As a value orientation that prioritizes the collective interests of groups and organizations over individuals’ self-interests, collectivism has received a lion’s share of scholarly attention in cross-cultural (e.g., Earley, Reference Earley1989; Hofstede, Reference Hofstede1980; Kirkman, Lowe, & Gibson, Reference Kirkman, Lowe and Gibson2006; Markus & Kitayama, Reference Markus and Kitayama1991; Parsons & Shils, Reference Parsons and Shils1951; Triandis, Reference Triandis1995) and organizational behavior research (e.g., Chatman & Barsade, Reference Chatman and Barsade1995; He, Chen, & Zhang, Reference He, Chen and Zhang2004; Wagner, Reference Wagner, Staw and Kramer2002; Wagner & Moch, Reference Wagner and Moch1986). Although value orientations are intricately related to ethical decisions and ethical behaviors (Haidt & Joseph, Reference Haidt and Joseph2004; Schwartz, Reference Schwartz1994), researchers have only recently started to address collectivism’s impact on ethical conduct in organizations. Unfortunately, several streams of research on the subject tell two conflicting stories.
On one hand, collectivism is thought to promote ethical conduct. Collectivism is recognized as other oriented (Markus & Kitayama, Reference Markus and Kitayama1991) and prosocial (Cohen & Morse, Reference Cohen and Morse2014), which means that a collectivist orientation should constrain selfishness and motivate good citizenship behaviors in organizations. Consistent with this argument, collectivism has been credited for promoting cooperation (Wagner, Reference Wagner1995), reducing social loafing (Earley, Reference Earley1989), blurring diversity fault lines (Chatman, Greer, Sherman, & Doerr, Reference Chatman, Greer, Sherman and Doerr2019), boosting support for organizational reform (Chen, Meindl, & Hunt, Reference Chen, Meindl and Hunt1997; He et al., Reference He, Chen and Zhang2004), and increasing workplace adaptivity (Chatman & Barsade, Reference Chatman and Barsade1995). Indeed, the bulk of organizational research on collectivism to date has focused on illuminating this construct’s numerous positive implications for employees, teams, and organizations.
On the other hand, several researchers have empirically demonstrated that collectivism can sometimes motivate unethical behaviors. For example, scholars have reported (sometimes contrary to what was hypothesized) that collectivism leads to stronger tolerance and endorsement of unethical judgments and practices as compared to individualism (Armstrong, Reference Armstrong1996; Christie, Kwon, Stoeberl, & Baumhart, Reference Christie, Kwon, Stoeberl and Baumhart2003; Danon-Leva, Cavico, & Mujtaba, Reference Danon-Leva, Cavico and Mujtaba2010; Scholtens & Dam, Reference Scholtens and Dam2007). Similarly, others have theorized or empirically demonstrated that collectivism is associated with lying in business negotiations (Triandis et al., Reference Triandis, Carnevale, Gelfand, Robert, Wasti, Probst, Kashima, Dragonas, Chan, Chen, Kim, De Dreu, Van De Vliert, Iwao, Ohbuchi and Schmitz2001), opportunism in business transactions (Chen, Peng, & Saparito, Reference Chen, Peng and Saparito2002), aggression against outgroup members (Cohen, Montoya, & Insko, Reference Cohen, Montoya and Insko2006), and violating social and moral norms to maximize ingroup interests (Wildschut, Insko, & Gaertner, Reference Wildschut, Insko and Gaertner2002). How can collectivism simultaneously motivate both ethical and unethical behaviors in the workplace?
In the present research, we seek to clarify these literatures and their incongruous findings by proposing a model of collectivism and ethical behavior. At its core, our model focuses on the primacy of ingroup interest (over that of both self-interest and outgroup interest) at the heart of collectivist concerns (Hofstede, Reference Hofstede1980; Parsons & Shils, Reference Parsons and Shils1951; Triandis, Reference Triandis1995). Generally speaking, we integrate collectivism theory and research with prior work on social identity, moral boundedness (Crimston, Bain, Hornsey, & Bastian, Reference Crimston, Bain, Hornsey and Bastian2016; Singer, Reference Singer2011), group morality (Cohen et al., Reference Cohen, Montoya and Insko2006; Wildschut & Insko, Reference Wildschut, Insko and Van Lange2006), and moral identity (Aquino & Reed, Reference Aquino and Reed2002) to argue that collectivists’ other orientation is morally bounded depending on who the other is. Specifically, we propose that collectivistically oriented employees are motivated to place greater priority on and feel more committed to their organization’s goals compared not only to their own goals but also to those of organizational outsiders (e.g., customers or competitors). Thus our model proposes that collectivism constrains unethical pro-self behaviors but promotes unethical pro-organization behaviors through its effect on organizational goal commitment—a key psychological construct that binds employees to organizations in the service of organizational performance (Erez & Earley, Reference Erez and Earley1993). We further contend that the strength of an individual’s moral identity moderates these effects by broadening collectivists’ moral concerns to others beyond the ingroup. We report three studies that test our arguments and provide evidence supporting our view of collectivism’s double-edged effects. In the end, our research reveals that collectivism is not a simple cure-all to what some scholars view as the downsides of individualism at work (Chen, Chen, & Meindl, Reference Chen, Chen and Meindl1998; Haynes, Campbell, & Hitt, Reference Haynes, Campbell and Hitt2017) but rather a value orientation whose expression is bounded the individual’s scope of moral concern (Crimston et al., Reference Crimston, Bain, Hornsey and Bastian2016; Singer, Reference Singer2011).
In addition to its important empirical implications, this research makes at least two significant theoretical contributions to the literature. First, our research extends the literatures on collectivism and ethics by providing an explanation for why and how collectivism simultaneously motivates both ethical and unethical behavior. Organizational life presents challenging moral dilemmas because of conflicts of interest embedded in complex stakeholder relationships. By revealing collectivism’s paradoxical effects on behaviors involving two different types of conflicts of interest in the workplace, namely, self-ingroup and self-outgroup, this model shows that employees’ ethicality depends on their relational values and the nature of the socioethical dilemmas in which they find themselves. We suggest that this approach can serve as an important template for exploring the impact of other prosocial values (e.g., loyalty) as well.
Second, the present research furthers ongoing conversations on moral boundedness (Crimston et al., Reference Crimston, Bain, Hornsey and Bastian2016; Singer, Reference Singer2011). In particular, our model provides new insights for better understanding how relational values and features of identity limit or expand people’s circle of moral concern in organizational contexts, thereby affecting ethical and unethical behavior. As such, it creates opportunities for investigating potentially similar effects of other prosocial constructs in organizational contexts, such as organization-specific prosocial helping identities (Farmer & Van Dyne, Reference Farmer and Van Dyne2010, Reference Farmer and Van Dyne2017) and group and organizational identities and identifications (Bolino & Grant, Reference Bolino and Grant2016; Brickson, Reference Brickson2005; Zhang, Chen, Chen, Liu, & Johnson, Reference Zhang, Chen, Chen, Liu and Johnson2014). To the extent that scholars interested in such topics are similarly willing to accept moral boundedness as a theoretical assumption of their theoretical models, entirely new insights might be imminent.
THEORY AND HYPOTHESES
The Boundedly Moral Nature of Collectivism
Collectivism has been conceptualized in diverse ways by anthropologists, sociologists, social psychologists, and management researchers (Chen, Meindl, & Hunt, Reference Chen, Meindl and Hunt1997; Erez & Earley, Reference Erez and Earley1993; Hofstede, Reference Hofstede1980; Markus & Kitayama, Reference Markus and Kitayama1991; Parsons & Shils, Reference Parsons and Shils1951; Triandis, Reference Triandis1995; Wagner, Reference Wagner, Staw and Kramer2002; Wagner & Moch, Reference Wagner and Moch1986). Nevertheless, at its core, collectivism describes differences in people’s general value orientation regarding self-identities and self–other relationships, particularly in situations involving self–other conflicts of interest. Rooted in an interdependent self-construal and a salient collective identity of “We,” collectivism orients people to prioritize the interests of collectivities over their individual self-interests. In contrast, individualism, its conceptual complement, is based on an independent self-construal and a salient personal identity of “I” and calls for prioritizing self-interests above all else (Earley, Reference Earley1989; Hofstede, Reference Hofstede1980; Markus & Kitayama, Reference Markus and Kitayama1991; Triandis, Reference Triandis1995; Wagner, Reference Wagner1995).Footnote 1
Given collectivism’s emphasis on and prioritization of group interests over individual self-interest, it is hardly surprising that organizational scholars have focused largely on its positive implications for workplace performance and behavior, particularly for employee prosociality. However, as prior cross-cultural management and organizational behavior research (e.g., Chen et al., Reference Chen, Peng and Saparito2002) suggests, even if a collectivist orientation can discourage selfish behavior at work, there is also evidence that it can motivate antisocial behavior (e.g., opportunism, aggressiveness) toward external stakeholders as well.
