incomplete responses, the final sample used in our study comprised 481 complete responses. Their demographic characteristics were as follows: 58.2% of the participants were male; age distribution was as follows: 30 years or below (9.1%), 31 to 45 years old (41.4%), and 46 to 60 years old (49.5%); for participants’ education, 51.5% of them had a bachelor’s or higher degree.

Measurements

The surveys were designed to capture the four concepts investigated in this research: ethical leadership, moral identity, subordinates’ other-praising moral emotion and online helping behavior. All items were measured on a five-point Likert scale where 1 = ‘strongly disagree’ and 5 = ‘strongly agree’. The English versions of all the scales were translated into Chinese by researchers qualified in the field of organizational behavior and then back translated to English by Chinese native speakers to ensure its content reliability. To measure the extent of ethical leadership, we used the ten-item scale from Brown et al. (2005). A sample item is “my supervisor defines success not just by results but also the way that they are obtained”. The Cronbach’s alpha coefficient was .924. Moral identity was measured following the ten items from Aquino and Reed (2002). Above the questionnaire items, we gave clear indication that “Listed below are some characteristics (such as caring, compassionate, fair, friendly, generous, hardworking, helpful, honest and kind) that may describe a person. The person with these characteristics could be you or it could be someone else. For a moment, visualize in your mind the kind of person who has these characteristics. Imagine how that person would think, feel, and act. When you have a clear image of what this person would be like, answer the following questions”. Sample items are “I am actively involved in activities that communicate to others that I have these characteristics” and “it would make me feel good to be a person who has these characteristics”. The Cronbach’s alpha coefficient was .906. We adopted the evaluation method of affect developed by Brunstein (1993) and Brown and Mitchell (2010) to measure the subordinates’ other-praising moral emotion, including compassion, empathy and gratitude. Participants indicated the extent to which they had felt these moods during “the past few days”. The Cronbach’s alpha coefficient was .853. Based on the four-item scale developed by Tang et al. (2008), we removed the item regarding assisting supervisor (I assist supervisor with his or her work), and retained the other three items, which are “I help my colleagues who have heavy workloads online (e.g. via WeChat, blog or e-mail)”, “I orient new colleagues even though it is not required at work by electronic tools” and “I help my colleagues who are off line”. The Cronbach’s alpha coefficient was .857. Following previous research, we controlled four demographic variables (age, gender, educational level and job tenure) which found to relate to ethical leadership, moral identity, other-praising moral emotion and online helping behaviors (Tang et al., 2008; Eisenbeiss & Knippenberg, 2015).

Confirmatory factor analyses

To examine the discriminant validity of our measures, we conducted confirmatory factor analyses. As shown in Table 1, the four-factor model (ethical leadership, moral identity, subordinates’ other-praising moral emotion and online helping behavior) fits data well (χ² = 1288.32, df = 293, RMSEA = .095, CFI = .96, NFI = .95), yielded a better fit to data than the other models. The CFA results also provide support for the distinctiveness of the four study variables for subsequent analyses.

Table 1.

Results of confirmatory factor analyses

Model	χ 2	df	χ 2/ df	RMSEA	CFI	NFI
Four-factor model: The Measurement Model	1288.32	293	4.396	.095	.96	.95
Three-factor model 1: EL and ME combined	1755.36	296	5.930	.112	.95	.94
Three-factor model 2: EL and MI combined	1710.12	296	5.777	.113	.95	.94
Three-factor model 3: ME and MI combined	1758.08	296	5.939	.112	.95	.94
Two-factor model: EL, ME and MI combined	2163.28	298	7.259	.127	.94	.93
Single-factor model: EL, ME, MI and OHB combined	2341.22	299	7.830	.132	.93	.92

n = 481. RMSEA: Root Mean Square Error of Approximation, CFI: Comparative Fit Index, NFI: Normed Fit Index. EL: Ethical Leadership, ME: subordinates’ other-praising moral emotion, MI: Moral Identity, OHB: Online Helping Behavior.

Descriptive statistics

Table 2 displays the means, standard deviations and correlations among the four variables range from .470 to .755 (all p < .01). Given the validity of data, we calculated the square roots of the average variance extracted (AVE) values of each variable (see the brackets on diagonal line in Table 2 and Appendix I). As shown in Table 2, the square roots of AVE of each variable is greater than its corresponding correlation coefficients, indicating a good discriminant validity.

