Voice or silence: antecedents of whistleblowing intentions

Abstract

Purpose

This study aims to examine the effects of peer ethical behavior and individual differences in valuation of fairness vs loyalty on whistleblowing intentions in academic settings. This study also tests the underlying mechanism responsible for the effects of peer behavior on reporting intentions, namely, fear of reprisal.

Design/methodology/approach

A survey was conducted with 947 undergraduate students. The model was tested using ordinary least squares regression models followed by bootstrapped mediation analyses.

Findings

Results showed that the effects of peer ethical behavior on whistleblowing intentions are mediated by fear of retaliation. Moreover, the findings indicated that, for low-severity transgressions, there is an interactive effect between fear of retaliation and endorsement of fairness over loyalty on whistleblowing intentions.

Research limitations/implications

When the misconduct is seen as minor, a potential whistleblower may understand that the expected costs outweigh the possible benefits of blowing the whistle. In such situations, higher fear of retaliation would undermine the effects of individual’s endorsement of fairness over loyalty on reporting intentions.

Practical implications

As the social environment significantly affects someone’s whistleblowing intentions, there should be visible efforts to improve and to foster an ethical infrastructure in organizations.

Social implications

As fear of retaliation by peers is one of the most important determinants affecting the decision to report misconduct in general, there must be serious efforts from leaders to mitigate any threat of retaliation to those who come forward.

Originality/value

This work contributes to the discussion about individual and situational antecedents of whistleblowing. More importantly, it sheds light on one potential boundary condition for the influence of the fairness–loyalty tradeoff on whistleblowing decisions: severity of the transgression. The findings provide initial evidence that, for low-severity transgressions, fear of retaliation weakens the positive effects of one’s moral compass in terms of preference for fairness over loyalty on whistleblowing intentions.

Citation

Iwai, T., Yeung, L. and Artes, R. (2021), "Voice or silence: antecedents of whistleblowing intentions", RAUSP Management Journal, Vol. 56 No. 2, pp. 186-201. https://doi.org/10.1108/RAUSP-06-2020-0126

1. Introduction

Whistleblowers play a central role in uncovering frauds and organizational wrongdoing in general. Recent data from cases of occupational fraud in companies shows that tips provided by individuals who observe misconduct are responsible for the detection of half of the cases of frauds perpetrated in Latin America (Association of Certified Fraud Examiners, 2020). Moreover, by reporting misconduct in place, whistleblowers can help organizations to avoid financial losses (e.g. employee embezzlement), lawsuits filed (e.g. employee discrimination or moral assault cases) and reputational damages (Near & Miceli, 2016). In the fiscal year of 2017, whistleblowers helped to recover almost $4bn in fraudulent charges in the USA (U.S. Department of Justice, 2017). Given its importance, understanding the factors that encourage or prevent individuals from reporting observed unethical conduct is of critical importance to theory and practice on ethics. Identifying these factors may have concrete consequences in the formulation of practical organizational policies that may promote whistleblowing.

Although some consider the act of whistleblowing to be driven by a sense of moral duty (Watts & Buckley, 2017) and others compare whistleblowers to saints (Avakian & Roberts, 2012), individuals who decide to blow the whistle are more likely to be seen as rats than heroes by members of their communities (Grant, 2002). Because of the many personal risks they undergo by deciding to report misconduct, many individuals choose to remain silent (Kish-Gephart, Detert, Treviño, & Edmondson, 2009; Morrison, 2014), which raises the question of what leads some individuals to blow the whistle while others do not.