We contend that one plausible explanation for collectivism’s negative effect on ethical behavior is that the collectivist value orientation is morally bounded. Over the past few decades, social categorization and social identity scholars have accumulated robust evidence of people’s ingroup favoritism and outgroup derogation across all societies and cultures (Brewer & Chen, Reference Brewer and Chen2007; Graham, Nosek, Haidt, Iyer, Koleva, & Ditto, Reference Graham, Nosek, Haidt, Iyer, Koleva and Ditto2011; Hogg & Terry, Reference Hogg and Terry2000; Tajfel & Turner, Reference Tajfel, Turner, Worchel and Austin1985). At the same time, behavioral ethics researchers have emphasized the boundedness of human beings’ moral circle (Crimston et al., Reference Crimston, Bain, Hornsey and Bastian2016; Crimston, Hornsey, Bain, & Bastian, Reference Crimston, Hornsey, Bain and Bastian2018; Singer, Reference Singer2011) in that social identities (Hogg & Terry, Reference Hogg and Terry2000), group morality (Cohen & Morse, Reference Cohen and Morse2014; Haidt & Joseph, Reference Haidt and Joseph2004), and cognitive limitations (Bazerman & Tenbrunsel, Reference Bazerman and Tenbrunsel2011) tend to circumscribe people’s felt moral obligation to the particular groups to which they belong and engender varying degrees of moral exclusion (Opotow, Reference Opotow1990). Although these literatures have not looked specifically at how prosocial values, such as collectivism, may induce psychological dynamics that lead to unethical behaviors against organizations’ external stakeholders, the logic of bounded morality suggests that prosocial values and social identities may combine to predict seemingly paradoxical ethical behaviors. How might this work?
Although collectivism, as a value orientation, is thought of as being “other oriented” and involving a willingness to be accommodating and self-sacrificial for the sake of others (e.g., Markus & Kitayama, Reference Markus and Kitayama1991), some have emphasized that collectivism’s other orientation is restricted to the membership of ingroups (Chen et al., Reference Chen, Peng and Saparito2002; Triandis, Reference Triandis1995). This suggests that pro-sociality stemming from a collectivist value orientation is contingent on a determination of who is “us” and who is “them” and that, relative to individualists, collectivists are more other oriented in ingroup social interactions. Although this ingroup orientation may not in and of itself be malevolent toward outgroups (Brewer, Reference Brewer1999), it nevertheless raises questions about whether collectivists’ moral obligation toward ingroups might motivate harmful behaviors toward outgroups in situations characterized by competition and conflicts of interest, as well as what psychological mechanisms facilitate such outcomes. To address these questions, we propose a model (see Figure 1) of collectivism’s double-edged effects on employee ethicality in business organizations where the interests of the individual conflict with the interests of the organization.
Figure 1: Theoretical Model
Double-Edged Effects of Collectivism on Unethical Behavior
Collectivism and Unethical Pro-Self Behavior
To begin, we submit that individuals higher in collectivism, because of their stronger moral concern for the well-being (e.g., needs, goals, and interests) of the ingroup, will be less likely to undertake unethical behaviors that benefit the self at the expense of their coworkers, work group, or organization (Akaah, Reference Akaah1996; Mitchell, Baer, Ambrose, Folger, & Palmer, Reference Mitchell, Baer, Ambrose, Folger and Palmer2018), namely, unethical pro-self behavior (UPSB) (Veetikazhi, Kamalanabhan, Malhotra, Arora, & Mueller, Reference Veetikazhi, Kamalanabhan, Malhotra, Arora and Mueller2022). Three arguments, each rooted in the principle that collectivists prioritize group interests and goals over those of individuals, support this proposition. First, high collectivists have a more salient interdependent self-construal and, as a result, tend to perceive and experience less conflict between their personal interests and those of collectivities to which they belong (Erez & Earley, Reference Erez and Earley1993). In other words, they may experience less tension between selfish and altruistic motives because they are more likely to fuse the “I” self with the “We” organization (Janoff-Bulman & Carnes, Reference Janoff-Bulman and Carnes2013; Swann & Buhrmester, Reference Swann and Buhrmester2015). Second, several scholars have suggested that when a conflict of interest does arise and requires a trade-off, individuals with a stronger collectivist value orientation are more attentive to the welfare of the collective, including its needs, goals, and interests, than they are to the self (Chen et al., Reference Chen, Peng and Saparito2002; Oyserman, Coon, & Kemmelmeier, Reference Oyserman, Coon and Kemmelmeier2002). Finally, we suggest that to the extent that self-ingroup ethical dilemmas are embedded in the broader context of intergroup business and social relationships, high collectivists are likely to be more vigilant about external competition and threat, making them more willing to set aside self-interest for the common good of the ingroup (Wildschut & Insko, Reference Wildschut, Insko and Van Lange2006). Thus we propose that individuals high (as opposed to low) in collectivism are less likely to undertake UPSB in organizations due to heightened moral concern for the welfare of the ingroup and fellow ingroup (organizational) members (Opotow, Reference Opotow1990).
Collectivism and Unethical Pro-Organization Behavior
Unethical pro-organization behavior (UPOB) refers to actions that are intended to promote the effective functioning of the organization yet violate societal standards of proper conduct (e.g., lying about product quality to make one’s company look good) (Umphress & Bingham, Reference Umphress and Bingham2011). Although UPOB and UPSB both violate moral norms of society at large, their essential distinction is that UPSB benefits the self, often at the expense of the organization, whereas UPOB is committed on behalf of the organization, often at the expense of those outside of it.
We propose that individuals with a stronger collectivist value orientation will be more likely to commit UPOB for several reasons. First, when intergroup demarcations are clear (as in most organizations), more collectivistic individuals are more likely to engage in group-based social comparisons and seek positive distinctions (Tajfel & Turner, Reference Tajfel, Turner, Worchel and Austin1985), thereby lowering their moral concern for the welfare of members of the outgroup. Second, the common good of the ingroup, including its needs, goals, and interests, is a dominant concern of high collectivists and justifies the use of unethical means to disadvantage the outgroup (Bandura, Reference Bandura1991; Opotow, Reference Opotow1990). Thus internal sanctions (arising from personal standards) against misconduct toward the outgroup will be overpowered by collectivists’ moral norm of ingroup favoring and a felt need to socially support competitive ingroup champions (Wildschut, Pinter, Vevea, Insko, & Schopler, Reference Wildschut, Pinter, Vevea, Insko and Schopler2003). Finally, blaming the victim, or victimization (Bandura, Reference Bandura1999), is made easier when the target is outgroup members and ingroup members share negative expectations about them (Cikara, Reference Cikara2015; Insko et al., Reference Insko, Schopler, Pemberton, Wieselquist, McIlraith, Curno and Gaertner1998; Wildschut et al., Reference Wildschut, Insko and Gaertner2002). For instance, employees in a sales context may explain their UPOB on grounds that buyers (if perceived as outgroup members) are simply out to maximize their own bottom line and hence it is their own fault for not knowing better.
Extant theory and research provide both direct and indirect support for our arguments. Regarding collectivism inhibiting UPSB, individuals higher in collectivism have been found to exhibit more self-sacrificial behavior (Triandis, Reference Triandis1995) and less intragroup social loafing (Wagner, Reference Wagner1995), even in the absence of individual accountability (Earley, Reference Earley1989). Regarding collectivism promoting UPOB, research has shown that collectivists who attach greater priority to ingroup goals and interests exhibit a greater tendency to lie in business negotiations (Triandis et al., Reference Triandis, Carnevale, Gelfand, Robert, Wasti, Probst, Kashima, Dragonas, Chan, Chen, Kim, De Dreu, Van De Vliert, Iwao, Ohbuchi and Schmitz2001). Similarly, Cohen et al. (Reference Cohen, Montoya and Insko2006) found that high-guilt-prone individuals also score high on ingroup empathy and are more concerned with maximizing ingroup outcomes, which leads to a more competitive stance against the outgroup that often violates moral norms. While neither of these latter two studies focused directly on UPOB, Yang, Lu, Yao, and Liang (Reference Yang, Lu, Yao and Liang2020), in a recent study of how leaders inspire follower UPOB, found that followers’ collectivist orientation enhances the positive effect of leaders’ self-sacrificial behavior on follower UPOB via identification with the leader. On the basis of the preceding theorizing and findings, we thus hypothesize the following:
Hypothesis 1a: Collectivism is negatively related to unethical pro-self behavior.
Hypothesis 1b: Collectivism is positively related to unethical pro-organization behavior.
The Mediating Role of Organizational Goal Commitment
We suggest that collectivism’s effects on unethical behaviors are mediated by organizational goal commitment. Goals and goal setting are powerful motivational tools for organizational and employee performance (Kanfer & Chen, Reference Kanfer and Chen2016), but their performance effects depend on an individual’s organizational goal commitment, that is, their “determination to try for a goal and the persistence in pursuing it over time” (Hollenbeck, Williams, & Klein, Reference Hollenbeck, Williams and Klein1989: 18). An organization’s goals may be somewhat aligned with individual members’ goals, but these goals are nevertheless distinct from each other and therefore give rise to conflicts of interest (Goodpaster, Reference Goodpaster1991). For example, an organization’s productivity and profitability goals are often likely to conflict with employees’ financial and quality-of-life goals (e.g., Wu, Ferris, Kwan, Chiang, Snape, & Liang, Reference Wu, Ferris, Kwan, Chiang, Snape and Liang2015). With this in mind, we argue that organizational goal commitment will affect how people navigate or resolve conflicts of interest between the self and organization (e.g., UPSB) and between the organization and its external stakeholders (e.g., UPOB) (Wildschut et al., Reference Wildschut, Pinter, Vevea, Insko and Schopler2003).
