A Fault Confessed Is Half Redressed: The Impact of Deviant Workplace Behavior on Proactive Behavior

,


Introduction
Deviant workplace behavior (DWB), conceptualized as "voluntary behavior that violates organizational norms and thereby threatens the well-being of the organization, its members, or both" (Robinson and Bennett 1995, p. 556), is a widespread manifestation of negative workplace behaviors (Bennett and Robinson 2000;Zhong and Robinson 2021), which includes theft, lying, or breaking workplace rules.As one pervasive form of negative workplace behaviors (Dhanani and LaPalme 2019;Zhong and Robinson 2021), many employees engage in such behaviors to some degree, which leads to numerous monetary and societal costs to the organization.Previous studies have extensively focused on the antecedents of DWB (e.g., Bolton et al. 2010;Ferguson et al. 2012;Mackey et al. 2021;Li et al. 2023) and the strategies that organizations can employ to curb it (e.g., Di Stefano et al. 2019;Thornton and Rupp 2016;Shahzad et al. 2023).
However, most research on DWB has adopted a "victim-centric" perspective, with little attention given to the "victimizer-centric" perspective, which means we know little about how the "victimizers" view themselves and behave as a result (Fan et al. 2021).This is a critical omission because employees continuously engage in various behaviors, and their previous behaviors may shape how they behave later (e.g., Lin et al. 2016).Some scholars have adopted a "cognitive dissonance" perspective through which to view employees' negative behaviors (see a review by Zhong and Robinson 2021).Specifically, following moral cleansing theory (Miller and Effron 2010;Mullen and Monin 2016), people may experience a dissonance between their desired moral self-image and their current moral level after engaging in immoral behaviors (e.g., Barkan et al. 2015;Ilies et al. 2013;Liao et al. 2018), and the perceived moral deficit may motivate individuals to subsequently engage in compensatory behaviors to restore their moral self-concept (Zhong et al. 2009).Therefore, we propose that, after engaging in DWB, employees may subsequently act in ways that benefit organizational effectiveness to compensate for their previous morally questionable behaviors.Nevertheless, to the best of our knowledge, there are few empirical studies that have investigated how and when DWB will lead to subsequent compensatory behaviors (e.g., Dadaboyev et al. 2023;Fan et al. 2021).Our research aims to fill this research gap by examining the underlying mechanism of the relationship between DWB and subsequently positive workplace behaviors.
In this study, we focus on a specific form of positive workplace behavior, namely proactive behavior, defined as "anticipatory action that employees take to impact themselves and/or their environments" (Grant and Ashford 2008, p. 5).Examples of proactive behaviors include feedback-seeking (Ashford et al. 2003), personal initiative (Frese and Fay 2001), taking charge (Morrison and Phelps 1999), and voice (LePine and Dyne 1998).Moral cleansing theory posits that individuals are inclined to undertake reparative actions that are designed to compensate victims (Zhong et al. 2009;West and Zhong 2015) and, since organizations are generally the victims of employees' DWB, employees may engage in proactive behaviors as a form of compensation towards the company because such behavior can enhance organizational efficiency, financial performance, innovation, and various other positive organizational outcomes (e.g., Bindl and Parker 2011;Thomas et al. 2010).In sum, we hypothesize that the victimizer exhibiting DWB will perceive a sense of moral deficit which, in turn, prompts them to subsequently engage in proactive behaviors to repair this deficit.
Additionally, moral cleansing theory suggests that the moral cleansing process is contingent upon an individual's moral characteristics (Liao et al. 2018).Specifically, we focus on an individual difference that reflects the level of morality in one's behaviormoral courage-which refers to "the strength of will is needed to face and resolve ethical challenges and to confront barriers that may inhibit the ability to proceed toward right action" (Sekerka et al. 2009, p. 566).We contend that moral courage moderates the relationship between employees' moral deficit and their subsequent proactive behavior; morally courageous employees are more likely to take a proactive approach (e.g., Sekerka 2015;Sekerka et al. 2009) and engage in compensatory behaviors than those with low moral courage.We offer a moderated mediation model to investigate how and when an employee's DWB will predict subsequent proactive behavior.Our overall theoretical model is depicted in Figure 1.
moral level after engaging in immoral behaviors (e.g., Barkan et al. 2015;Ilies et al. 2013;Liao et al. 2018), and the perceived moral deficit may motivate individuals to subsequently engage in compensatory behaviors to restore their moral self-concept (Zhong et al. 2009).Therefore, we propose that, after engaging in DWB, employees may subsequently act in ways that benefit organizational effectiveness to compensate for their previous morally questionable behaviors.Nevertheless, to the best of our knowledge, there are few empirical studies that have investigated how and when DWB will lead to subsequent compensatory behaviors (e.g., Dadaboyev et al. 2023;Fan et al. 2021).Our research aims to fill this research gap by examining the underlying mechanism of the relationship between DWB and subsequently positive workplace behaviors.
In this study, we focus on a specific form of positive workplace behavior, namely proactive behavior, defined as "anticipatory action that employees take to impact themselves and/or their environments" (Grant and Ashford 2008, p. 5).Examples of proactive behaviors include feedback-seeking (Ashford et al. 2003), personal initiative (Frese and Fay 2001), taking charge (Morrison and Phelps 1999), and voice (LePine and Van Dyne 1998).Moral cleansing theory posits that individuals are inclined to undertake reparative actions that are designed to compensate victims (Zhong et al. 2009;West and Zhong 2015) and, since organizations are generally the victims of employees' DWB, employees may engage in proactive behaviors as a form of compensation towards the company because such behavior can enhance organizational efficiency, financial performance, innovation, and various other positive organizational outcomes (e.g., Bindl and Parker 2011;Thomas et al. 2010).In sum, we hypothesize that the victimizer exhibiting DWB will perceive a sense of moral deficit which, in turn, prompts them to subsequently engage in proactive behaviors to repair this deficit.
Additionally, moral cleansing theory suggests that the moral cleansing process is contingent upon an individual's moral characteristics (Liao et al. 2018).Specifically, we focus on an individual difference that reflects the level of morality in one's behaviormoral courage-which refers to "the strength of will is needed to face and resolve ethical challenges and to confront barriers that may inhibit the ability to proceed toward right action" (Sekerka et al. 2009, p. 566).We contend that moral courage moderates the relationship between employees' moral deficit and their subsequent proactive behavior; morally courageous employees are more likely to take a proactive approach (e.g., Sekerka 2015;Sekerka et al. 2009) and engage in compensatory behaviors than those with low moral courage.We offer a moderated mediation model to investigate how and when an employee's DWB will predict subsequent proactive behavior.Our overall theoretical model is depicted in Figure 1.We test our hypotheses across two studies, namely an experimental study conducted in the United Kingdom and a survey-based field study conducted in the United States.Our research contributes to the literature in several ways.First, the literature on DWB primarily focuses on the detrimental consequences, ignoring its potential positive outcomes; our research helps counter this imbalance by introducing a novel "victimizer-centric" perspective, through which we investigate the psychological and behavioral outcomes of DWB.In doing so, we suggest that DWB can subsequently lead to morally laudable acts (i.e., proactive behavior) once they experience a sense of moral deficit.Second, we integrate the broader proactive behavior literature with the moral literature to provide additional insights into elaborating the antecedents of proactive behavior from a moral perspective.Finally, we contribute to the moral cleansing literature by identifying moral courage as a vital boundary condition that influences the relationship between a moral deficit and socially desirable outcomes.

