Blame the shepherd not the sheep: Imitating high-ranking transgressors mitigates punishment for unethical behavior

Do bad role models exonerate others’ unethical behavior? Based on social learning theory and psychological theories of blame, we predicted that unethical behavior by higher-ranking individuals changes how people respond to lower-ranking individuals who subsequently commit the same transgression. Five studies explored when and why this rank-dependent imitation effect occurs. Across all five studies, we found that people were less punitive when low-ranking transgressors imitated high-ranking members of their organization. However, imitation only reduced punishment when the two transgressors were from the same organization (Study 2), when the transgressions were highly similar (Study 3), and when it was unclear whether the initial transgressor was punished (Study 5). Results also indicated that imitation affects punishment because it influences whom people blame for the transgression. These findings reveal actor-observer differences in social learning and identify a way that unethical behavior spreads through organizations. Major scandals caused by corporate executives receive a great deal of attention from the media and scholars alike, but the aggregated cost of relatively minor transgressions committed by the average employee is substantial. Asset misappropriations, such as expense report manipulation and inventory theft, are by far the most common type of fraud within organizations (Association of Certified Fraud Examiners, 2014). Expense report fraud alone costs companies in the United States $1 billion annually (J.P. Morgan Chase, 2011). Employee theft of retail goods causes $15.1 billion in lost revenue, which is a larger loss than is caused by shoplifting (National Retail Federation, 2012). Tips from employees remain the most effective means of detecting these types of fraud (Association of Fraud Examiners, 2014; see also Weaver, Treviño, & Cochran, 1999). However, employees also can create and maintain a culture that “normalizes” bad behavior. For example, it is an open secret in some organizations that employees pad their expense reports by ten percent or more (Strout, 2001). Currently, it is poorly understood how people come to tolerate unethical behavior in some instances more than in others. What increases the likelihood that people will look the other way rather than punish those who violate the rules? Behavioral ethics research has tended to examine ethical transgressions as isolated, one-off occurrences, rather than in relation to other transgressions that have occurred within the organization (Ashforth, Gioia, Robinson, & Treviño, 2008; Greve, Palmer, & Pozner, 2010; Moore, 2009). Recent work, however, has begun to focus on how bad behavior propagates through organizations by exploring social contagion as a contributor to abusive supervision (Brown, Treviño, & Harrison, 2005; Mawritz, Mayer, Hoobler, Wayne, & Marinova, 2012), anti-social employee behavior (Mayer, Kuenzi, & Greenbaum, 2010; Robinson & O’Leary-Kelly, Bauman et al., Organizational Behavior and Human Decision Process (in press) 1998), levels of deviance across workgroups (Mayer, Kuenzi, Greenbaum, Bardes, & Salvador, 2009), collective acts of corruption that benefit the organization (Smith-Crowe & Warren, 2014), as well as exemplary behaviors (Brown et al., 2005; Mayer, Nurmohamed, Treviño, Shapiro, & Schminke, 2013; Schaubroeck et al., 2012). Placing greater emphasis on understanding the connections among unethical behaviors enacted by different people within organizations as they unfold over time has identified important processes that are often underspecified in models of individual ethical decision making. In the spirit of this emerging area of research, we examine how prior instances of unethical behavior change how people evaluate subsequent transgressions and punish imitators. Our contention is that people are less apt to punish bad behavior when transgressors imitate those who outrank them compared to when they imitate peers or commit a transgression no one else committed recently. That is, we expect there to be a rank-dependent imitation effect on punishment. As we explain below, social learning theory (Bandura, 1977, 1986; Robinson & O’Leary-Kelly, 1998) and theories of blame (Malle, Guglielmo, & Monroe, 2014) suggest that a rank-dependent imitation effect should emerge because bad behavior by high-ranking others affects how observers assign blame, which in turn affects punishment. Moreover, psychological research on descriptive norms suggests that high-ranking individuals’ behavior can alter observers’ perceptions of what is typical for group members (Cialdini, Reno, & Kallgren, 1990), which also may mitigate punishment. In sum, we expect that bad role models at least partially exonerate others' subsequent transgressions of the same