Past research has documented goal commitment as motivating task performance (Kanfer & Chen, Reference Kanfer and Chen2016; Locke & Latham, Reference Locke and Latham2002; Rodgers & Hunter, Reference Rodgers and Hunter1991), but several scholars have warned that the single-mindedness of goal commitment can lead to deviant behaviors (e.g., Salancik, Reference Salancik, Staw and Salancik1977). Specifically, they have referenced the motivation-constraining principle to argue that commitment to performance goals can be so dominant as to leave little cognitive and affective space for moral considerations (such as the legitimate interests of competitors and society) (e.g., Moore, Reference Moore and De Cremer2009; Shah, Friedman, & Kruglanski, Reference Shah, Friedman and Kruglanski2002). Thus organizational goal commitment represents a plausible key mechanism through which collectivism exerts its double-edged effects on employees’ workplace ethical behavior. All organizations generally encourage and expect employees to be highly committed to organizational goals, but employees with a high collectivist value orientation are most likely to perceive personal and organizational goals as cooperatively interdependent and are therefore more willing to commit to organizational goals, even when their personal goals are not well served (Chen et al., Reference Chen, Chen and Meindl1998). In this light, we propose that high commitment to organizational goals acts as a self-regulating mechanism that serves to inhibit unethical behavior that advances an employee’s individual goals at the expense of the organization (i.e., UPSB).
These arguments notwithstanding, we suggest that organizational goal commitment should also act, in contrast to its prohibitive effect on UPSB, as a powerful motivator of UPOB (Moore & Gino, Reference Moore and Gino2013; Randall, Reference Randall1987). When faced with dilemmas involving conflicts of interest between the organization and its external stakeholders, high commitment to the organization’s performance goals is likely to lead employees to prioritize goal accomplishment above the interests of external stakeholders and societal moral dictates (Opotow, Reference Opotow1990; Randall, Reference Randall1987). For example, employee buy-in to Volkswagen’s “faster, higher, and farther” corporate goal has been identified as a key contributor to the creation of the cheating device in its emission control system, which temporarily advantaged the company but violated government regulations and damaged both consumers and the environment (Ewing, Reference Ewing2017). Thus organizational goal commitment is likely to play a critical role in facilitating the double-edge effects of collectivism on unethical pro-self and unethical pro-organization behavior.
Hypothesis 2a: Organizational goal commitment mediates the negative relationship between collectivism and unethical pro-self behavior such that collectivism is positively related to organizational goal commitment, which in turn is negatively related to unethical pro-self behavior.
Hypothesis 2b: Organizational goal commitment mediates the positive relationship between collectivism and unethical pro-organization behavior such that collectivism is positively related to organizational goal commitment, which in turn is positively related to unethical pro-organization behavior.
The Moderating Role of Moral Identity
Finally, we suggest that an individual’s moral identity acts as a key boundary condition on the relationships hypothesized herein. We focus on moral identity as a boundary condition because of its potential to simultaneously enhance collectivists’ moral concern for ingroup others while expanding their moral concern for outgroup others (Reed & Aquino, Reference Reed and Aquino2003). Moral identity is a self-schema organized around a set of universalistic prosocial attributes such as being compassionate, caring, honest, and fair (Aquino & Reed, Reference Aquino and Reed2002). Moral identity has been identified as a powerful driver of moral self-regulation (e.g., Aquino & Reed, Reference Aquino and Reed2002; Lapsley, Reference Lapsley1998; Lapsley & Lasky, Reference Lapsley and Lasky2001; Mayer, Aquino, Greenbaum, & Kuenzi, Reference Mayer, Aquino, Greenbaum and Kuenzi2012; Shao, Aquino, & Freeman, Reference Shao, Aquino and Freeman2008), connecting moral judgment and moral action to both promote ethical behavior (Cohen, Panter, Turan, Morse, & Kim, Reference Cohen, Panter, Turan, Morse and Kim2014; Damon & Hart, Reference Damon, Hart, Bornstein and Lamb1992; Mayer et al., Reference Mayer, Aquino, Greenbaum and Kuenzi2012; for review, see also Shao et al., Reference Shao, Aquino and Freeman2008) and inhibit unethical behavior (Aquino & Reed, Reference Aquino and Reed2002; Cohen et al., Reference Cohen, Panter, Turan, Morse and Kim2014; Mayer et al., Reference Mayer, Aquino, Greenbaum and Kuenzi2012; Skarlicki, van Jaarsveld, Shao, Song, & Wang, Reference Skarlicki, van Jaarsveld, Shao, Song and Wang2016).
To understand moral identity’s moderating effect, it is worth noting that collectivism motivates organizational goal commitment because of its other-oriented value system—an orientation it shares in common with moral identity. As argued earlier, however, collectivists’ organizational commitment is limited to the business goals and objectives of the organization. Moral identity, by contrast, constitutes a set of universalistic moral attributes that go beyond commitment to self- and ingroup interests (Reed & Aquino, Reference Reed and Aquino2003). We therefore contend that moral identity, with its capacity to expand an individual’s circle of moral concern (Crimston et al., Reference Crimston, Bain, Hornsey and Bastian2016; Reed & Aquino, Reference Reed and Aquino2003), serves as a key construct that moderates the impact of collectivists’ organizational goal commitment on UPSB and UPOB. Specifically, we contend that moral identity strengthens the inhibiting effect of organizational goal commitment (motivated by collectivism) on UPSB but weakens the promotional effect of organizational goal commitment on UPOB. To the extent that moral identity and organizational goal commitment converge when it comes to resolving self-organization conflicts of interest, moral identity naturally reinforces collectivists’ organization-serving cognitions and motives (Aquino & Reed, Reference Aquino and Reed2002; Reynolds & Ceranic, Reference Reynolds and Ceranic2007), thereby affirming their organizational goal commitment and reducing UPSB.
With regard to UPOB, however, we contend that moral identity will act as a countervailing force to collectivists’ organizational goal commitment for several reasons. First, a strong moral identity raises collectivists’ moral awareness about the legitimate interests and goals of outgroups and about the potentially harmful effect that ingroup goal commitment could cause to these outgroups. Second, a strong moral identity makes salient one’s personal, more inclusive moral obligation that potentially buffers or counteracts the effects of ingroup goal commitment on one’s behavior toward outgroups. Last, moral identity, by expanding an individual’s circle of moral regard (Crimston et al., Reference Crimston, Bain, Hornsey and Bastian2016; Reed & Aquino, Reference Reed and Aquino2003), may enable collectivists to see outgroups as relevant stakeholders whose interests are intricately related to the ingroup. For all these reasons, we propose that collectivists with a strong (as opposed to weak) moral identity are more likely to be aware and considerate of the needs and goals of outgroups, thereby reducing the positive effect of organizational goal commitment on unethical pro-organizational behaviors. Indeed, research by Skarlicki et al. (Reference Skarlicki, van Jaarsveld, Shao, Song and Wang2016) provides evidence for just such a buffering effect, showing that employees who internalize moral ideals are less likely to retaliate in response to mistreatment from customers. Thus, we propose the following:
Hypothesis 3a: The negative indirect relationship between collectivism and unethical pro-self behavior via organizational goal commitment is moderated by moral identity such that it is stronger for individuals with stronger moral identity.
Hypothesis 3b: The positive indirect relationship between collectivism and unethical pro-organization behavior via organizational goal commitment is moderated by moral identity such that it is weaker for those with stronger moral identity.
OVERVIEW OF STUDIES
We conducted three studies to examine our hypotheses. The studies employed both survey and experimental methodologies in an effort to strengthen the ecological and internal validity of the findings. In study 1, we utilized multiwave survey data in a Chinese retail company to test the proposed double-edged effects of collectivism on workplace unethical behaviors and the proposed mediating mechanism. In study 2, we explored the generalizability of these findings by administering a vignette-based survey to a sample of US working adults. Finally, in study 3, we tested our full theoretical model via a vignette-based experiment that manipulated collectivist value orientation and measured moral identity.
STUDY 1
Pilot Study
Prior to study 1, we conducted a pilot study to establish that unethical pro-organization behavior in the workplace, as originally conceived of and measured in the United States by Umphress and colleagues (Reference Umphress, Bingham and Mitchell2010), is also judged as unethical in China. We viewed such confirmation as important because these behaviors might be viewed differently in a collectivist society. We recruited 238 working adults (female = 52.9 percent; male = 47.1 percent; M age = 32.29 years; M tenure = 9.16 years) through Sojump, a widely used Chinese crowdsourcing platform (e.g., Keller, Loewenstein, & Yan, Reference Keller, Loewenstein and Yan2017). Respondents were asked to read the six items that compose Umphress and colleagues’ (Reference Umphress, Bingham and Mitchell2010) original self-report measure of unethical pro-organization behavior and rate the extent to which they thought each behavior was morally right on a 7-point scale ranging from 1 (not at all) to 7 (very much) (α = .88). Their mean rating on this measure was 2.98 (SD = 1.32), which is significantly lower than 4 (the scale midpoint), t(237) = −11.97, p < .001. The results confirm that Chinese working adults perceive UPOB in the workplace to be unethical.