Deviant Workplace Behavior (DWB)
DWB is defined as any voluntary behavior where employees violate the expectations of the social context in the workplace (Kaplan 1975), and it comprises a range of behaviors, such as theft, harassment, and dishonesty, that go against workplace rules and regulations (Robinson and Bennett 1995).Employees may engage in DWB for both individual and organizational reasons, including self-interest, emotional exhaustion, mistreatment or injustice, unethical collogues or leadership, or issues with the organizational climate (Chen and King 2018;Mackey et al. 2021;Miao et al. 2020;Park et al. 2019;Shafer 2008;Scherer et al. 2001;Wilson et al. 2015).
DWB has detrimental effects on both organizations and employees (Robinson and Bennett 1995).For instance, it can exacerbate workplace incivility among members (Meier and Spector 2013), contribute to abusive supervision (Lian et al. 2014), and even lead to insomnia (Yuan et al. 2018).However, recent research has also begun to explore the potential positive effects of DWB; for instance, Fan et al. (2021) found that DWB can positively contribute to the recovery of employees' resources, which, in turn, enhances their work engagement.Similarly, Ilies et al. (2013) demonstrated that undesirable workplace behaviors are positively associated with organizational citizenship behavior (OCB) through the emotional mechanism of guilt.Diverging from these studies, our research seeks to explore the potential positive impacts of DWB on proactive behavior and the underlying mechanisms from a moral cognitive perspective.