Table 2.

Means, standard deviations and correlations

Individual level	Mean	SD	1	2	3	4	5	6	7	8
1.Follower gender	1.417	.493
2.Follower age	2.403	.651	.032
3.Follower education	2.590	.814	-.118**	.233**
4.Job tenure	2.083	.560	-.051	.490**	.152**
5.Ethical leadership	3.922	.710	.029	.106*	.139**	.109*	(.773)
6.Moral identity	3.730	.659	.003	.095*	.194**	.090*	.755**	(.773)
7.Other-praising emotion	3.381	.922	.058	.045	.046	-.005	.489**	.470**	(.880)
8.Online helping behavior	3.704	.862	.083	.057	.136**	.048	.682**	.712**	.488**	(.883)

n = 481. ^*p < .05, ^**p < .01. Values in parentheses represent the square root of AVE for each variable.

Hypothesis testing

We used hierarchical regression analysis to test the main effect. First, as shown in Table 3, combining with model 1, the results in model 2 suggested that ethical leadership has a positive effect on other-praising moral emotion (β = .494, p < .01), supporting Hypothesis. Model 4 indicated that other-praising moral emotion positively affect online helping behavior (β = .479, p < .01), and thus Hypothesis 2 was supported. Second, we turn our attention to the mediation effect stipulating that other-praising moral emotion will mediate the relationship between ethical leadership and online helping behavior. As Model 5 depicted, the other-praising moral emotion has a positive impact on online helping behavior (β = .200, p < .01) when controlling ethical leadership. The results verify Hypothesis 3.

Table 3.

Results of hierarchical regression analysis

Variables	Other-praising emotion		Online helping behavior
	Model1	Model2	Model3	Model4	Model5
Follower gender	.060	.037	.070*	.072	.062
Follower age	.048	.028	-.021	-.017	-.026
Follower education	.047	-.015	.058	.120**	.061
Job tenure	-.033	-.069	-.021	.044	-.007
Ethical leadership		.494**	.676**		.577**
Other-praising moral emotion				.479**	.200**
R2	.008	.245	.473	.257	.503
F	.949**	30.798**	85.193**	32.898**	79.983**

^*p < .05, ^**p < .01.

Following the approach outlined by Preacher et al. (2007), we used IBM SPSS Statistics 23 in conjunction with the PROCESS macro (model 4 in PROCESS, 95% confidence interval, sample = 5000) (Hayes, 2013) to further test Hypothesis 3. Table 4 displays the results of bootstrapping analysis. The direct effect of ethical leadership on online helping behavior is significant (β = .408) with 95% CI [.297, .519], and the indirect effect of subordinates’ other-praising moral emotion is significant (β = .420) with 95% CI [.322, .522], excluding zero. This suggested that subordinates’ other-praising moral emotion mediates the relationship between ethical leadership and online helping behavior.

Table 4.

The test of mediating effect of other-praising moral emotion

ediator	Sobel test	Effects	coeff	se	95%CI
					LLCI	ULCI
Other-praising moral emotion	9.172**	Indirect effect	.420	.052	.322	.522
		Direct effect	.408	.057	.297	.519
		Total effect	.828	.041	.748	.908

To test moderated mediation effects where moral identity moderates the indirect effect of ethical leadership on online helping behavior via subordinates’ other-praising moral emotion (hypothesis 4), we used Model 7 (95% confidence interval, sample = 5000) of the SPSS macro developed by Hayes (2013). Table 5 illustrates these moderated indirect effects through changes in the level of moral identity. Under high moral identity, the effect of ethical leadership on online helping behavior via other-praising moral emotion is .071, with 95% CI [.041, .117], excluding 0. While under low moral identity, the effect of ethical leadership on online helping behavior via other-praising moral emotion is .100, with 95% CI [.056, .162], excluding 0. In addition, the INDEX values of moral identity are significant, which means significant differences between high and low moral identity. These support our argument (H4) that moral identity moderates the mediation effects of subordinates’ other-praising moral emotion.

Table 5.