Main Study
Respondents and Procedure
Consistent with other research in the UPOB literature (Chen, Chen, & Sheldon, Reference Chen, Chen and Sheldon2016; Jiang, Hu, Hong, Liao, & Liu, Reference Jiang, Hu, Hong, Liao and Liu2016; Tang, Yam, & Koopman, Reference Tang, Yam and Koopman2020), we collected data from employees of a Chinese retail company at three different points in time. We introduced temporal separation between stages of the data collection process to help reduce potential common method bias in respondents’ retrieval and reporting of responses (Podsakoff, MacKenzie, & Podsakoff, Reference Podsakoff, MacKenzie and Podsakoff2012). All respondents had frequent interactions with clients, conducting business on behalf of their company in the course of their daily work. These employees, assured that their individual responses were confidential and for academic research only, filled out the surveys at their workplace either before or after work. In wave 1, respondents filled out survey questions assessing collectivism and organizational goal commitment, as well as a set of control variables, including age, gender, education, tenure, social desirability bias, and organizational identification. One month after wave 1, respondents reported their frequency of engaging in UPOB. Two months after wave 2, to reduce potential common method bias, we followed Vincent and Kouchaki’s (Reference Vincent and Kouchaki2016) procedure and invited supervisors to assess their subordinates’ frequency of engaging in UPSB.
Five hundred sixty salespersons of the company were contacted about voluntarily participating in the study. Of these employees, 301 completed the first survey. Of these 301 employees, 247 completed the second survey. Ultimately, we received 225 usable questionnaires from 38 supervisors of these 247 employees, resulting in a response rate to our entire set of measures of 40.2 percent (female = 76 percent; male = 24 percent; M age = 31–40 years; M tenure = 3.19 years).
To examine attrition effects, we compared three respondent groups: 1) those for whom we had data from both time points as well as from their supervisor (n = 225), 2) those who completed waves 1 and 2 but for whom we did not obtain supervisor-provided UPSB data (n = 22), and 3) those who completed the first wave only (n = 54). We found no difference among the three groups in terms of age, gender, tenure, position, and education.
Measures
We used 5-point Likert scales to measure all key variables. The English items were translated into Chinese using translation and back-translation procedures (Brislin, Reference Brislin, Lonner and Berry1986). All materials were presented in Chinese.
Employees’ collectivist value orientation was measured using an eight-item measure developed by Triandis and Gelfand (Reference Triandis and Gelfand1998) that assesses an individual’s concern for ingroup members (horizontal collectivism) and for the collective as a whole (vertical collectivism). Sample items are “The well-being of my coworkers is important to me” and “It is my duty to take care of my family, even when I have to sacrifice what I want,” rated on a 5-point scale ranging from 1 (strongly disagree) to 5 (strongly agree) (α = .85).
To assess respondents’ organizational goal commitment, we employed a three-item measure adapted from Hollenbeck et al. (Reference Hollenbeck, Williams and Klein1989): “I take my organization’s goals very seriously”; “I am strongly committed to pursuing my organization’s goals”; and “Quite frankly, I don’t care if my organization’s goals are achieved or not” (reverse scored), rated on a 5-point scale ranging from 1 (strongly disagree) to 5 (strongly agree) (α = .82).
Finally, to assess respondents’ tendency to engage in UPOB and UPSB, we asked our respondents to rate the former and their supervisors to rate the latter. Using different raters was intended to help reduce common source bias between the two measures themselves and in these measures’ relationships with measured antecedents (Podsakoff et al., Reference Podsakoff, MacKenzie and Podsakoff2012). It was also guided by a concern that respondents might be less candid about their willingness to engage UPSB than UPOB, whereas their supervisors might be more aware of and candid about the former.Footnote 2
To assess respondents’ UPSB, each respondent’s supervisor rated the frequency of their subordinates’ UPSB on six items from Akaah’s (Reference Akaah1996) seventeen-item measure of unethical behavior, a widely used measure of self-serving unethical behavior (e.g., Mayer et al., Reference Mayer, Aquino, Greenbaum and Kuenzi2012; Moore, Mayer, Chiang, Crossley, Karlesky, & Birtch, Reference Moore, Mayer, Chiang, Crossley, Karlesky and Birtch2019; Vincent & Kouchaki, Reference Vincent and Kouchaki2016). To identify the most ecologically relevant items from the scale, we subjected fifteen of them to a separate frequency assessment (we removed one item, “authorize a subordinate to violate company rules,” because the employees had no subordinates of their own and replaced two similar items, “pad an expense account up to 10 percent” and “pad an expense account more than 10 percent,” with a single item, “reimburse personal expenses from a business expense account”). We invited eleven senior and regional managers (female = 81.8 percent; male = 18.2 percent; M age = 38.55 years; M tenure = 7.27 years) at the company to rate the frequency with which these fifteen unethical behaviors occurred in their retail stores on a 7-point scale ranging from 1 (never) to 7 (always). Nine items were removed because supervisors rated them as “never” or “rarely” happening among retail employees. The remaining six items were rated by at least one rater as occurring at least “sometimes.” These six items were “use company services for personal use,” “conceal one’s errors,” “pass blame for errors to an innocent coworker,” “claim credit for someone else’s work,” “falsify time/quality/quantity reports,” and “reimburse personal expenses from a business expense account,” rated on a 5-point scale ranging from 1 (never) to 5 (always) (α = .94).
To measure UPOB, respondents completed Umphress et al.’s (Reference Umphress, Bingham and Mitchell2010) six-item measure about how frequently they engaged in each behavior. Sample items are “Since it helped my organization, I misrepresented the truth to make my organization look good” and “To help my organization, I exaggerated the truth about my company’s products and services to customers and clients,” rated on a 5-point scale ranging from 1 (never) to 5 (always) (α = .92).
Control variables. A past meta-analysis by Kish-Gephart, Harrison, and Treviño (Reference Kish-Gephart, Harrison and Treviño2010) suggested that age, gender, and education may relate to unethical decision-making and behavior. Furthermore, Umphress et al. (Reference Umphress, Bingham and Mitchell2010) suggested that position and tenure may potentially affect UPB. Considering the possibility that these demographic variables may be related to both a collectivist value orientation and the unethical behavior under study, we controlled for each of these five demographic variables. Gender was measured using a dummy variable (1 = female, 0 = male), education was measured using four continuous categories (1 = middle school or below, 2 = vocational school, 3 = university, and 4 = graduate school), position was measured using four continuous categories (1 = employee, 2 = frontline manager, 3 = middle manager, and 4 = senior manager), and tenure was measured by the number of years the employee had worked at the company. To account for socially desirable responding, we also measured and controlled for impression management bias, using a subscale from Steenkamp, De Long, and Baumgartner’s (Reference Steenkamp, De Long and Baumgartner2010) scale (α = .85). Finally, as previous research has found that organizational identification is a key antecedent of UPOB (e.g., Chen et al., Reference Chen, Chen and Sheldon2016) and could also be positively related to collectivism (Lee, Park, & Koo, Reference Lee, Park and Koo2015), we measured and controlled for organizational identification using the six-item scale developed by Mael and Ashforth (Reference Mael and Ashforth1992) (α = .74). Following Bernerth and Aguinis’s (Reference Bernerth and Aguinis2016) suggestion, we conducted the analyses with and without control variables. The sets of analyses generated the same patterns of results and levels of significance.
Analytical Strategy
All constructs were conceptualized and measured at the individual level. Because our data were nested (i.e., a single supervisor provided assessments for several subordinates), we employed multilevel structural equation modeling (MSEM) procedures to test our hypotheses via Mplus 7.20. MSEM can capture the nested nature of the data and examine multiple mediated and moderated relationships simultaneously (Preacher, Zyphur, & Zhang, Reference Preacher, Zyphur and Zhang2010). To account for nonnormal sampling distributions of indirect relations, we applied Monte Carlo simulation to calculate confidence intervals for indirect effects by using twenty thousand resamples via the R program (Preacher & Selig, Reference Preacher and Selig2012).
Results
Confirmatory factor analysis. Confirmatory factor analysis showed that the hypothesized four-factor model was a good fit to the data (χ2 = 338.58, df = 113, RMSEA = .09, SRMR = .05, CFI = .92, TLI = .91) and that all factor loadings were significant. Furthermore, this model achieved a better fit to data compared to models that combined collectivism and organizational goal commitment (Δχ2[3] = 49.43, p < .01, TLI = .89, CFI = .91, RMSEA = .10, SRMR = .06) or collectivism, organizational goal commitment, and unethical pro-organization behavior (Δχ2[5] = 381.63, p < .01, TLI = .76, CFI = .80, RMSEA = .15, SRMR = .14), supporting the distinctiveness of our key measures.