Moral Cognitive Responses to DWB
The fundamental argument of moral cleansing theory is based on a moral balance model (Nisan and Horenczyk 1990) that proposes that individuals have a moral self-regard, which refers to a self-assessment of their current moral degree; the level of moral selfregard increases when they engage in morally laudable behaviors and decreases when they engage in morally questionable behaviors.Every individual has a relatively stable moral equilibrium point, which is the ideal level for one's moral self-regard, and people are motivated to maintain their moral self-regard so that it aligns with their equilibrium point (e.g., Ahmad et al. 2021;Sachdeva et al. 2009).When there is a discrepancy between people's moral self-regard and their moral equilibrium point, they are more likely to act either morally or immorally to maintain the balance.Specifically, if individuals engage in morally commendable behaviors, this endows them with a psychological "license" that allows them to act in a morally questionable way without concerning their self-image (e.g., Loi et al. 2020;Yam et al. 2017).Conversely, the principle applies when their moral self-regard is lower than the moral equilibrium point.If individuals experience a decline in their moral self-regard due to engaging in immoral actions, they are inclined to engage in compensatory behaviors to "cleanse" their moral impurity and boost their moral self-regard to keep it consistent with the moral equilibrium point.
Furthermore, moral cleansing theory suggests that previous unethical behaviors shape subsequently favorable behaviors via both emotional and cognitive mechanisms (Zhong et al. 2009).Several studies have demonstrated the emotional ramifications that occur after immoral behaviors.For instance, Ilies et al. (2013) showed that, when individuals realized their past behaviors violate organizational or social norms, they feel guilty.
Unlike these studies, our research predominantly focuses on the cognitive mechanism.As we discussed above, DWB exerts detrimental consequences on the organizations and simultaneously jeopardizes the individual's moral self-regard (e.g., Aquino and Reed 2002).
Therefore, following moral cleansing theory, we contend that, when employees engaged in DWB, they are more likely to perceive a high moral deficit, which refers to a deficit in personal moral credits resulting from one's own unethical behaviors (Liao et al. 2023).The construct of moral credit is analogous to keeping a moral bank account.Individuals tend to evaluate their behaviors regarding ethicality, where good deeds accumulate moral credits that can be used to pay for subsequent immoral behaviors (Lin et al. 2016), and bad deeds lead to a moral debt, necessitating engagement in compensatory behaviors to "repay" this debt (Miller and Effron 2010;Zhong et al. 2009).Therefore, the fluctuation of moral credits in response to one's actions mirrors an individual's cognitive response to subsequent moral or immoral behaviors.Building on this, perpetrating DWB not only violates social norms but also inflicts harm on the organization and on colleagues (e.g., Robinson et al. 2014).Thus, employees may perceive a loss of moral credit (i.e., moral deficit).Supporting our argument, several studies have identified a positive relationship between moral deficit and previous ethical transgressions.For example, Zhong et al. (2010) conducted two experiments and found that individuals felt a moral deficit upon making an unethical choice.In a similar vein, Liao et al. (2018) showed that leaders experienced a loss of moral credit after perpetrating abusive leadership behaviors.
We, therefore, hypothesize that employees are more likely to perceive moral deficit as a cognitive response to their DWB.

Hypothesis 1 (H1).
The act of engaging in DWB is positively associated with an increase in moral deficit.

Compensatory Response to Perceived Moral Deficit
As previously discussed, moral cleansing theory suggests that individuals are inclined to align their moral self-regard with their moral equilibrium and are, therefore, motivated to engage in morally laudable behaviors to compensate for their previous immoral behaviors.A handful of empirical studies found that employees often engaged in multiple subsequent behaviors, such as apologizing and compensating (Ghorbani et al. 2013), OCB (Ilies et al. 2013), service-oriented helping behavior (Liao et al. 2023), and person-oriented leadership behavior (Liao et al. 2018), to repair their self-image and raise their moral self-regard.Although these studies imply that individuals proactively engage in reparative behaviors to atone for their past transgressions, it is surprising that few studies have, so far, empirically tested whether proactive behavior can "cleanse" a moral deficit (except Wang et al. 2022).Consequently, our research contends that proactive behavior is an appropriately compensatory behavior in response to a moral deficit caused by engagement in DWB.
Proactive behavior involves employees' acting in anticipation to bring about change in the workplace to enhance organizational effectiveness.It has three vital attributes; self-initiation, change-oriented, and future-focused (Parker et al. 2010;Parker and Collins 2010).Proactive employees take the initiative to make things happen, rather than waiting for things happen, and it is considered to be a specific form of prosocial behavior (e.g., Belschak and Den Hartog 2017;Mackenzie et al. 2011), which means it overlaps to some extent with OCB.However, a consensus in much of the proactive behavior literature is that proactive behavior is not an extra role behavior like OCB.As proactivity is a "goal-driven process" (Parker et al. 2010), individuals engaging in it broadly comprehend and redefine their roles so that they can encapsulate the desirable future (i.e., goals) that they wish to achieve (Frese and Fay 2001;Grant and Ashford 2008;Griffin et al. 2007;Parker et al. 2006).In other words, both in-role and extra-role performance can, to some extent, be manifested in a proactive way.
We posit that proactive behavior may be a plausible behavioral response to redress a perceived moral deficit for two reasons.First, when employees experience a moral deficit in themselves, the compensatory behaviors they subsequently use to amend their past immoral behaviors are typically self-started.Second, since organizations are often the primary victims of DWB (e.g., Zhong and Robinson 2021), individuals are likely to prioritize compensating the organization and enhancing its efficiency as their foremost goal (Cryder et al. 2012;Wang et al. 2022).Proactive behavior, which is dedicated to changing the status quo for a better future, can yield numerous positive organizational outcomes, including innovations, resilience, and improved organizational effectiveness (e.g., Detert and Burris 2007;Liang et al. 2012;Moon et al. 2008), that can counterbalance the harm inflicted by a previous engagement in DWB.Furthermore, employees frequently receive tangible rewards, such as promotions or increased salaries, from engaging in proactive behaviors (e.g., Seibert et al. 2001;Thompson 2005), which can serve to enhance their moral self-regard.Employees can leverage proactive behavior to repair the perceived moral deficit resulting from their previous DWB.Based on these arguments, we propose the following: Hypothesis 2 (H2).A perceived moral deficit is positively associated with subsequent proactive behavior.
Combining H1 with H2, we contend that a perceived moral deficit is the underlying mechanism that elaborates the positive relationship between DWB and subsequent proactive behavior.

Hypothesis 3 (H3).
The act of engaging in DWB is positively associated with subsequent proactive behavior via a perceived moral deficit.