To tackle this issue, extant research has been arguing for the role of individual and contextual determinants in the decision to report unethical conduct. Regarding individual factors, studies have investigated the role of demographic characteristics (e.g. gender, age and education) in whistleblowing (Cassematis & Wortley, 2013; Mesmer-Magnus & Viswesvaran, 2005) as well as the influence of other personal characteristics such as personality traits (Culiberg & Mihelič, 2017) and value orientations (Park, Blenkinsopp, & Park, 2014). About situational factors, studies have focused on characteristics of the wrongdoing (Near & Miceli, 1996) or the organizational context, such as organizational culture and climate (Kaptein, 2011), threat of retaliation (Mesmer-Magnus & Viswesvaran, 2005) and supervisor and coworker support (Mayer, Nurmohamed, Treviño, Shapiro, & Schminke, 2013). This literature has shown that both individual and situational factors are determinants in whistleblowing (some argue that more on intentions than on effective decisions to act), but there is also sound evidence that situational variables tend to be more influential than individual ones (Cassematis & Wortley, 2013; Mesmer-Magnus & Viswesvaran, 2005, for a meta-analytic review; Near & Miceli, 1996). In Brazil, the literature on corporate whistleblowing has focused on the importance of cultural aspects (Behrens, 2015; Sampaio & Sobral, 2013). Building on this tradition of explaining human behavior as the result of personal and contextual variables, this work aims to investigate the joint effects of the individual’s moral values and the immediate social environment in which he or she is embedded. More specifically, we examine individual differences in preferences for fairness vs loyalty and peer ethical behavior on intentions to blow the whistle in academic settings. In our view, this topic has not been dealt in depth by the aforementioned literature. Moreover, we explore the mediating effect of fear of retaliation on the relationship between peer behavior and whistleblowing.

We chose to investigate whistleblowing in academic environment, specifically among undergraduate students in a private college, because fraud is a pervasive problem at most academic institutions (MacGregor & Stuebs, 2014; McCabe, Trevino, & Butterfield, 2001) and it may lead to the normalization of unethical behavior in other domains (Fida, Tramontano, Paciello, Ghezzi, & Barbaranelli, 2018). Furthermore, some suggest that misconduct in vocational education can influence future ethical behavior at the workplace (McCabe, Butterfield, & Trevino, 2012). Although this work does not aim at focusing on the specificities of academic whistleblowing as compared to other environments, the contributions of these authors are crucial to the understanding of our problem.

In this paper, we analyze less severe misconducts – such as cheating during exams and signing presence list for others – but also very severe ones – such as bribing employees to access exam content in advance and hacking the academic system for personal gains.

Within this scope, this paper provides the following contributions. First, we add to the discussion of academic dishonesty by examining the influence of individual and situational antecedents of whistleblowing not already tested in academic settings, namely, peer behavior (in our case, students) and valuations of fairness vs loyalty. Moreover, by including fear of retaliation in our analysis, our work sheds light on the psychological mechanism by which the immediate social environment influences the decision to blow the whistle or not. Additionally, we extend literature on whistleblowing by investigating how fear of reprisal from one’s peers may interact with valuations of fairness vs loyalty on whistleblowing intentions.

2. Theory and hypotheses

2.1. Peer behavior and fear of retaliation

Given the central importance of whistleblowing, defined as the “disclosure by organization members (former or current) of illegal, immoral, or illegitimate practices under the control of their employees, to persons or organizations that may be able to effect action” (Near & Miceli, 1985, p. 4), much work has been done to identity conditions under which whistleblowing is more likely to occur. In the literature of behavioral ethics in general, the social environment has been identified as one of the most powerful forces in fostering ethical conduct (Tenbrunsel & Chugh, 2015). In academic settings, although many contextual factors may play a role (such as honor code, sanctions and monitoring systems), peer behavior (i.e. other students’ behavior) is considered one key determinant to explain academic cheating (McCabe, Butterfield, & Treviño, 2006). This is because much of the learning occurs via vicarious processes, that is, people do not have to be directly reinforced to learn behaviors; rather, they learn through the observations of others’ behaviors (Bandura, 1986). Thus, when observing others cheating and getting ahead, individuals are more likely to display the same cheating behavior. By the same token, if individuals see whistleblowers suffering reprisal for speaking up, then whistleblowing should be less likely to occur.