Descriptive statistics. Table 1 reports the means, standard deviations, and zero-order Pearson correlations of all variables in study 1.
Table 1: Means, Standard Deviations, Correlations, and Internal Consistency Estimate (Study 1)
Note. N = 225. T1 = time 1. T2 = time 2, one month after time 1. T3 = time 3, two months after T2. Coefficient alphas are given in parentheses on the diagonal.
* p < .05.
** p < .01, two-tailed.
Hypothesis testing. Hypotheses 1a and 1b propose a significant relationship between collectivism and both types of unethical behaviors studied (negative for UPSB, positive for UPOB). Results from our MSEM procedures show that, after taking control variables into account (results for control variables on UPSB: age = −.05 [.04] ns; gender = −.07 [.07] ns; education = .01 [.06], p < .05; tenure = .01 [.01] ns; position = −.13 [.09] ns; impression management bias = −.15 [.07] p<.05; organizational identification =.01 [.06] ns; results for control variables on UPOB: age = .05 [.05] ns; gender = −.03 [.13] ns; education = .05 [.09], p < .05; tenure = .01 [.01] ns; position = .02 [.07] ns; impression management bias = −.08 [.09] ns; organizational identification = .16 [.08], p < .10), collectivism was indeed negatively related to UPSB (B = −.22, SE = .11, p < .05) yet positively related to UPOB, as predicted (B = .32, SE = .10, p < .01). Thus hypotheses 1a and 1b were supported.
As can be seen in Figure 2, MSEM procedures also revealed that collectivism was significantly related, in the predicted manner, to organizational goal commitment (B = .55, SE = .13, p < .01) and that organizational goal commitment, in turn, was negatively related to UPSB (B = −.17, SE = .07, p < .05) but positively related to UPOB (B = .19, SE = .08, p < .05). We therefore tested the proposed indirect relationships specified by hypotheses 2a and 2b via R using twenty thousand resamples. In doing so, we found that the indirect relationship between collectivism and UPSB via organizational goal commitment was significant and negative (indirect effect = −.09, 95 percent CI = [−.19, −.02]), whereas the indirect relationship between collectivism and UPOB via organizational goal commitment was significant and positive (indirect effect = .10, 95 percent CI = [.02, .22]). Thus hypotheses 2a and 2b were supported.
Figure 2: Path Analysis Results for the Mediation Model (Study 1)
Note. N = 225. Unstandardized effects were reported, and standard errors appear in parentheses. For ease of readability, we did not present the coefficients of the paths from control variables (i.e., age, gender, education, tenure, position, impression management bias, and organizational identification) in the model. Results for control variables on organizational goal commitment are age = .08 [.04], p < .05; gender = .02 [.08] ns; education = .14 [.06], p < .05; tenure = −.01 [.01] ns; position = .12 [.05], p < .05; impression management bias = .13 [.09] ns; organizational identification = .10 [.08] ns; on UPSB are age = −.03 [.04] ns; gender = −.06 [.06] ns; education = .03 [.06] ns; tenure = .01 [.01] ns; position = −.11 [.09] ns; impression management bias = −.13 [.07], p < .10; organizational identification = .03 [.07] ns; and on UPOB are age = .04 [.05] ns; gender = −.03 [.13] ns; education = .03 [.09] ns; tenure = .01 [.01] ns; position = −.001 [.07] ns; impression management bias = −.11 [.08] ns; organizational identification = .14 [.09] ns. The explained variance was R 2 = .22 for organizational goal commitment, R 2 = .10 for UPSB, and R 2 = .11 for UPOB. *p < .10. **p < .05. ***p < .01.
Discussion
Study 1 provides support for the proposed double-edged effects of a collectivist value orientation on workplace unethical behaviors. It also offers strong support for our proposed mechanism. Given that the data were collected within China, whose collectivist national culture may have magnified the results, questions remained about the generalizability of the results. To replicate our findings in a more individualistic culture, we conducted a follow-up study with a sample of United States–based working adults.
STUDY 2
In our second study, we recruited a sample of United States–based working adults and assigned them to read one of two workplace decision-making scenarios, each of which posed a unique ethical dilemma: whether to engage in UPSB or whether to engage in UPOB. Each dilemma focused on one specific form of UPSB or UPOB that either pitted participants’ self-interest against their organization’s interests or their organization’s interests against those of an outgroup (or outgroups), respectively.Footnote 3 We then assessed whether individual differences in collectivism shaped participants’ responses. Because each participant was assigned to only one of these two dilemmas, we conducted this second study in two parts (study 2a and study 2b), testing hypotheses 1a and 2a and 1b and 2b separately.
Study 2a: UPSB
Participants and Procedure
Participants were 226 US working adults (female = 45.1 percent; male = 54.9 percent; M age = 35–44 years; M work experience = 25.09 years) recruited through Amazon’s Mechanical Turk (Buhrmester, Kwang, & Gosling, Reference Buhrmester, Kwang and Gosling2011). Thirteen participants were removed because they failed to correctly answer an attention check item, resulting in a final sample of 213.
Upon agreeing to participate, participants were placed in the role of a purchasing agent and asked to read a business decision-making scenario (see Appendix A) that posed an ethical dilemma (i.e., conflict of interest) about whether to include a supplier who regularly offered them personal perks on a list for contract renegotiation. In the scenario, the participant’s company was described as having recently decided to require all purchasing agents to reevaluate their current suppliers to shorten contract lengths so as to “increase the company’s flexibility to select new and better suppliers.” Though such action would reduce the company’s costs and increase its profit margins, it would likely upset targeted suppliers. The supplier in question was described as a prime candidate for contract renegotiation. Thus, although targeting this supplier for contract renegotiation would clearly help the participant’s company (and be in line with the company’s wishes), it would jeopardize future personal perks and VIP treatment that this supplier had previously directed toward the participant. We then assessed participants’ likely behavior, which we construed as the participant’s willingness to be swayed by a personal bribe, with two items: “How likely are you to leave this supplier off your list of suppliers to be targeted for short-term contracts?” and “How likely are you to recommend the old supplier of yours in the scenario as a target for renegotiation?” (reverse coded), rated on a 7-point scale ranging from 1 (not at all) to 7 (very likely) (α = .88). We also assessed participants’ level of organizational goal commitment by administering a context-specific version of the same three-item scale of organizational goal commitment used in study 1 (i.e., a version of this measure in which the items were adapted to refer specifically to the company’s stated goal of shortening the contract length with current suppliers), rated on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree). A sample item is “To what extent would you care about your company accomplishing its goal of signing shorter contracts with its existing suppliers?” Finally, we assessed participants’ level of collectivism by administering the eight-item collectivism scale used in study 1 (α = .82).
Results
Descriptive statistics. Table 2 reports the means, standard deviations, and zero-order Pearson correlations of all variables in study 2a.
Table 2: Means, Standard Deviations, Correlations, and Internal Consistency Estimate (Study 2a)
Note. N = 213.
* p < .05.
** p < .01, two-tailed.
Coefficient alphas are given in parentheses on the diagonal.
Hypothesis testing. To test the proposed negative and indirect effect of collectivism on UPSB (hypotheses 1a and 2a), we conducted two regression analyses and a bootstrapping analysis. Consistent with the results of study 1, after taking our control variables into account (results for control variables on organizational goal commitment: age = .26 [.06], p <.01; gender = −.14 [.16] ns; education = −.04 [.06] ns; impression management bias = .09 [.07] ns; results for control variables on UPSB: age = −.19 [.10], p = .05; gender = .22 [.23] ns; education = .01 [.09] ns; impression management bias = −.20 [.10], p = .05), participants’ level of collectivism was once again positively related to their level of organizational goal commitment (B = .43, SE = .09, p < .01). Unexpectedly, however, the previously observed negative relationship between collectivism and UPSB, though still negative, did not reach significance (B = −.15, SE = .14, ns) (a finding to which we will return in our discussion of this study’s results). Thus hypothesis 1a was not supported in this second study. That said, a follow-up bootstrapping analysis did reveal that the proposed indirect effect of collectivism on UPSB via organizational goal commitment was both negative and significant (indirect effect = −.18, 95 percent CI = [−.32, −.07]). Thus hypothesis 2a was once again supported.
Study 2b: UPOB
Participants and Procedure
To explore the proposed positive relationship between collectivism and UPOB, a new sample of 226 US working adults (female = 44.7 percent; male = 55.3 percent; M age = 35–44 years; M work experience = 23.80 years) was recruited, again from Amazon’s Mechanical Turk platform. Nine were removed for incorrectly answering an attention check item, leaving a final sample of 217 participants.