The Moderating Role of Moral Courage
Thus far, we have emphasized the underlying mechanism of the moral cleansing process from a "within-employee" perspective.However, moral cleansing theory also suggests that the mediation effect of this process may differ among individuals who possess varying moral characteristics (Mullen and Monin 2016).A recent study showed that individuals activated a "double-distancing mechanism" to their immoral behaviors, treating the morally questionable behaviors of others overly harshly, rather than engaging in compensatory behaviors themselves (Barkan et al. 2012).Drawing upon these insights, we identify moral courage as a vital factor in shaping ethical decision-making.
Moral courage refers to the extent to which individuals are committed to moral principles and their willingness to act in response to their own immoral acts (Liao et al. 2018).Prior studies have shown that moral courage can appear both as a stable personal trait (e.g., Peralta et al. 2021;Sekerka et al. 2009) and as a malleable state (e.g., Jonas et al. 2007;Hannah et al. 2011;Hannah et al. 2013).For example, moral courage can be enhanced by training (Jonas et al. 2007) or a positive leadership style (e.g., Alshehri and Elsaied 2022;Cheng et al. 2019), and morally courageous individuals are more likely to intentionally conduct in morally good ways, despite any perceived risks or obstacles (e.g., Peralta et al. 2021).
Building on these arguments, we propose that employees with high moral courage are more likely to engage in proactive behavior when they perceive a moral deficit caused by their previous DWB than those with low moral courage; indeed, proactive behaviors that contrast with prior DWB are an implicit and public acknowledgment of an employee's previous misconduct.Therefore, it may be more expedient for employees to ignore their moral deficits or state that their engagement in DWB is attributed to several contextual factors (e.g., work stress, unfair treatment) (e.g., Dhanani and LaPalme 2019;Li et al. 2017).Even if employees perceive a moral deficit, they might opt to engage in compensatory behaviors that are directed towards individuals outside of the workplace (Sachdeva et al. 2009) or find a plausible excuse to justify their immoral behaviors, rather than proactively helping the organization and make amends for their mistakes.However, employees with high moral courage are more likely to bravely acknowledge their previous wrongdoings without fearing the risks associated with such an admission.Moreover, such employees tend to confront their immoral deeds and are more likely to conduct effective compensatory behaviors to assuage their moral deficit.In addition, scholars have argued that engaging in proactive behavior is not easy (e.g., Griffin et al. 2007), as employees need to be courageous in challenging the status quo, overcoming difficulties, and obstacles, eventually contributing to a more desirable future.
Based on these arguments, we hypothesize that: Hypothesis 4 (H4).Employees' moral courage moderates the relationship between perceived moral deficit and proactive behavior, such that, the positive relationship is stronger when their moral courage is high rather than low.
Integrating our theoretical arguments about the moderating role of moral courage and the mediating role of moral deficit, we contend the following first stage moderated mediation hypothesis: Hypothesis 5 (H5).Employees' moral courage moderates the indirect effect of their DWB on subsequent proactive behavior via perceived moral deficit, such that the indirect effect is stronger when their moral courage is high rather than low.

Method
We conducted two studies to test our hypotheses.In Study 1, we used experimental settings to test the causal effects of DWB on employees' subsequent perceived moral deficit and proactive behavior, since previous research has found that employees will engage in unethical behavior subsequent to exhibiting pro-social behavior (e.g., Yam et al. 2017;Loi et al. 2020).In Study 2, we employed a multi-wave survey-based study to examine our entire hypothesized model.

Study 1 3.1.1. Design and Participants
We recruited 130 full-time employees working across the United Kingdom using Prolific, an online survey platform widely used for research purposes (Peer et al. 2017).The participants were from various industries, including manufacturing, retail, and finance, and were randomly allocated to either the experiment group or the control group.Participants in the experimental condition were presented with an overview of DWB, supplemented by some examples, such as "taking an additional or longer break" (Bennett and Robinson 2000).Subsequently, they were required to recall and detail any such actions that they had exhibited in the past month.Two sample responses were "I took a longer lunch break so I could have a nap" and "I have a daydream when I should have been writing functional specs".On the contrary, participants in the control condition were asked to recall and describe a routine work activity they had performed.Two sample responses were "creating reports and analyzing data to provide insights to stakeholders," and "using project management software to manage budgets".We meticulously examined all of the responses to ensure they aligned with our instructions.Two participants were excluded, as their responses did not meet our minimum word-count requirement, resulting in a final sample of 128 participants, of whom 51.6% were male; their average was 43.81 (SD = 11.33);86.5% of them had a bachelor's degree or higher; 51.6% did not hold a managerial position, and the average organizational tenure was 10.37 years (SD = 8.67).After completing the text item, they were asked to respond to questions about manipulation check, moral deficit, and subsequent proactive behavior.

Measures
Unless otherwise stated, the specific variables that are described below were rated on a five-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree).
The mediator is moral deficit.We adopted a five-item scale developed by Yuan et al. (2018) to measure moral deficit (α = 0.97).The participants were instructed to indicate their agreement with each item based on how they thought about themselves after engaging in DWB.One representative item was, "When I engaged in this behavior, I lost moral credit for performing a morally questionable behavior".
The dependent variable is proactive behavior.We measured proactive behavior using nine items from Griffin et al. (2007) (α = 0.91).The participants were asked to indicate how frequently they have engaged in proactive behaviors after exhibiting DWB on a five-point scale from 1 (never) to 5 (always).A sample item was, "I initiate better ways of doing my core tasks".