Bootstrapping of moderated mediation

Mediator	Conditional indirect effect					Index of moderated mediation
	Moderator (moral identity)	effects	se.	95%  confidence interval		INDEX	se	95%  confidence interval
				LLCI	ULCI			LLCI	ULCI
Other-praising moral emotion	High	.071	.019	.041	.117	.022	.013	.003	.055
	Low	.100	.027	.056	.162

To further investigate the moderation effect, we derived the “regions of significance” for the conditional effect of ethical leadership on online helping behavior using the Johnson–Neyman technique in SPSS macro proposed by Preacher et al. (2007, p.213). Figure 2 plots the conditional effect of ethical leadership on online helping behavior across distributions of moral identity as a solid line as well as the lower and upper bounds of a 95% confidence interval as dashed lines. It can be seen from Fig. 2 that dashed lines confidence band has no intersection with the horizontal axis, showing that the indirect effect was found to be significant for any value of moral identity. Therefore, Hypothesis 4 was further supported.

Fig. 2.

Conditional indirect effect for moral identity

Discussion

Online helping behavior has become an increasingly important topic to organizational scholars and practitioners alike, given internet is a significant tool for daily work communication. In times of COVID-19 online communication has become the only way for employees to interact with co-workers. Unfortunately, replies are not guaranteed, as e-mail or WeChat recipients often disregard the messages they receive. Sociological studies attempted to shed light on the mechanism of online helping behavior from the perspective of social loafing, the bystander effect, the diffusion of responsibility and network density (Blair et al., 2005; Lewis et al., 2004; Guéguen et al., 2005; Kuchmaner et al., 2019). However, these studies have limited practical implications for organizational management. Recently, scholars argued that there is a basic hypothesis behind helping behavior, that is, employees’ ethics is a triggering factor of helping behavior (Tang et al., 2008; Clercq et al., 2019). According to social learning theory and leadership behavior research suggested that employees’ ethical characteristics are largely affected by ethical leader (Men et al., 2020; Wang et al., 2018; Mayer et al., 2012). Thus, the purpose of this study is to examine when and how ethical leadership influences online helping behavior.

To achieve this purpose, we constructed a moderated mediation model. Drawing on social learning theory, we examined the relationships among ethical leadership, other-praising moral emotion, moral identity, and online helping behavior. The present study echoes Brown and Mitchell’s (2010) work, by breaking the limitation of identity/identification and fit/congruence, analyzes the effect of ethical leadership on online helping behavior from the emotion’s perspective. This study extends the influential mechanism of ethical leadership, and provides a new perspective to understand the moral motivation of employees’ prosocial behavior. In addition, instead of focusing on helping behavior, this study introduces online helping behavior of sociology (Blair et al., 2005; Lewis et al., 2004; Guéguen et al., 2005) into organizational management research considering the development of network. This is an extension for organizational management research, and makes contributions to the integration of disciplines.

Theoretical implications

The findings of our study contribute to the ethical leadership literature in three ways. First, we contribute the online helping behavior literature by introducing ethical leadership. With the application of network communication tools at work, a lot of helping behaviors among co-workers occur online (Wang & Wang, 2008). More often, communication through e-mail or WeChat is regarded as low efficiency because the messages are easily to be ignored by recipients inadvertently. Sociological literature primarily focuses on the antecedents of online helping behavior from the aspects of social loafing, the bystander effect, the diffusion of responsibility and network density (Blair et al., 2005; Lewis et al., 2004; Guéguen et al., 2005; Kuchmaner et al., 2019). However, to our knowledge there remains limited knowledge about how to enhance online helping behavior within a firm. Clercq et al. (2019, p.1169) proposed that “helping behaviors targeted at coworkers are intrinsically connected to workplace ethics, and employees perceived as more ethical tend to engage in helping behaviors to a greater extent than those who are perceived as less ethical”. It shows an implicit hypothesis that personal values and ethics are the prerequisite of this kind of behaviors (Tang et al., 2008; Clercq et al., 2019). A great number of studies on ethical leadership found that individuals’ ethical characteristics are derived from the imitation of the behaviors of their ethical leader. Hence, we addressed this research gap by incorporating ethical leadership into the research regarding online helping behavior, and found that ethical leadership can significantly enhance online helping behavior. On the one hand, this advances the understanding of helping behavior from off-line to online. On the other hand, this provides the ethical perspective to expound the trigging mechanism of online helping behavior, expanding online helping behavior research from sociology to organizational management. Second, by emphasizing the role of other-praising moral emotion, we identify other-praising moral emotion as an important mechanism through which ethical leadership translates to online helping behavior, advancing our understanding of the relationship between ethical leadership and online helping behavior. Brown and Mitchell (2010) stated that cognitive category (identity and matching) and affective category (emotion) are primary mechanisms of ethical leadership influencing employees’ behaviors. However, most of existing research examines the impact of ethical leadership on employees’ behaviors from the perspective of cognitive category, such as psychological ownership (Avey & Palanski, 2012), moral efficacy (Lee et al., 2017), and trust (Xu et al., 2016), and overlooks the significance of emotion. To fill up this research gap, we analyze the effect of other-suffering or -praising emotions on the relationship between ethical leadership and online helping behavior, extending ethical leadership research into affective category (e.g., emotion).