Additionally, not only individuals use peers as potential role models for their own behavior, but they also observe others to gather signals and cues of what is considered appropriate. As such, peer ethical behavior may also provide normative support for whistleblowing decisions. Consistent with it, social information processing theory (Salancik & Pfeffer, 1978) suggests that under conditions of uncertainty about acceptable behavior, individuals look for cues about what to do in their social environment. According to the authors, the social context provides valuable information about what behaviors, attitudes and opinions one should display and follow. Actors from the immediate social circle, such as coworkers and immediate supervisors, exert particularly important influence because individuals are motivated to follow and agree with their peers to fit in. In line with this reasoning, Mayer et al. (2013) have found support for the joint effects of leader and peer ethical behavior on employees’ willingness to report unethical conduct in work environment. Based on these arguments, we hypothesize the following:

H1.

Peer ethical behavior is positively related to whistleblowing intentions.

Despite the many organizational benefits associated with whistleblowing, almost half of the individuals aware of misconduct remain silent (Fredin, 2012). That number is not surprising considering that coming forward to report organizational wrongdoing is risky (Near & Miceli, 2016). The literature on voice and silence presents consistent evidence that a safety calculus is crucial to the decision of voice and silence (Morrison, 2014). Extant research has shown that most whistleblowers face some form of retaliation from colleagues or supervisors after disclosing misconduct (Mesmer-Magnus & Viswesvaran, 2005; Park, Bjørkelo, & Blenkinsopp, 2020). For instance, Rothschild and Miethe (1999) identified a retaliation rate of nearly two-thirds among actual whistleblowers, ranging from more formal undesirable actions against the person who speak up (such as termination, demotion, unfavorable job performance evaluations, involuntary transfer, assignment of unmanageable tasks and professional blacklisting) to more informal and unofficial ones (e.g. social ostracism and bullying).

Given the potential threats of blowing the whistle, individuals may choose for what has been called defensive silence, a type of silence driven by the fear of the anticipated negative consequences for the self (Kish-Gephart et al., 2009). Such fear is triggered by cues in one’s environment signaling the existence of some threat. According to the informal rules and norms of the group about how to deal with misconduct from one of its members (Trevino & Victor, 1992), individuals then may choose to withhold information regarding another person’s misconduct. In this line of reasoning, the immediate social environment in which the person is embedded may activate the fear of being punished for speaking up. For instance, Mayer et al. (2013) have found evidence that, in work environments, the effects of supervisory ethical leadership and coworker behavior on reporting unethical conduct internally occur via fear of retaliation.

Although work-related retaliation (such as terminations or demotions) is absent in academic settings, whistleblowers may still face ostracism, name-calling and other forms of social sanctions from their academic peers. For these reasons, if individuals perceive their academic peers as more ethical, they will be less likely to be perceived as a snitch or a tattletale. Also, they will feel less threatened to be reprimanded by the group and that will increase their likelihood of engaging in voice. As such, fear of retaliation may be the underlying mechanism by which peer behavior influence whistleblowing intentions in academic environment. Following this, we expected that:

H2.

The effect of peer ethical behavior on whistleblowing intentions is mediated by fear of retaliation.

2.2 Interplay between the whistleblower’s dilemma and the social environment

When deciding whether or not to disclosure another person’s misconduct, individuals face what has been called the whistleblower’s dilemma (Waytz, Dungan, & Young, 2013), in which two important moral values, namely, fairness and loyalty, may be in conflict. Although both values guide people in judging whether a behavior is appropriate or not (Graham et al., 2011; Hildreth, Gino, & Bazerman, 2016), they clash when it comes to whistleblowing decisions. While norms of fairness require treating everyone equally, loyalty norms entail preferential treatment for in-group members over others. In whistleblowing decisions, by reporting other’s misdeeds, individuals are acting in the name of justice and the benefit of the society more broadly. But at the same time, their act may be considered a betrayal to the closer group. As such, there is a tradeoff between concerns regarding the well-being of one’s immediate group and concerns for the well-being of others that goes beyond one’s group (Dungan, Young, & Waytz, 2019). As Dungan, Waytz, and Young (2015, p. 129) put it: “from one perspective, whistleblowing is the ultimate act of justice, serving to right a wrong. From another perspective, whistleblowing is the ultimate breach, a grave betrayal.”