Manipulation Check
Prior to the primary analysis, we conducted an analysis of variance (ANOVA) to check the validity of the effectiveness of our manipulation check using the item "The behavior you have recalled was ethical" (Wang et al. 2022).The results show that participants in the experimental condition rated this item lower (M = 2.75, SD = 0.78) than those in the control condition (M = 4.63, SD = 0.63; F [1, 126] = 225, p < 0.01, η p 2 = 0.64), indicating that DWB was successfully manipulated in our experiment.

Study 1 Results
We conducted a one-way ANOVA to examine whether the participants in the experimental condition perceived more of a moral deficit than those in the control condition.H1 posits that the act of engaging in DWB is positively associated with an increase in moral deficit.Figure 2 reveals that those who recalled a DWB reported a higher degree of moral deficit (M = 2.47, SD = 0.86) than those who recalled a routine workplace behavior (M = 1.20,SD = 0.46; F [1, 126] = 107.43,p < 0.01, η p 2 = 0.46), suggesting that employees' DWB has an impact on perceived moral deficit, thereby supporting H1.Furthermore, we tested whether the participants in the experimental condition engaged in more proactive behavior than those in the control condition.The results show that those who recalled a DWB reported more proactive behaviors (M = 3.52, SD = 0.61) than those who recalled a routine workplace behavior (M = 3.22, SD = 0.78; F [1, 126] = 5.83, p < 0.05, η p 2 = 0.04), providing preliminary support for H3.Overall, we leveraged Study 1 to demonstrate the causal effect of DWB on moral deficit and subsequent proactive behavior.

Design and Participants
We conducted a survey-based study to test all our hypotheses.We recruited full-time employees working in the United States through Prolific and employed a multi-wave survey design to collect data across two different time points, separated by a one-week interval, to mitigate common method bias (Podsakoff et al. 2003).At Time 1, 203 participants responded to questions regarding demographic information, DWB, and moral deficit.After removing the respondents who failed the attention check, we retained 197 effective responses.One week later, at Time 2, 180 participants (response rate 91.4%) completed a follow-up survey including questions on proactive behavior, moral courage, and the con-

Study 2 3.2.1. Design and Participants
We conducted a survey-based study to test all our hypotheses.We recruited full-time employees working in the United States through Prolific and employed a multi-wave survey design to collect data across two different time points, separated by a one-week interval, to mitigate common method bias (Podsakoff et al. 2003).At Time 1, 203 participants responded to questions regarding demographic information, DWB, and moral deficit.After removing the respondents who failed the attention check, we retained 197 effective responses.One week later, at Time 2, 180 participants (response rate 91.4%) completed a follow-up survey including questions on proactive behavior, moral courage, and the control variables.Each participant was compensated with GPB 0.50 (approximately USD 0.65) at Time 1 and GBP 0.63 (approximately USD 0.80) at Time 2. The final sample of 180 valid participants worked in various industries, including finance, insurance, retail, manufacturing, and IT.Among them, 51.7% were female; 79.4% were identified as Caucasian; the average age was 38.29 (SD = 10.80);78.4% held a degree of bachelor's degree or higher; and 59.4% held a non-managerial position.Their average tenure at their current organization was 6.91 years (SD = 6.96).

Measures
Unless otherwise noted, the specific variables described below were rated on a fivepoint Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree).
The independent variable is DWB.We measured DWB using the six-item self-reported measure developed by Bennett and Robinson (2000) on a five-point scale from 1 (never) to 5 (always) (α = 0.83).A sample item was "I take an additional or longer break than is acceptable at my workplace".
The moderator is moral courage.Moral courage was assessed using the three-item scale from Hannah and Avolio (2010) (α = 0.73).An example item was, "I demonstrate courage to do the right thing, even at personal cost".
For the mediator and dependent variable, we used the same items from Study 1 to measure moral deficit (α = 0.97) and the employee's proactive behavior (α = 0.90).
For the control variables, we included a range of control variables to enhance the validity of the relationships among the core constructs (Bernerth and Aguinis 2016).First, referring to previous studies on DWB (e.g., Yuan et al. 2018), we controlled for demographic variables, specifically gender, age, education, organizational tenure, and position, as these variables have impacts on DWBs and proactive behaviors.Further, we controlled for three personality traits that may influence proactive behavior, namely agreeableness, consciousness, and proactive personality (e.g., Fuller and Marler 2009;Parker et al. 2010).Agreeableness (α = 0.69) and consciousness (α = 0.81) were measured using the scale developed by Soto and John (2017), and proactive personality was measured using Seibert et al.'s (1999) ten-item scale (α = 0.87).