We employed the same procedure as in study 2a. All aspects of studies 2a and 2b were identical, except for the business decision-making scenario, which presented an opportunity to engage in corporate bribery (i.e., UPOB) as opposed to personal bribery (i.e., UPSB). In this study, the company at which the participants were employed was described as facing a saturated domestic market and having recently decided to shift its focus to international business. The goal of the new strategy was “to cultivate international business so that it becomes the company’s major center of growth in business and profit.” In their role as a sales agent, participants were confronted with a decision about whether to pay a bribe to an international buyer to win a lucrative contract (order) for the company. To assess their likelihood of engaging in unethical pro-organization behavior, participants responded to two items—“How inclined are you to pay the money to win the order?” and “How likely are you to pay the money to satisfy the buyer’s request?”—rated on a 7-point scale ranging from 1 (not at all) to 7 (very likely) (α =.96). Participants’ level of organizational goal commitment was measured using the same, context-specific three-item scale of organizational goal commitment employed in study 2a, adapted to this new scenario (α = .93). We also used the same eight-item scale to measure their level of collectivism.
Results
Descriptive statistics. Table 3 reports the means, standard deviations, and zero-order Pearson correlations of all variables in study 2b.
Table 3: Means, Standard Deviations, Correlations, and Internal Consistency Estimate (Study 2b)
Note. N = 217.
* p < .05.
** p < .01, two-tailed.
Coefficient alphas are given in parentheses on the diagonal.
Hypothesis testing. Employing the same analytical strategy as in study 2a, we found, as predicted, that after taking control variables into account (results for control variables on organizational goal commitment: age = .14 [.04], p < .01; gender = −.01 [.10] ns; education = −.04 [.05] ns; impression management bias = .02 [.05] ns; results for control variables on UPOB: age = −.01 [.11] ns; gender = −.07 [.24] ns; education = −.25 [.11] ns; impression management bias = −.27 [.12], p = .02), collectivism was positively related to both participants’ level of organizational goal commitment (B = .34, SE = .06, p < .01) and their likelihood of engaging in UPOB (B = .42, SE = .15, p < .01). Furthermore, we found that the proposed indirect effect of collectivism on UPOB via organizational goal commitment was positive and significant (indirect effect = .14, 95 percent CI = [.03, .26]). Thus hypotheses 2a and 2b were supported once again.
Discussion
Our first two studies (study 1 and study 2) provide strong, cross-national support for the proposed double-edged effects of a collectivist value orientation on workplace unethical behaviors. However, both studies test only the main effects of collectivism, and both rely on survey methodology, which limits our ability to draw causal inferences from the data. Furthermore, our mixed findings regarding hypothesis 1a leave some uncertainty regarding collectivism’s true relationship to UPSB. To address these limitations, we conducted an experiment in which we directly manipulated participants’ cultural value orientation and measured their moral identity. In treating cultural orientations as malleable, we follow prior experimental research on individualism-collectivism that has used priming activities (e.g., Gardner, Gabriel, & Lee, Reference Gardner, Gabriel and Lee1999; Goncalo & Staw, Reference Goncalo and Staw2006) and scenarios (Trafimow, Triandis, & Goto, Reference Trafimow, Triandis and Goto1991) to activate the unique psychological states associated with holding an individualist versus collectivist value orientation.
STUDY 3
For the third and final study, we recruited working adults in China and manipulated their cultural value orientation. As in study 2, we conducted the study in two parts (study 3a and study 3b) to examine UPSB and UPOB separately.
Study 3a: UPSB
Participants and Design
We recruited 174 participants (female = 67.2 percent; male = 32.8 percent; M age = 31–40 years; M work experience = 8.32 years) working in various industries through Sojump in exchange for 10 RMB. Fifteen were excluded after incorrectly answering a simple attention check item included in the study, thereby leaving a final sample of 159 participants.
Upon agreeing to participate, participants were invited to complete a task, adopted from Goncalo and Duguid (Reference Goncalo and Duguid2012), that served as our manipulation of collectivism. In this activity, participants responded to three questions designed to prime either a collectivist or an individualist value orientation. Specifically, those assigned to the collectivist value orientation condition were instructed to write 1) “three statements describing groups to which you belong,” 2) “three statements describing why you think you are like most other people,” and 3) “three statements describing why you think it might be advantageous to ‘blend in’ with other people.” In contrast, those assigned to the individualist value orientation condition were instructed to write 1) “three statements describing yourself,” 2) “three statements describing why you think you are not like most other people,” and 3) “three statements describing why you think it might be advantageous to ‘stand out’ from other people.” Directly afterward, all participants responded to two items that assessed the extent to which their self-interests or those of their general ingroups were more salient. These two items, which served as a manipulation check, were “My responses to this first activity make me care more about my personal self-interest” and “My responses to this first activity make me care more about the interests of groups to which I belong,” rated on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree).
Following this task, participants read a scenario (see Appendix B) that posed a dilemma about whether to submit receipts for personal expenses incurred during business trips for reimbursement (i.e., pad expenses). This ethical dilemma, based on one of the items in our study 1 measure of UPSB, was different from that featured in study 2a, which was conducted with US participants, because requesting reimbursement for personal expenses is a more commonplace unethical behavior in China as opposed to obtaining personal perks, which is more commonplace in the United States (Liu, Liu, Hong, Brockner, Tam, & Li, Reference Liu, Liu, Hong, Brockner, Tam and Li2017; Lovett, Simmons, & Kali, Reference Lovett, Simmons and Kali1999). Although the underlying conflict of interest is the same, this adjustment reflects the practice in cross-cultural research of adapting construct operationalizations to local cultural practices (Gelfand, Raver, & Ehrhart, Reference Gelfand, Raver, Ehrhart and Rogelberg2002).
In the scenario, the company was described as having recently set up a large budget to be used by sales representatives for traveling to faraway regions to meet customers on-site so as to “grow the company’s client network and increase sales.” We then assessed participants’ likely behavior with two items: “How inclined would you be to claim all the expenses you can in this situation?” and “How likely are you to present all expense receipts for reimbursement that you can, including those for personal expenses incurred by you and work-related receipts from relatives and friends?” rated on a 7-point scale ranging from 1 (not at all) to 7 (very likely) (α = .86). We also assessed participants’ level of organizational goal commitment by administering the same, context-specific three-item scale of organizational goal commitment as used in studies 2a and 2b, albeit adapted to the present study (α = .79).
After participants had completed all other measures, we assessed participants’ level of moral identity by administering the five items from Aquino, Freeman, Reed, Lim, and Felps’s (Reference Aquino, Freeman, Reed, Lim and Felps2009) moral identity internalization subscale. Though the original scale consists of two dimensions, internalization and symbolization, we focused solely on internalization because it has received greater validation than symbolization and been widely employed as a freestanding measure of moral identity in the behavioral ethics literature (e.g., Mitchell, Vogel, & Folger, Reference Mitchell, Vogel and Folger2015; Taylor, Griffith, Vadera, Folger, & Letwin, Reference Taylor, Griffith, Vadera, Folger and Letwin2019). Indeed, as Jennings, Mitchell, and Hannah (Reference Jennings, Mitchell and Hannah2015) noted in a recent review, compared with symbolization, internalization is a more robust predictor of ethical and unethical behaviors. In completing the internalization measure, participants were asked to imagine a person who possesses nine moral attributes (e.g., caring, compassionate) and to then indicate, in five statements, the extent to which having these attributes is important to their sense of self, rated on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree) (α = .70). Sample items are “Being someone who has these characteristics is an important part of who I am” and “I strongly desire to have these characteristics.”
Results
Manipulation checks. Our manipulation of a collectivist orientation was successful. Participants assigned to the collectivist orientation condition reported caring significantly more about the interests of groups to which they belonged (M = 5.76, SD = 1.07) than did those in the individualist orientation condition (M = 4.72, SD = 1.36), t(157) = 5.34, p < .001, d = .85, whereas participants in the individualist orientation condition cared significantly more about their own personal self-interest (M = 4.22, SD = 1.44) than did those in the collectivist orientation condition (M = 3.56, SD = 1.58), t(175) = −2.74, p = .007, d = .44.
Hypothesis testing. To test hypothesis 1a, we conducted an analysis of variance (ANOVA) of participants’ likelihood of engaging in UPSB. Consistent with our survey data from studies 1 and 2a, we found that participants in the collectivist orientation condition reported a weaker intention to engage in UPSB (M = 4.33, SD = 1.52) than did those in the individualist orientation condition (M = 4.98, SD = 1.52), t(157) = −2.69, p < .01; F(1, 157) = 7.24, p < .01; η2 = .04.
We next tested our proposed explanation for and boundary condition hypotheses (i.e., hypotheses 2a and 3a). Consistent, once again, with our prior findings, we found that respondents in the collectivist orientation condition reported significantly higher organizational goal commitment (M = 5.72, SD = 0.82) than did those in the individualist orientation condition (M = 5.12, SD = 1.35), t(157) = 3.36, p < .01, d = .55. Using PROCESS (Hayes, Reference Hayes2013), we employed a bootstrapping procedure with twenty thousand random resamples to simultaneously test the proposed indirect and conditional indirect effects. As predicted, the proposed indirect relationship between collectivism and UPSB via organizational goal commitment was negative and significant (indirect effect = −.16, 95 percent CI = [−.37, −.02]). Thus hypothesis 2a was supported.