Extant evidence has shown that loyalty leads to corruption in some domains (Hildreth et al., 2016) as well as prior research has found that the endorsement of fairness over loyalty is positively associated with whistleblowing in general (Waytz et al., 2013). In this paper, we aim to test the effects of this whistleblowing dilemma on willingness to report wrongdoing in academic settings. Moreover, we are not aware of empirical studies testing the interplay between the fairness-loyalty tradeoff and the fear of reprisal in whistleblowing intentions. Examining this interactive effect would be important, because the effects of the endorsement of fairness over loyalty on whistleblowing decisions may depend on the degree to which one fears being socially reprimanded by their peers. For instance, prior work has provided evidence that the risk of exclusion from the group predicts employees’ engagement in pro-group unethical behavior (Thau, Derfler-Rozin, Pitesa, Mitchell, & Pillutla, 2015).

Although some people value fairness more than loyalty to a group, social inclusion is an important universal need (Baumeister & Leary, 1995), which means that to a certain degree everyone cares about affiliation to a group and their inclusionary status in the group. Therefore, we posit that fear of retaliation can suppress the positive effect of preference for fairness over loyalty norms in whistleblowing intentions. Based on the aforementioned arguments, we hypothesize the following:

H3.

The preference for fairness over loyalty is positively related to whistleblowing intentions.

H4.

There is an interactive effect of preference for fairness over loyalty and fear of retaliation on whistleblowing intentions such that the effect of preference for fairness over loyalty on willingness to blow the whistle is weakened when fear of retaliation is higher.

3. Method

3.1 Sample

Data for analysis were gathered from a sample of students from a Brazilian business school. We targeted this sample for two reasons. First, we intended to examine antecedents of whistleblowing intentions in academic environment. Also, we chose business students because prior research has shown that they cheat more than their nonbusiness-student peers (McCabe et al., 2006). Students from the 1st, 2nd and 3rd years of the courses were invited to participate in the study as volunteers. They were informed that the study was of a scientific nature and they would not be identified at any time, as the data collected was anonymous. Out of 1,076 participants, 41% were freshmen and 32% were female. The mean age was 19.6 years old (standard deviation = 1.6 years). A total of 32 participants did not complete the entire survey. The final sample was of 947.

3.2 Measures

Fairness-loyalty trade-off. The first measure was based on the Moral Foundations Questionnaire (Graham et al., 2011). On a six-point scale, six items were used to assess participants’ opinions about the importance of fairness and loyalty for their judgments of right and wrong. Sample items included: “whether or not someone showed a lack of loyalty” and “whether or not someone had his or her rights denied.” We computed a composite values score by averaging the three fairness items and the three loyalty items separately and then subtracting the loyalty score from the fairness score. The resulting values were standardized to obtain a variable [Fairness-loyalty trade-off (“FL”)] with mean 0 and standard deviation 1. The higher the score, the more an individual values fairness over loyalty.

Peer ethical behavior (“Peers”). We used three items adapted from Mayer et al. (2013) to assess peer ethical behavior. A sample item is: “my peers support me in following my school’s standards of ethical behavior.” We averaged the three items to compute the “Peers” score. All items were measured on a seven-point Likert scale (1 = strongly disagree and 7 = strongly agree).

Fear of retaliation (“Ret”). Two items adapted from Mayer et al. (2013) were used to assess fear of retaliation on a seven-point Likert scale (1 = strongly disagree and 7 = strongly agree): “if I report a dishonesty act, I will be seen as a snitch by my peers” and “if I report a dishonesty act, I will be seen as a trouble-maker by my peers.” We averaged the two items to compute the “Ret” score.