Study 2 Results
For the confirmatory factor analysis, prior to testing the hypotheses, we performed a series of confirmatory factor analyses (CFAs) using RStudio and the Lavaan package (Rosseel 2012) to examine the construct validity of the measures (i.e., DWB, moral deficit, moral courage, and proactive behavior).Table 1 shows that the hypothesized four-factor model had a better fit (χ 2 = 473.04,df = 225, χ 2 /df = 2.10; comparative fit index [CFI] = 0.91, incremental fit index [IFI] = 0.91, standardized root mean square residual [SRMR] = 0.06, root-mean-square error of approximation [RMSEA] = 0.08) to the data than any other alternative models, such as a three-factor model that combined moral courage and moral deficit (χ 2 = 594.65,df = 227, χ 2 /df = 2.62; CFI = 0.86, IFI = 0.86, SRMR = 0.09, RMSEA = 0.10) and a two-factor model that combined DWB, moral courage, and moral deficit (χ 2 = 918.49,df = 229, χ 2 /df = 4.01; CFI = 0.74, IFI = 0.74, SRMR = 0.14, RMSEA = 0.13).The results demonstrate the discriminant validity of our focal constructs.For hypotheses testing, Table 2 presents the descriptive statistics and correlations for the core variables and control variables.To test our hypotheses, we employed a path analysis model in RStudio, which allowed us to simultaneously estimate the coefficients for all the hypothesized relationships.Table 3 displays the results of the path analysis model.Model 2 in Table 3 indicates that employees' DWB is positively related to moral deficit (b = 0.34, p < 0.05); Model 4 in Table 3 shows that moral deficit is positively related to subsequent proactive behavior (b = 0.14, p < 0.01).These results imply that DWB has a significantly positive effect on moral deficit, and moral deficit also has a positive impact on employees' proactive behavior, supporting H1 and H2.Moreover, Model 6 in Table 3 shows that the interaction of moral deficit and moral courage is positively related to proactive behavior (b = 0.13, p < 0.05), indicating that moral courage positively moderates the relationship between moral deficit and employees' proactive behavior, which supports H4.Further simple slope tests also reveal that, when the employee's moral courage is high (1 SD above the mean), the relationship between moral deficit and employees' proactive behavior is statistically significant (b = 0.21, p < 0.01).However, this relationship becomes non-significant when the employee's moral courage is low (1 SD below the mean) (b = 0.06, n.s.). Figure 3 depicts the pattern of the interaction effect.Then, we used the PROCESS macro (Model 4) to examine the mediation effect of moral deficit between DWB and proactive behavior.We conducted the bootstrap approach with 5000 samplings to obtain the 95% bias-corrected confidence intervals (CI).We find that the indirect effect of DWB on proactive behavior via moral deficit is statistically significant (indirect effect = 0.05, bootstrapped SE = 0.02, 95% CI: [0.01, 0.09]), which supports H3.
Finally, we employed the PROCESS macro (Model 14) to test the moderated mediation effect.Similar to the test of the mediation effect, we performed the bootstrap approach with 5000 samplings to obtain the 95% CI.As shown in Table 4, employees' DWB is more positively associated with proactive behavior via moral deficit when the moral courage is  Note: N = 180.a dummy coded: 0 = male, 1 = female; b dummy coded: 0 = non-managerial, 1 = managerial.Cronbach alpha reliabilities are reported in parentheses.* p < 0.05, ** p < 0.01.Then, we used the PROCESS macro (Model 4) to examine the mediation effect of moral deficit between DWB and proactive behavior.We conducted the bootstrap approach with 5000 samplings to obtain the 95% bias-corrected confidence intervals (CI).We find that the indirect effect of DWB on proactive behavior via moral deficit is statistically significant (indirect effect = 0.05, bootstrapped SE = 0.02, 95% CI: [0.01, 0.09]), which supports H3.
Finally, we employed the PROCESS macro (Model 14) to test the moderated mediation effect.Similar to the test of the mediation effect, we performed the bootstrap approach with 5000 samplings to obtain the 95% CI.As shown in Table 4, employees' DWB is more positively associated with proactive behavior via moral deficit when the moral courage is high (Estimate = 0.07, bootstrapped SE= 0.03, 95% CI: [0.01, 0.09]) than when it is low (Estimate = 0.02, bootstrapped SE = 0.02, 95% CI: [−0.01, 0.06]).The index of the moderated mediation effect is also significant (Estimate = 0.04, bootstrapped SE = 0.03, 95% CI: [0.00, 0.10]), all of which support H5. Notes: SE: standard error; CI: confidence interval.Lower limit and upper limit are reflected in parentheses.a : 95% CI does not include zero.

Discussion
Our research leverages moral cleansing theory to explore the potentially positive outcome of DWB from a "victimizer-centric" perspective.Specifically, we contended that DWB is positively associated with subsequent proactive behavior via a perceived moral deficit, and such an effect is especially heightened when the employee's moral courage is high rather than low.Our arguments received empirical support from an experimental study conducted in the United Kingdom and a survey-based field study performed in the United States.We will discuss the theoretical and practical implications in the following sections.