Finally, our study demonstrates that the influence of ethical leadership on online helping behavior is contingent on moral identity. Although extant studies have extensively explored the link between ethical leadership and moral identity (Qin et al., 2018; Li et al., 2020; Van Quaquebeke et al., 2019), most of them treat moral identity as a direct consequence of ethical leadership, without considering the differences of moral identity among individuals. Nevertheless, Mesdaghinia et al.’s (2019) work indicated that the perceptions of leaders’ ethical behaviors vary among employees with different degree of moral identity. By highlighting the contingent role of moral identity, we found that in contrast with employees with low moral identity, the other-praising moral emotions of employees with high moral identity are more likely to be subject to ethical leadership. This enriches both the literature on ethical leadership and the literature on moral identity, and provides a novel reference for future work on the influence of ethical leadership behaviors on employees’ emotions and behaviors.

Practical implications

The present study also offers some managerial insights for the organization. First, we have displayed the positive mechanism of ethical leadership on employees’ online helping behaviors. From a leader’s point of view, on one hand, they should promote employees’ online helping behaviors by establishing high moral standards and maintaining high moral images in their daily management work. On the other hand, leaders should endeavor to moderately motivate employees’ other-praising moral emotions, and thus catalyze employees’ online helping behaviors, especially for those employees involving in cross-organizational and cross-regional collaborations. More basically, the organization should put more efforts into cultivating ethical leaders, for example, inculcating the benefits of ethics through training items. Second, in the respect of subordinates, individuals with high moral identity should be given priority for employment, and those employees are the optimal candidates for help. Additionally, ethical leadership are more likely to induce the emotion changes of high moral identity employees, such as other-praising moral emotion. Thus, the organization could make good use of this to advocate online helping behaviors by creating a positive emotional climate and a culture of kindness.

Limitations and future research directions

Despite these contributions, there are some limitations. First, our study examines the mechanism of ethical leadership on online helping behavior under different degree of moral identity. Yet, other potential mechanisms should not be ruled out. In addition to cognitive factors, cultural factors may also play critical roles in the relationship between ethical leadership and online helping behavior. For example, using the Meta-analysis method, Li et al. (2020) found that compared with the Anglo culture, ethical leadership in the Asian Confucian culture can stimulate employees’ voice better while ethical leadership in the Anglo culture is more effective in enhancing Leader-Member Exchange (LMX) and job satisfaction better. Resick et al.’s (2011) work revealed the importance of cultural factors as situational conditions to the ethical leadership research. For instance, Confucian Asia culture possesses typical characteristics such as team collectivism, institutional collectivism and performance orientation, and hence appreciates people-oriented leadership. People-oriented leaders are deemed as the supporters of employees as they prefer to adopt indirect communication ways to avoid ruining team’s harmony, and maintain a positive relationship. However, there emerges higher gender discrimination in Anglo culture, and Anglo culture presents individualism and performance orientation, values charm, participation and humanization. Future studies should examine the moderating effects of the cultural factors on the linkage of ethical leadership and online helping behavior.

Second, this study primarily sheds light on the consequences and conditions of ethical leadership, making its antecedents remained undeveloped. Ahn et al. (2013) regarded leaders’ core self-evaluation as the antecedent of ethical leadership, and believed that leaders’ ethical behaviors are related to their individual characteristics. Moreover, leaders’ conscientiousness and reflectiveness have been shown to play an important role in shaping behaviors (Babalola et al., 2019). However, existing studies are inadequate to explore the effect of leaders’ individual characteristics on ethical leadership, especially empirical analysis under different circumstances. The ensuing research could unravel the triggering mechanisms of ethical leadership utilizing meta-analysis method or qualitative comparative analysis method.