Whistleblowing intentions. To assess willingness to blow the whistle, eight scenarios of academic cheating with different levels of severity adapted from McCabe and Trevino (1993) were presented to participants. For each situation, respondents indicated their whistleblowing intentions on a seven-point scale ranging from 1 (very unlikely) to 7 (very likely).

For each scenario, participants were also asked how likely they would be to blow the whistle if the transgressor was: “an acquaintance you barely interact” or “a close friend.” We averaged all the 16 responses (2 different transgressors for 8 different scenarios) to compute an overall whistleblowing score (“OWS”). An exploratory factor analysis with varimax rotation was used to uncover whether respondents perceived differently the seven cheating behaviors. The analysis yielded a two-factor solution with 77% of explanation of the total variance of the data. Items were retained if they loaded higher than 0.50 on a single factor. As Table 1 indicates, the two factors distinguish between high (factor 1) and low (factor 2) severity transgressions. For this reason, we also computed a whistleblowing score for high (“HWS”) and low (“LWS”) severity transgressions.

Control variable. Research has shown that not only is moral disengagement a strong predictor of a myriad of unethical behaviors (Moore, 2015; Moore, Detert, Treviño, Baker, & Mayer, 2012), but it also influences observers’ reactions to unethical situations (Aquino, Reed, Thau, & Freeman, 2007). Thus, we controlled for it using the 24-item moral disengagement propensity measure (“MD”) from Detert, Treviño, and Sweitzer (2008) on a seven-point scale (1 = strongly disagree and 7 = strongly agree).

Figure 1 represents the expected relations among the variables as specified by our hypotheses. The letter y stands for the whistleblowing score (OWS, HWS or LWS).

3.3 Statistical model

Based on Figure 1 and using MD as a control variable, we estimated regression models to measure the effects of “Peers,” “Ret” and “FL,” and the interaction (moderation) effect between “Ret” and “FL” over the “OWS,” “HWS” and “LWS.” The coefficients of the models are associated with the connections illustrated in Figure 1. We also evaluated the effect of “Peers” over “Ret” using “MD” as a control variable. The econometric models are given by

(1) y=β0+β1 Peers+β2 Ret+β3 FL+β4 Ret*FL+β5 MD+ε1,

(2) Ret=α0+α1 Peers+α2 MD+ε2,

where y denotes the dependent variable (OWS, HWS or LWS), β and α are the parameters of the models and ε₁ and ε₁ represent the errors of the models. Because of the large sample size (947) results by an ordinary least squares estimation of these parameters are robust to the lack of normality of the errors. To control for heteroscedasticity, we used Huber–Whites’s robust standard errors.

The evaluations of the indirect effect of “Peer” over “OWS,” “HWS” and “LWS” were based on bootstrap estimates of the 95% confidence interval of the conditional indirect effects of peers over whistleblowing score variables for 16th, 50th and 84th percentile of FL. For each interval, 50,000 bootstrap samples were drawn.

To estimate and analyze our moderated mediation model, we used the computational routines provided in Hayes’ PROCESS macro [see Hayes (2017) for a more detailed review of the statistical treatment], which is appropriate for mediation analysis and widely used to examine psychological processes (Rucker, Preacher, Tormala, & Petty, 2011).

4. Results

Descriptive statistics for the variables used in the models are shown in Table 2. As expected, the intention to report tends to be higher, in general terms, for HWS than for LWS. “Peers” and “Ret” have mean and median closer to the highest value of the scale (7) than to the lowest (1), indicating a strong presence of these two characteristics in the sample. The unstandardized FL median indicates that most students value fairness more than loyalty.

Table 3 describes the estimated models. The effects of “Peers” and “FL” over the whistleblowing scores were positive and significant (p < 0.05); the positive signs support H1 and H3. The moderation effect between Ret and FL was significant only for LWS (p = 0.0240), which is coherent with H4.