Theoretical Implications
First, our research broadens the DWB literature by focusing on the employees' psychological and behavioral consequences from a "victimizer-centric" perspective.Specifically, following the moral cleansing theory, we highlight the "bright" side of DWB by examining how it can influence employees' proactive behavior via moral deficit.To date, the overwhelming majority of the DWB literature has focused on the "dark" side of it, suggesting that DWB is negatively related to organizational outcomes, such as workplace incivility (e.g., Meier and Spector 2013) and insomnia (Yuan et al. 2018).A key limitation of this approach is its exclusive focus on the victim.Employees who conducted DWB may recognize that their behaviors violate organizational and social norms, prompting them to display compensatory behaviors to cleanse their perceived immorality.Indeed, some scholars have called for future research to concentrate more on the potentially positive influence of DWB on the victimizer (e.g., Krischer et al. 2010), while several studies have argued that individuals who engage in unethical behavior are more likely to help others (Ding et al. 2016), resist subsequent temptation to cheat on financial issues (Gaspar et al. 2015), and exhibit more prosocial leadership behaviors (Liao et al. 2018).Diverging from these studies, our study uncovers a unique outcome of DWB, namely proactive behavior, which is beneficial and constructive for enhancing organizational effectiveness.By providing seemingly distinctive and paradoxical evidence about how and when DWB facilitates their subsequent proactive behavior, our findings advance the understanding of the consequences of DWB and suggest that certain bad deeds performed by employees may unexpectedly lead to good behaviors afterward.Therefore, our research offers a vital contribution to the DWB literature.
Second, our research makes important contributions to the broader literature on proactive behavior by introducing DWB as a novel antecedent of proactive behavior.Previous research in the proactive behavior literature has identified three motivational states, which are "can do" (i.e., an individual's belief in his or her ability to engage in proactive behavior); "reason to do" (i.e., the perceived necessity to undertake proactive behavior); and "energized to do" (i.e., perceived positive states from which to engage in proactive behavior) (Parker et al. 2010).Our research extends the "reason to do" motivational state by adopting a moral cleansing approach.Specifically, our results showed that proactive behavior could be a reparative response to address the moral deficit resulting from perpetrating DWB.Other studies on proactive behavior primarily found that negative behaviors are less likely to predict various proactive behaviors, as employees who encounter negative behaviors lack sufficient resources to act proactively (e.g., Xu et al. 2019;Wu et al. 2018).Our findings supplement these studies by suggesting that the two paradoxical behaviors-DWB and proactive behavior-may coexist within the same individual, which indicates that the relationship between negative workplace behaviors and proactive behavior may be more complex than was previously understood.
Third, we extend the moral cleansing theory by empirically demonstrating employees' fluctuations in morality at the within-person level.While past research in the moral cleansing literature has indicated that certain bad deeds may result in subsequently positive behavior through a moral cleansing mechanism (e.g., Gino and Margolis 2011;Zhong et al. 2009), little research empirically tested the underlying mechanisms.A recent study by Liao et al. (2018) employed a moral reparative model to reveal the positive relationship between abusive leadership behaviors and constructive leadership behaviors.Nevertheless, the empirical study of the moral cleansing process is still nascent.Our research offers some robust evidence of how, why, and when employees' moral transgressions can lead to subsequent compensatory actions.By illustrating the changes in employees' moral self-regard triggered by DWB and proactive behavior, we highlight that, in organizations, the moral and immoral deeds of both employees and supervisors vary across time.More importantly, we add to the moral cleansing theory by offering insights regarding the boundary condition of the moral cleansing process.Our research particularly diverges from previous studies that mainly focused on identifying the factors that either enhance or mitigate the dissonance between an individual's desired moral self-regard and their actual moral degree after behaving unethically (e.g., Van Tongeren et al. 2015;Ward and King 2018).We extend the literature by revealing the moderating effect of moral courage on the relationship between perceived moral deficit and proactive behavior.Specifically, only morally courageous employees are able to confront their moral deficit (e.g., Liao et al. 2018;Sekerka et al. 2009), thereby effectively translating the perceived moral deficit to subsequent positive outcomes.In sum, our findings complement moral cleansing theory by using a between-person approach to identify the vital role of moral characteristics in the moral cleansing process, leading to a more holistic understanding of the transition from negative behaviors to positive behaviors.

Practical Implications
Our research has several practical implications for employees and organizations.First, our results demonstrated that engaging in DWB yields a perception of moral deficit, which can, in turn, lead to a subsequent proactive behavior to rectify past transgressions.However, proactive behavior is an inherently time-and resource-consuming behavior (e.g., Sonnentag and Starzyk 2015;Tsai 2023).Engaging in proactive behavior to compensate for previous DWB may deplete employees' energy, leading to increased fatigue or even emotional exhaustion (Pan et al. 2023).Therefore, the primary objective of organizations should be to minimize employees' engagement in DWB.For example, some employees engage in DWB to cope with work-related stressors (e.g., Junaedi and Wulani 2021; Wu and Wei 2024), so organizations can conduct regular surveys to assess employees' stress levels.Additionally, providing clearer and more transparent performance-evaluation mechanisms can help alleviate employees' "appraisal-related" pressures.Moreover, organizations should reinforce organizational moral and ethical norms, ensuring that the employees are fully aware of the potential destructiveness of DWB from the outset.
Second, our results indicated that employees are neither absolutely "good" nor "bad"; rather, their behaviors may fluctuate over time, from "bad behaviors" to "good behaviors" or vice versa.As a result, organizations should not hastily label an employee as "bad" solely based on immoral behavior exhibited at a specific time, since employees may engage in a variety of behaviors that are theoretically contradictory and opposing, yet transformable within a given period.However, it is neither realistic nor advisable to encourage employees to engage in DWB first with the hope of subsequently boosting their proactivity.For employees who occasionally engage in morally questionable behaviors because of low selfregulation (e.g., Yam et al. 2016), we recommend that they can engage in proactive behaviors to counterbalance their previous unethical behaviors and mitigate the harmful effect of their actions on the organization.Moreover, organizations are advised not to promptly penalize employees who have committed wrongdoing.Instead, exercising patience and observing over time is recommended, as these employees may engage in subsequent compensatory behaviors to rectify their previous negative actions.
Third, our findings regarding the boundary condition of the moral cleansing process show that employees with a high level of moral courage are more likely to exhibit compensatory behaviors when they perceive a deficit in terms of their moral self-regard.Consequently, organizations could use tests in the selection process that measure moral characteristics to screen for employees with high moral courage.Furthermore, organi-zations and managers may also implement training programs to intentionally enhance employees' moral courage (Sekerka et al. 2014).For instance, organizations may choose to provide experiential training courses designed to develop employees' moral courage by requesting them to express their actual opinions about a range of ethical issues or establish certain company policies that prompt apologies and forgiveness for immoral actions.Such policies could foster a favorable climate where employees are brave enough to report recent unethical behaviors that they have either committed or witnessed.