kind in the eyes of observers. Our research contributes to the literature in three main ways. First, it contributes to the literature on retributive justice by examining whether people become more tolerant of bad behavior after it has been modeled by higher-ranking members of organizations. Prior research has largely focused on relatively stable characteristics of punishers, transgressors, and contexts (e.g., Arvey & Jones, 1985; Butterfield, Treviño, & Ball, 1996; Podsakoff, 1982). Our research is the first to consider more transient features of situations (e.g., recent misconduct) as a unique influence on culpability and punishment. As in prior research on retributive justice (e.g., Darly & Pittman, 2003; Fragale, Rosen, Xu, & Merideth, 2009; Okimoto & Wenzel, 2014), we focus on lay observers’ reactions to transgressions. Although leaders and supervisors have the formal authority and responsibility to punish undesirable behavior, employees often scold, sabotage, or ostracize their coworkers for misbehaving (e.g., Barker, 1993; Gromet & Okimoto, 2014; Hollinger & Clark, 1982; O’Reilly & Aquino, 2011; Struthers, Miller, Boudens, & Briggs, 2001), and this type of punishment from peers is a very effective deterrent of unethical behavior (e.g., Hollinger & Clark, 1982; Tittle, 1977; Tittle & Logan, 1973; Zimring & Hawkins, 1973). Second, our research extends recent work in behavioral ethics that has begun to address how prior behaviors affect subsequent behaviors across levels of the organization. Our focus is novel because most other work in this area examines people’s propensity to commit unethical behavior, whereas we investigate when and how prior transgressions change people’s evaluations and responses to others’ unethical behavior. Therefore, we offer a new and complementary perspective on unethical contagion within organizations because we directly examine how prior transgressions— especially those committed by higher-ranking members of organizations—change the environment in which subsequent transgressions occur. If, as we suggest, people are less apt to punish those who imitate unethical behavior committed by higher-ranking members of their organization, then social systems may become less responsive to certain transgressions over time, which may disinhibit others from acting similarly. This dynamic represents one mechanism through which unethical behavior may become prevalent in organizations. Third, it is well-established that modeling influences others’ propensity to act similarly (e.g., Bandura, 1986; Gino, Ayal, & Ariely, 2009; Robinson & O’Leary-Kelly, 1998), but very little is known about whether, when, and why third-party observers take bad role models into account when evaluating and responding to others’ behavior. That is, we investigate whether people take modeling and social learning processes into account when evaluating individuals who followed a bad role model (i.e., third-party judgments) rather than examine how modeling and social learning influences individuals contemplating an action (i.e., second-party judgment and behavior). Thus, the current research has implications for social learning theory as well. THEORETICAL BACKGROUND Recent work on ethical leadership and contagion focuses on sequences of unethical behaviors in organizations and has sought to understand how one individual’s behavioral output becomes an input to other individuals’ judgments and behaviors (e.g., Brown et al., 2005; Mayer et al., 2009; SmithCrowe & Warren, 2014). This work builds on insights from social learning theory, which emphasizes that people learn how to behave in a given situation by observing others Bauman et al., Organizational Behavior and Human Decision Process (in press) (Bandura, 1977, 1986). Leaders who model bad behavior embolden their subordinates to engage in bad behavior (Brown et al., 2005; Mawritz et al., 2012; Mayer et al., 2010; Mayer et al., 2009). Modeling can also exert an influence up or across the organizational hierarchy as well (e.g., Gino et al., 2009; Robinson & O’Leary-Kelly, 1998; Zey-Ferrell & Ferrell, 1982). In sum, research that draws from social learning theory has begun to articulate processes that explain how unethical behavior within organizations unfolds over time. Although modeling and social learning processes are well established as antecedents of behavior, research has not considered whether third-party observers take bad role models into account when evaluating and responding to transgressions. In the sections that follow, we discuss when and why we expect people to punish misbehavior differently depending on whether a higher-ranking member of the organization has recently committed a similar transgression. We argue that unethical behavior from higher-ranking individuals—but not peers—influences perceived descriptive norms for behavior, alters attributions of blame, and reduces punishment.