To test H2, Table 4 presents the bootstrap confidence intervals for the indirect peer mediation effect. Except for the lowest values of FL for the LWS model, the intervals do not contain the value 0, which do not contradict H2.

5. Discussion and theoretical and practical implications

This work aimed to investigate the effects of individual and situational factors on whistleblowing intentions. Specifically, we examined the combined effects of peer ethical behavior and individual differences in preferences for fairness vs loyalty on whistleblowing intentions. Furthermore, we tested the mediating role of fear of retaliation in the relationship between peer behavior and whistleblowing.

Our results provided support for the aforementioned relationships. Consistent with prior studies in organizational contexts, we found that peers’ ethical behavior (Mayer et al., 2013) and the preference of fairness over loyalty (Waytz et al., 2013) have a positive impact on whistleblowing intentions. Moreover, in line with social information processing theory (Salancik & Pfeffer, 1978) and previous studies on silence in corporate settings (Mesmer-Magnus & Viswesvaran, 2005; Morrison, 2014; Park et al., 2020), we found that peers’ behavior on whistleblowing intentions is mediated by fear of retaliation. Finally, we found support for the interplay between preference for fairness over loyalty and fear of retaliation on willingness to blow the whistle.

Yet, it is worth noting some important differences in the moderation effects that were found. Although it was not the primary objective of this work, our results provide initial evidence that the joint effects of the fairness–loyalty tradeoff and the fear of reprisal depend on the severity of transgression. For low-severity transgressions, the positive effect of preference for fairness over loyalty on whistleblowing intentions is weakened by the perceived risk of being retaliated by peers. However, the interactive effect is not observed for high-severity transgressions neither for the overall whistleblowing score. These results can be explained by a cost–benefit analysis, in which the whistleblower weighs the economic and psychological outcomes of speaking up (Gundlach, Douglas, & Martinko, 2003). When the misconduct is seen as minor, a potential whistleblower may understand that the expected costs (social and work-related retaliations) outweigh the possible benefits of blowing the whistle. In such situations, higher fear of retaliation would undermine the effects of individual’s endorsement of fairness over loyalty in reporting intentions. This is consistent with past research that relates characteristics of the wrongdoing, specifically severity, with the choice to blow the whistle (Culiberg & Mihelič, 2017; Mesmer-Magnus & Viswesvaran, 2005; Near & Miceli, 1985, 1996). The less salient the wrongdoing, the less likely it is for people to speak up against it. Salience can be lower because ethical erosion happens gradually (Gino & Bazerman, 2009) or because of the low perceived seriousness of the wrongdoing, for instance. Likewise, the moral intensity literature (Jones, 1991) indicates that there is a positive relationship between magnitude of consequences of the wrongdoing and reporting intent (Chen & Lai, 2014). Additionally, individuals are motivated to maintain a positive sense of self (Sumanth, Mayer, & Kay, 2011). As such, those who value fairness more than loyalty are more willing to blow the whistle because that is the “right thing to do.” Therefore, if they decide not to blow the whistle because of fear of retaliation, such action would damage their positive self-image because that would be inconsistent with their values. To avoid such negative feelings, individuals may engage in self-deception to convince themselves that their decision to remain silent is not morally questionable. That is, they need to rationalize this cognitive dissonance between attitude and behavior (Festinger, 1957; Latan, Chiappetta Jabbour, & Lopes de Sousa Jabbour, 2019). For low-severity transgressions, that internal justification would be easier because the misconduct is less serious. Curiously, our empirical results corroborate many of the propositions posed by Sampaio and Sobral (2013). According to their theoretical model about whistleblowing in Brazil, attitude toward whistleblowing is influenced by the seriousness of the wrongdoing and fear of retaliation, among other factors.