Limitations and Directions for Future Research
Our research has several limitations that should be improved upon in future research.First, although our results show that DWB has a linear effect on perceived moral deficit and subsequent proactive behavior, it is also possible that such a relationship could be a Ushaped curve.In other words, the intensity of immoral perception resulting from immoral behaviors might vary depending on the severity or frequency of such behavior.Specifically, if an employee occasionally engages in DWB that causes minimal harm to the organization, he or she may not perceive it to be an ethical issue, but rather view such behaviors as an approach to recover from work-related stress (e.g., Eschleman et al. 2015).However, if an employee frequently perpetrates DWBs that significantly harm the organization, he or she may recognize such behaviors are excessive, which can lead to a subsequent moral cleansing process.Although we did not observe any U-shaped curvilinear effects across the two studies, we encourage researchers to investigate the potential nonlinear effect of DWB on subsequent compensatory behaviors to extend our understanding of the moral cleansing mechanism between "bad" deeds and "good" deeds.
Second, although we used a combination of multiple research methods (i.e., an experimental study and a survey-based field study) to provide robust evidence to support our hypotheses, our study also has several methodological limitations.First, while Study 1 demonstrates the causal effect of DWB on perceived moral deficit and proactive behavior, we did not examine the causal effect of moral deficit on proactive behavior.Although it is unlikely that employees feel a deficit in their morality after conducting proactive behavior, it is still necessary to test this causal effect using another experiment to ensure the validity of our conclusions.Second, we used self-reported data to measure employees' DWB and proactive behavior and, although this approach is prevalent in the counterproductive behavior literature (e.g., Ilies et al. 2013;Loi et al. 2020), individuals are inclined to report more positive behaviors following an admission of engaging in unethical behaviors.Therefore, we recommend that future research should collect data from multiple sources (e.g., supervisors and colleagues) to mitigate potential common method bias.In addition, it is fruitful for further research to focus on collecting longitudinal data to help us gain a deeper understanding of the dynamics of employees' cognitive and behavioral changes.
Third, since our study only tested the boundary condition of the moral cleansing process at the individual level, future research should explore more possible moderators at the team or organizational level.For instance, if employees are in an organization with low organizational justice, engaging in DWB might make them feel more entitled than experiencing a moral deficit (e.g., Klotz and Bolino 2013), leading to fewer compensatory actions.Therefore, it would be meaningful to incorporate several higher-level variables that could either strengthen or weaken the overall moral cleansing process.
Finally, since both of our studies were performed in Western countries (i.e., the United Kingdom and the United States), the generalizability of our findings is limited.Contrary to the individualistic focus of Western cultures, Eastern cultures place greater emphasis on collectivism (Hofstede 2001).Such cultural distinction could lead to employees who engage in DWB in an Eastern context experiencing a heightened sense of moral deficit.Hence, we encourage future research to further examine our theoretical model in various cultural settings, such as China and Japan.

Figure 1 .
Figure 1.Overall hypothesized model.We test our hypotheses across two studies, namely an experimental study conducted in the United Kingdom and a survey-based field study conducted in the United States.Our research contributes to the literature in several ways.First, the literature on DWB primarily focuses on the detrimental consequences, ignoring its potential positive outcomes; our research helps counter this imbalance by introducing a novel "victimizer-

Figure 2 .
Figure 2. Outcomes of experimental conditions on moral deficit and proactive behavior (Study 1).

Figure 2 .
Figure 2. Outcomes of experimental conditions on moral deficit and proactive behavior (Study 1).

Figure 3 .
Figure 3. Interactive effect of moral deficit and moral courage on proactive behavior (Study 2).

Figure 3 .
Figure 3. Interactive effect of moral deficit and moral courage on proactive behavior (Study 2).

Table 1 .
Comparison of measurement models for focal constructs (Study 2).

Table 3 .
The results of the path analysis model (Study 2).

Table 4 .
The results of the moderated mediation effect test (Study 2).