Furthermore, as shown in Figure 3, the relationship between organizational goal commitment and UPSB was stronger when moral identity was high (B = −.53, SE = .13, p < .01) than when it was low (B = .05, SE = .17, ns). Additionally, the indirect effect of collectivism on UPSB via organizational goal commitment was significant when moral identity was high (indirect effect = −.28, 95 percent CI = [−.55, −.08]), but not when moral identity was low (indirect effect = .06, 95 percent CI = [−.16, .33]). The difference between the indirect relationships was likewise significant (difference index = −.34, 95 percent CI = [−.72, −.07]). Thus hypothesis 3a was also supported.
Figure 3: Interaction of Goal Commitment and Moral Identity on UPSB (Study 3a)
Study 3b: UPOB
Participants and Design
For study 3b, a sample of 172 working adults (female = 54.7 percent; male = 45.3 percent; M age = 31–40 years; M work experience = 9.38 years) was recruited through Sojump, again in exchange for 10 RMB. Thirteen were dropped after incorrectly answering an attention check item, thereby leaving a final sample of 159 participants.
We employed the same procedure, manipulation of a collectivist orientation, and measurement of moral identity as in study 3a. All aspects of the two studies were identical, except that we used the same business decision-making scenario as in study 2b.
Results
Manipulation checks. Our manipulation was successful. Participants assigned to the collectivist orientation condition reported caring significantly more about the interests of groups to which they belonged (M = 5.77, SD = .91) than did those in the individualist orientation condition (M = 4.85, SD = 1.65), t(157) = 4.37, p < .001, d = .69, whereas participants in the individualist orientation condition cared significantly more about their own personal self-interest (M = 4.01, SD = 1.68) than did those in the collectivist orientation condition (M = 3.29, SD = 1.53), t(157) = −2.83, p < .01, d = .45.
Hypothesis testing. To test hypothesis 1b, we conducted an ANOVA and found, consistent with our survey data from studies 1 and 2b, that participants in the collectivist orientation condition reported being significantly more likely to engage in UPOB (M = 5.18, SD = 1.22) than did those in the individualist orientation condition (M = 4.63, SD = 1.61), t(157) = 2.41, p = .02, F(1, 157) = 5.79, p = .02, η2 = .04.
Next, we performed the same analyses as we performed in study 3a. We found that participants in the collectivist orientation condition reported a significantly higher level of organizational goal commitment (M = 5.94, SD = 0.75) than did those in the individualist orientation condition (M = 5.34, SD = 1.28), t(157) = 3.60, p < .01, d = .57. In addition, the indirect relationship between collectivism and UPOB via organizational goal commitment was positive and significant (indirect effect = .25, 95 percent CI = [.08, .51]). Thus hypothesis 2b was supported.
Finally, we found support for hypothesis 3b. As shown in Figure 4, the relationship between organizational goal commitment and UPOB was stronger when moral identity was low (B = .81, SE =.15, p < .01) than when it was high (B = .24, SE = .03, ns). Additionally, the indirect positive effect of collectivism on UPOB via organizational goal commitment was stronger when moral identity was low (indirect effect = .47, 95 percent CI = [.20, .85]) than when it was high (indirect effect = .12, 95 percent CI = [−.10, .39]). The difference between the indirect relationships was likewise significant (difference index = −.35, 95 percent CI = [−.68, −.10]).
Figure 4: Interaction of Goal Commitment and Moral Identity on UPOB (Study 3b)
Discussion
Study 3 replicates the double-edged effects of collectivism observed in our prior two studies while at the same time bolstering causal support for our main proposition. Results also provide evidence of a key boundary condition on the effects of collectivism on workplace unethical behaviors. Namely, the data reveal that an individual’s moral identity moderates both the indirect positive relationship between collectivism and UPOB and the indirect negative relationship between collectivism and UPSB. Thus study 3 addresses the limitations of our previous studies and offers support for our full theoretical model.Footnote 4
GENERAL DISCUSSION
Using a series of complementary studies, this research explored whether, how, and when collectivists exhibit bounded morality in the workplace. The results indicate that, owing to higher commitment to organizational goals, more collectivistic employees are less likely to engage in UPSB yet more likely to engage in UPOB. This pattern was observed across different operationalizations of UPSB and UPOB, as well as different national contexts, and held regardless of whether employees’ value orientation was measured or manipulated. We also found evidence that moral identity strengthens the negative indirect relationship between collectivism and UPSB, while weakening the positive indirect relationship between collectivism and UPOB. This latter finding suggests that the strength of an employee’s moral identity serves as a key boundary condition on the proposed double-edged effects of collectivism, via goal commitment, on employees’ workplace (un)ethical behavior. In the following section, we discuss the contributions and implications of these findings.
Theoretical Implications
The present research makes three contributions to the literature. First, our research extends conversations about collectivism and various types of socioethical dilemmas prevalent in business organizations. Our model and empirical results suggest that individuals with a prosocial value, collectivism, act ethically or unethically depending on the nature of their conflicts of interest. We suggest that this observation and our theoretical explanation of it not only helps clarify a long-standing confusion in the field but also creates a path for additional conversations in the future. For example, to the extent that the effects of collectivism vary depending on the nature of the decision-making context, it seems entirely reasonable that individuals with other prosocial values, such as loyalty to groups of which they are a part, might be subject to similar constraints. We hope that our efforts in the present research not only address the specific puzzle about collectivism upon which we sought to shed light but also serve as a template for exploring the paradoxical effects of other types of prosocial values.
Second, this research enriches and contextualizes the study of moral boundedness in research (Crimston et al., Reference Crimston, Bain, Hornsey and Bastian2016; Crimston et al., Reference Crimston, Hornsey, Bain and Bastian2018; Singer, Reference Singer2011). In particular, our results demonstrate that the ethical behaviors of collectivists are a function of their scope of moral concern—the broader the circle of moral concern is, the more collectivism invokes ethical behavior, whereas a narrower circle of moral concern leads collectivists to act more unethically. Nevertheless, we recognize that scholars do not regularly account for the impact of moral boundedness in their theories and models of decision-making, despite ample evidence that individuals are not wholly rational actors but are instead predictably morally bounded (e.g., Bazerman & Tenbrunsel, Reference Bazerman and Tenbrunsel2011). Just as we turned to moral boundedness to explain a set of contradictory findings in the collectivism literature, we suggest that others might look to moral boundedness as a possible explanation for unanswered questions in other, related areas as well. For example, Bolino and Grant (Reference Bolino and Grant2016) recently recognized a dual nature to prosociality and prosocial motives but stopped short of recognizing moral boundedness as a limiting factor. We believe that our work can spark conversations in this and many other areas.
Finally, this research illustrates the value of exploring the role of value orientations in ethical decision-making. Despite ethical implications inherent in cultural and personal values (Haidt & Joseph, Reference Haidt and Joseph2004; Schwartz, Reference Schwartz1994, Triandis, Reference Triandis1995), ethics researchers have generally focused on explicitly moral constructs, such as moral foundations (Graham et al., Reference Graham, Nosek, Haidt, Iyer, Koleva and Ditto2011), moral ideologies (Kish-Gephart et al., Reference Kish-Gephart, Harrison and Treviño2010), and moral development (Treviño, Reference Treviño1986), as the primary antecedents of ethical decision-making and behavior. Cultural studies of ethics, in contrast, have relied on categorical variables (e.g., nationality and race) as proxies of culture or focused on effects of value differences on ethical attitudes and judgments without exploring their linkages to workplace behaviors. Ethics research benefits from a theoretical model that sheds light on how and why value orientations influence workplace ethical conduct, and we hope that this work inspires other such approaches in this area. For example, future research might include both moral and cultural antecedents to compare, contrast, or control for their distinct effects on individuals’ ethical decisions and behaviors, both in the workplace and beyond.
Limitations
The positive features of this research notwithstanding, its limitations are worth addressing. First, although the present research focuses on organizational goal commitment as the psychological mechanism that links collectivism to UPSB and UPOB, future research could explore alternative mediating mechanisms, such as shared responsibility or organizational identification. Second, evidence for the predicted negative relationship between collectivism and UPSB was stronger in our initial field study and concluding vignette experiment than it was in study 2. Although the sign of this relationship was the same, our confidence in the generalizability of the model across national contexts would nevertheless be strengthened by additional research replicating our model in more individualistic cultures. Third, studies 2 and 3 consisted of online crowdsourced samples. Whether their results would generalize to other samples is an open question. Furthermore, they employed vignettes involving ethical decision-making, rather than having participants respond to actual conflicts of interest in their workplace or placing them within a laboratory experiment in which they encountered an actual ethical dilemma. While such methodology is quite common in behavioral ethics research (Mitchell, Reynolds, & Treviño, Reference Mitchell, Reynolds and Treviño2020), and our concerns about it are significantly alleviated by the results of our initial field study, additional research is nevertheless needed to evaluate whether our full set of results would also emerge when employing additional methodologies (e.g., laboratory experimentation).
Practical Implications
This research has notable implications for managers and organizations. Although UPSB and UPOB can both be costly for organizations and society (Cialdini, Petrova, & Goldstein, Reference Cialdini, Petrova and Goldstein2004), management and employees alike are typically more cognizant of the costs of the former than of the latter. It is troubling to note, therefore, that in study 1, employees’ likelihood of enacting UPOB was significantly higher than their engagement in UPSB. Though both were low, such differences suggest that organizational ethical norms regarding UPOB may be more relaxed than those regarding UPSB and that employees may view the former as more justifiable than the latter. Organizations would thus be well served to begin taking seriously the costs of UPOB to themselves and society.