What are the implications of this all? As the social environment significantly affects someone’s whistleblowing intentions, there should be visible efforts to improve and to foster ethical infrastructure (Tenbrunsel, Smith-Crowe, & Umphress, 2003). That applies either to academic and corporate settings. In regards to the academic environment, it must be embedded in a climate that truly supports ethical behavior by its members (Chaudhary, Gupta, & Phoolka, 2019). Higher ranking members (professors, directors, head of departments, etc.) must be straightforward in dealing with ethical wrongdoings, being these of low or high severity (Miceli, Near, & Dworkin, 2009). The message must be clear that transgressions will not be tolerated, no matter its gravity, and that reports on any type of dishonest behavior will receive full support. As such, universities must relentlessly express – in words, actions and decisions from higher ranking members and professors, employees, etc. – that ethical behavior is expected and demanded from all members of the community. It must explicitly offer incentives and recognition to those who act according to the code of ethics. Also, directors and professors must emphasize – as much and as clear as it is possible – that retaliation to whistleblowers will not be accepted.

Our findings also have managerial implications. As we have shown, and very much in line with the previous literature, the fear of retaliation is one of the most important determinants – alone or as a mediator – affecting the decision to voice or silence. Thus, there must be permanent and serious efforts from managers and leaders to mitigate any threat of retaliation to those who come forward and denounce wrongdoings. This is a difficult task because many times, retaliation is informal and unofficial and hard to detect (Bjørkelo, 2013). In other cases, even worse, it is part of official measures (Park et al., 2020). Some distressing results were found in this sense by Dyck, Morse, and Zingales (2010): they analyzed 216 cases of whistleblowing in big corporations and found that in 82% of the cases in which the whistleblower did not conceal his/her identity, the blower was fired, quit under duress or had significantly altered responsibilities. “In addition, many blowers report having move to another industry and often to another town to escape personal harassment” (Dyck et al., 2010, pp. 2240–2245). Certainly, this ruins any attempt to create an ethical organizational climate. Managers should act as ethical leaders (Treviño, Hartman, & Brown, 2000) and send a clear message that employees who decide to speak up will be fully supported. Additionally, leaders can encourage open discussions among their subordinates about the tensions many of them might feel when attempting to balance loyalty and fairness demands (Dungan et al., 2019). The more open and candid the discussion about these moral conflicts, the clearer the message about which behavior is appropriate when facing these ethical dilemmas.

In any context, creating the incentives and norms for members to speak up must be an exercise which combines societal, legal and organizational effort, in which this last one has a crucial role. Yet, as Bjørkelo (2013) poses, many organizations are still unprepared for this task. Universities are not different.

6. Limitations and future research

Some limitations of this study should be addressed. This paper analyzed whistleblowing intentions within undergraduate students. One limitation is the sample of respondents: all members of a private university in Brazil. Even in the context of higher education, these students had little variance in terms of social background, age, preferences and ideologies. We were also unable to measure the extent of some group factors that may affect whistleblowing intentions, such as group cohesion (Alleyne, Haniffa, & Hudaib, 2019). Moreover, the sample was gathered from a single university. Although the current study focused on peer ethical behavior from the immediate social group, we acknowledge that collecting data from different universities with diverse culture and norms would add to our results.

There are also other opportunities for future research. In this work, we tested the joint effects of individual and contextual factors in whistleblowing intentions, but we did not measure actual reporting behavior. Therefore, collecting data from individuals who actually witnessed misconduct and assessing whether they reported those unethical conducts would be an interesting extension of the findings of this study. Additionally, we focused on the mediating role of fear of retaliation to explain the relationship between peer behavior and whistleblowing intentions. Future work could examine other mechanisms responsible for this effect, such as psychological safety at the group level. Recent research has shown that psychological safety, the shared belief by members of a group that is safe to take interpersonal risks within this group (Edmondson, 1999), is an important antecedent of silence at work (Sherf, Parke, & Isaakyan, 2020). As such, future studies could test whether psychological safety could mediate the link between peer ethical behavior and the decision to speak up in academic settings.