Third, although focusing on manufacturing enterprises under Chinese context can reduce the influence of culture and industry on the present research’s findings, this also limits the universality of our conclusions in other cultures and industries. Prior studies have indirectly indicated that industry has a significant impact on ethical leadership. For instance, Joplin et al. (2021) found that high ethical leadership cannot lead to high employee engagement in service sector; analyzing financial industry, Ahmad and Gao (2018) suggested that ethical leadership enables to enhance employee engagement through psychological empowerment. Therefore, we expect future research can clarify the variance of the influence of ethical leadership on employees’ outcomes in different industries, and thus provides specific management implications for enterprises in different industries.

Fourth, the results of this study show a significant moderating role of moral identity in the linkage of ethical leadership and other-praising moral emotion, yet this moderating effect is very weak judging from the INDEX value in Table 5. One possible explanation is that the level of ethical identity is likely to be too stable to form the variance of identity in a specific industry. Future research can use sample from multiple industries to test the conclusions of this study.

Conclusions

The emergence and thriving of home office directly induce a series of online helping behaviors. Previous research tends to link helping behavior with employees’ moral characteristics (Clercq et al., 2019; Tang et al., 2008). Social learning theory and leadership behavior research suggest that employees’ moral characteristics are largely influenced by ethical leaders (Men et al., 2020; Wang et al., 2018; Mayer et al., 2012). Yet, it remains unclear about the underlying mechanism of ethical leaders influencing followers’ online helping behavior. Thus, this study makes an important contribution by examining how and when ethical leadership is more effective in enhancing online helping behavior by highlighting the importance of other-praising moral emotion (influential path) and moral identity (influential boundary). In doing so, we depict a holistic picture of how to convert ethical leadership into employees’ online helping behavior. Our findings advance the understanding of the relationship between ethical leadership and employees’ behaviors, and provide some valuable references for the practices of organizational management.

Biographies

Shuwen Li

is an assistant professor at School of Economics and Management, Dalian University of Technology, China. His research interests include organizational emotion, strategic human resource management, organizational innovation management, and leadership. His researches have been published in Asia Pacific of Human Resource Management, Journal of Knowledge Management, and International Journal of Human Resource Management.

Ruiqian Jia

is a Postdoctoral student of management at Fudan University, China. His research interests include innovation management, strategic human resource management, and leadership. He has published dozens of articles in journals such as Asia Pacific of Human Resource Management, Journal of Knowledge Management, and Journal of Business Ethics.

Juergen H. Seufert

is an assistant professor at University of Shanghai for Science and technology, Sino British college. He received a PhD from the University of Wollongong, Australia, for his thesis on Qanta’s Marginal Abatement cost curve in the context of late capitalism and climate change. His research is multi-disciplinary ranging from environmental efficiency in economics to environmental management accounting in accounting. His work in environmental management accounting focuses particularly on the Marginal Abatement Cost Curve as investment tool to save CO2 and costs.

Jinlian Luo

is a professor at the school of economics and management, Tongji University, China. Her research interests include ambidextrous leadership and team innovation. Her researches have been published in International Journal of Human Resource Management, and Journal of Management.

Rui Sun

is a professor at the China Academy of Personnel Sciences. He received his PhD from Tsinghua University. His research interests include organizational emotional capability and innovation.

Appendix 1

Variables	Items	Factor loading	CR	The sqr of AVE	Variables	Items	Factor loading	CR	The sqr of AVE
Ethical leadership	EB1	.824	.936	.773	Moral identity	MI1	.730	.937	.773
	EB2	.855				MI2	.771
	EB3	.767				MI3	.763
	EB4	.813				MI4	.763
	EB5	.817				MI5	.757
	EB6	.768				MI6	.726
	EB7	.626				MI7	.778
	EB8	.709				MI8	.850
	EB9	.765				MI9	.815
	EB10	.760				MI10	.765
Other-praising emotion	ME1	.855	.912	.880	Online helping behavior	OHB1	.867	.914	.883
	ME2	.874				OHB2	.902
	ME3	.911				OHB3	.878

Declarations

Conflict of interest

No conflict of interest exits in the submission of this manuscript, and manuscript is approved by all authors for publication. I would like to declare on behalf of my co-authors that the work described was original research that has not been published previously, and not under consideration for publication elsewhere, in whole or in part. All the authors listed have approved the manuscript that is enclosed.