To this end, the present research suggests a few possible interventions for organizations to reduce UPOB. To begin with, organizations need to be aware of and guard against the potential downsides of prosocial concepts like collectivism, loyalty, and organizational identification, all of which have traditionally been viewed as desirable for motivating organization-benefiting behaviors. Organizations can take at least two types of measures to mitigate such downsides. One is to instill a balanced multistakeholder view and establish corporate policies to protect and advance the well-being of all stakeholders. Focusing exclusively on organizational well-being may lead to lapses in fulfilling corporate social responsibility toward internal and external stakeholders. The other is to promote ethical standards and establish ethical bottom lines to counteract or balance financial bottom lines in resolving conflicts of interest among different stakeholders. Relatedly, in view of the important role of goal commitment, organizations could use training, development, and performance metrics to highlight to employees the importance of larger social goods, such as corporate social responsibility and an improved natural environment. Finally, in view of the regulating effect of moral identity, organizations should take into consideration moral character when making selection and promotion decisions and emphasize the development of moral character in training and development. Each of these interventions represents a potentially promising route for mitigating UPOB.
CONCLUSION
This research extends knowledge by addressing the paradoxical effects of collectivism, a prototypical prosocial construct, on employee (un)ethical behaviors. Our results demonstrate that the relationship between collectivism and ethical behavior is much more nuanced and intriguing than it might appear. We are excited by the opportunities this research offers.
Acknowledgments
The work in this article was supported by the National Natural Science Foundation of China (project 72272044). The authors declare that this research was conducted in accordance with ethical standards and that we do not have any commercial or associative interest that represents a conflict of interest in connection with the work.
APPENDIX A: VIGNETTE SCENARIOS FOR STUDY 2A AND STUDY 2B
Study 2a (UPSB)
Imagine you work as a purchasing agent for a fast growing manufacturing company, Digital Appliances (DA). Current economic conditions in digital electronics favor buyers of materials and will continue for a few more years. This is great news for DA because it means great opportunities to renegotiate more favorable terms with its existing suppliers so that DA will be even more competitive in the industry. In the past, because material supply was uncertain and prices kept going up, DA signed many long-term contracts that lasted five years or more so as to ensure continuity of material supply at stable price. Considering the improved market conditions, however, DA has decided to require that all purchasing agents reevaluate their current suppliers in an effort to shorten the contract lengths to three years or less. The goal of this initiative is to reduce the average contract length so as to increase the company’s flexibility to select new and better suppliers. Such changes will significantly increase the workload of purchasing agents but will help the company to reduce costs and increase profit margins.
All purchasing agents must submit a list of candidates for contract renegotiation to the purchasing department within a week. Tomorrow is the deadline, and your manager just reminded you to submit your list in time. While you have already identified a list of potential companies, there is one supplier that you are still debating whether or not to put up for renegotiation. Normally, this company would be a clear candidate for renegotiation because the quality and price of their materials, while acceptable, is not outstanding. However, as an old supplier for over 10 years, it is among your most favorite partners, one that you particularly enjoy working with. Whenever you visit, the company’s president makes time to personally meet with you to express his appreciation for your business and your cooperation and support. This supplier is also among the few that have a very generous VIP policy for long-term purchasing agents, like yourself. For example, it offers a free five-year sports club membership to you, which also allows for bringing two guests. If you submit this supplier for contract renegotiation, you will succeed in obtaining a short-term contract. However, you are seriously concerned that you will lose the VIP special treatment, such as the free sports club membership and other perks, and that your work with the company will be less comfortable and less enjoyable.
You thus need to decide whether or not to put this supplier on the list for contract renegotiation. Note that this decision is entirely up to you because your department manager generally follows the suggestions of frontline purchasing agents.
Study 2b (UPOB)
Imagine that you are a sales agent for a company that does business internationally. Your main task involves communicating with international buyers and trying to win sales orders for your firm. As the domestic market has become saturated, the company has decided to focus on the development of its international business. The company’s main goal is to cultivate international business so that it becomes the company’s primary center of growth in business and profit. Successful achievement of this goal is critical to the future development of the company.
Recently, you have been negotiating with a potential buyer based in Mexico. The order is about 7 million dollars, the largest one that your company has won. You estimate that your company can make a profit of $700,000 out of this business. You are quite happy with the way negotiations have been going. Although there are two other firms competing for the same order, the buyer really likes your company’s products, and your prices are fairly competitive. You are quite confident that you will win the order.
Yesterday you heard back from the potential buyer. He told you that he hasn’t made up his mind. He suggested the quality of your company’s product is very competitive but the price is higher than that of the other two companies. However, he also hinted that he would give your company the order if your company is willing to offer him around $7,000 for his extra effort. The requested amount of money would be a very small portion of your company’s eventual profit, i.e., around 1%.
As a salesperson, you have the power to make the decision on behalf of your company, and you know that the payment can be made in the form of business expenses. You have to get back to the buyer by tomorrow to tell him your final decision. What do you decide to do?
APPENDIX B: VIGNETTE SCENARIOS FOR STUDY 3AFootnote 5
Imagine yourself working as a sales representative for a pharmacy company. Sales reps earn their remuneration partly from a fixed salary and partly through commissions on sales. The success of your company depends largely on institutional buyers of drugs, such as hospitals. Successful sales reps tend to be those capable of building long-term, trusting relationships with different hospitals. The company, therefore, encourages sales reps to travel to hospitals all over China to personally meet with hospital administrators to promote products and build relationships.
Recently, the company decided to set up a huge budget to subsidize sales reps for travel expenses. The initiative is called “the visit and promotion project.” The goal of the visit and promotion project is to create a specific fund that is designated for supporting salespeople to travel to faraway hospitals to meet customers on-site so as to promote proactive sales activities. On average, each sales rep is allowed to reimburse RMB 50,000 per year in expense receipts from legitimate business entities (e.g., registered transportation companies, hotels, and restaurants). For top sales reps who generate large sales, more reimbursement can be claimed in proportion to the amount of sales made that exceeds average sales, again, as long as legitimate work-related travel expense receipts are presented to the financial department.
In the past year, you have been one of the top sales reps in your company. Apart from having excellent sales skills, your success is also due to your network of family members and friends who happen to be well connected to some large hospitals. Through their connections, you were able to make large sales, even though you yourself did not do much traveling or in-person visits to these hospitals. Your large sales in turn allow you to reimburse up to a total of RMB 70,000 this year, as long as you can present all expense receipts. However, you currently have only RMB 40,000 of receipts for expenses that were truly related to your hospital visits. You do have RMB 10,000 in personal expenses incurred while traveling for work, and your relatives and friends can easily provide you with another RMB 20,000 worth of legitimate business-related receipts from their own work.
Today is the deadline for processing all work-related travel reimbursement. The administrative staff just called you asking for your expense receipts. You heard that most sales reps in the company have submitted enough expense receipts that enable them to reimburse the maximum amount of subsidies allowed to them. You have to decide whether or not to claim your full allowed amount by presenting all of the receipts available to you, including those for your own personal expenses and those of your relatives and friends.
Chao C. Chen is a distinguished professor of management and global business at Rutgers Business School, Rutgers University. His research interests include cross-cultural management, Chinese management, organizational justice, leadership, and business ethics. His business ethics works include prosocial motivation, moral boundedness, ethical climate, ethical voice, abusive supervision, and unethical pro-organizational behavior. His work has been published in several top organization and management journals, such as the Academy of Management Journal, Academy of Management Review, Administrative Science Quarterly, the Journal of Applied Psychology, Organization Science, and the Journal of International Business Studies.
Oliver Sheldon is an associate professor of management and global business at Rutgers Business School, Rutgers University. His work seeks to better understand motivational and relational triggers of interpersonal competition, conflict, and unethical behavior in groups, with the aim of shedding light on how organizations might improve coordination, collaboration, and ethical decision-making among employees. His work has been published in several management, social psychology, and applied psychology journals, such as Organization Science, the Journal of Applied Psychology, Organizational Behavior and Human Decision Processes, and the Journal of Experimental Social Psychology.
Mo Chen ([email protected], corresponding author) is an associate professor at the School of Management at Harbin Institute of Technology. His research interests include business ethics, leadership, proactivity, and the microfoundation of corporate social responsibility. His work has been published in several management and ethics journals, such as the Journal of Applied Psychology, the Journal of Business Ethics, and Management and Organization Review.
Scott J. Reynolds is a professor of business ethics and the Michael G. Foster Endowed Professor at the Michael G. Foster School of Business at the University of Washington. His research interests include topics related to moral decision-making, such as moral awareness, moral attentiveness, and moral intuition, and his work has appeared in several top management and ethics journals, such as the Academy of Management Journal, the Journal of Applied Psychology, Organizational Behavior and Human Decision Processes, Personnel Psychology, and Business Ethics Quarterly.