Figures

Figure 1.

The moderated mediation model

Table 1.

Rotated factor loadings and descriptive statistics

	Factor loadings
Item	1	2	Communalities	Mean	SD
Bribing a reprography employee to gain access to the exam in advance	0.937	0.059	0.880	3.96	1.96
Leaking the test content to other students	0.872	0.201	0.800	3.31	1.92
Hacking the absentee system to tamper with class attendance list	0.841	0.269	0.780	3.07	1.83
Hiring someone to solve tests and assignments in your place	0.817	0.290	0.752	3.52	1.99
Fully plagiarizing a course assignment found on the internet	0.595	0.631	0.753	2.50	1.51
Cheating during an exam	0.390	0.783	0.765	2.30	1.51
Signing a presence list as a colleague who is not present	−0.080	0.907	0.829	1.76	1.40
Putting his/her name in a group assignment that he/she did not participate in	0.332	0.708	0.611	2.63	1.70
% explanation	45.4	31.7	77.1

Note:

SD: standard deviation

Table 2.

Descriptive statistics for the variables of the models

Variables	Mean	Median	SD	Minimum	Maximum
OWS	2.88	2.75	1.33	1.00	6.94
HWS	3.46	3.25	1.72	1.00	7.00
LWS	2.23	1.83	1.29	1.00	7.00
Peers	5.00	5.33	1.40	1.00	7.00
Ret	4.79	5.00	1.90	1.00	7.00
Unstandardized FL	0.68	0.67	1.03	−3.33	4.67
MD (control)	59.49	57.00	19.10	24	142

Table 3.

Estimated effects

Effect	Parameter estimates for Ret				Model 1 OWS model				Model 2 HWS model				Model 3 Estimates for LWS model
	Coefficient	se	t	p	Coefficient	se	t	p	Coefficient	se	t	p	Coefficient	se	t	p
Constant	6.6675	0.2837	23.50	0.0000	3.0841	0.2502	12.33	0.0000	4.1286	0.3385	12.20	0.0000	1.9829	0.2537	7.82	0.0000
Peers	−0.4846	0.0396	−12.24	0.0000	0.1463	0.0291	5.03	0.0000	0.1468	0.0400	3.67	0.0003	0.1382	0.0286	4.83	0.0000
Ret					−0.1491	0.0221	−6.73	0.0000	−0.2012	0.0298	−6.76	0.0000	−0.0874	0.0219	−3.99	0.0001
FL					0.3604	0.1115	3.23	0.0013	0.3083	0.1513	2.04	0.0419	0.3909	0.1035	3.78	0.0002
Ret × FL					−0.0134	0.0204	−0.66	0.5118	0.0162	0.0275	0.59	0.5548	−0.0439	0.0194	−2.26	0.0240
MD (control)	0.0093	0.0031	3.02	0.0026	−0.0038	0.0022	−1.72	0.0850	−0.0073	0.0028	−2.63	0.0087	−0.0006	0.0024	−0.26	0.7937

Table 4.

Conditional indirect effects of peers over whistleblowing score variables

FL	Indirect effect for OWS				Indirect effect for HWS				Peer indirect effect for LWS
	Effect	BootSE	BootLLCI	BootULCI	Effect	BootSE	BootLLCI	BootULCI	Effect	BootSE	BootLLCI	BootULCI
−0.9799	0.0659	0.0146	0.0386	0.0953	0.1052	0.0204	0.0676	0.1481	0.0215	0.0130	−0.0037	0.0479
−0.0133	0.0721	0.0122	0.0489	0.0967	0.0976	0.0169	0.0656	0.1320	0.0420	0.0113	0.0203	0.0647
0.9534	0.0784	0.0167	0.0457	0.1122	0.0900	0.0223	0.0476	0.1359	0.0626	0.0162	0.0309	0.0953