Moral repair versus punishment: in ﬂ uences on managerial responses

Purpose – The current management literature suggests that when employees engage in wrongdoing, managers typically respond with punishment.The emergingmoral repair literature suggests an alternative to punishment: a reparative response that focuses on repairing harm and restoring damaged relationships. However, little is currently known about restorative managerial responses, including why managers respond to employee wrongdoing in a reparative versus punitive manner. The purpose of this paper is to examine a varietyof cognitive andemotionalin ﬂ uences onthis managerial decision. Design/methodology/approach – This study used a scenario-based survey methodology. The authors gathered data from 894 managers in sales and ﬁ nancial services contexts to test a set of hypotheses regarding individual-level in ﬂ uencesonmanagers ’ punitive versusrestorativeresponses. Findings – This study found that managers ’ restorative justice orientation, retributive justice orientation, social considerations (e.g. when employees are relatively interdependent versus independent), instrumental considerations (e.g. when the offender is highly valuable to the organization) and feelings of anger in ﬂ uenced their reparativeversuspunitive responses. Research limitations/implications – Data are cross-sectional, so causality inferences should be approached with caution. Another potential limitation is common method bias due to single-source and single-wave data. Practical implications – The ﬁ ndings of this study show that managers often opt for a restorative response to workplace transgressions, and this study surfaces a variety of reasons why managers choose a restorativeresponseinsteadof a punitive response.


Introduction
Management theory and research have traditionally focused on punishment as the manager's primary response to employee wrongdoing (Butterfield, Treviño, & Ball, 1996;Carlsmith, Darley, & Robinson, 2002).Punishment has been associated with a variety of desirable outcomes, including lower turnover intentions, deterring future wrongdoing, offender penitence and reintegration, organizational learning, improved employee attitudes, satisfaction, trust and commitmentparticularly when the punishment process and outcomes are perceived as fair (Treviño & Weaver, 2010).However, punishment also has important limitations.Research shows that unjust, arbitrary or inconsistent punishment can trigger negative outcomes such as lower self-efficacy, helplessness, anger, confusion, sadness, resentment, moral outrage, retaliation and sabotage (Treviño & Weaver, 2010).Theorists also argue that in many cases, punishment does little to repair harm for victims and coworkers or to repair damaged relationships (Goodstein & Butterfield, 2010).
The decades-long dominance of the punishment approach to employee wrongdoing has been challenged by research in moral repair (Goodstein & Butterfield, 2010;Goodstein, Aquino, & Skarlicki, 2011;Fehr & Gelfand, 2012;Goodstein, Butterfield, & Neale, 2016), which has begun to examine an alternative mode of responding to wrongdoing: reparative or restorative approaches.Unlike punitive responses, which focus on retribution and other forms of justice, reparative approaches orient managers toward repairing harm and restoring interpersonal relationships (Goodstein & Butterfield, 2010).However, the moral repair literature has generally focused on offenders, victims or third-party stakeholders, with little focus on managers.Therefore, little is known about reparative responses from the manager's perspective, including their role as initiators of reparative responses.
The limited manager-focused research in this area has generally examined managers as offenders or targets of harmful behavior (Folger & Skarlicki, 2005).We are aware of only one empirical study that explicitly examined factors that influence managers' decisions to engage in reparative versus punitive responses.Neale, Butterfield, Goodstein, and Tripp's (2020) qualitative study suggests that needs, managerial role responsibilities, how others respond to the wrongdoing incident and the organizational environment play a role in determining whether managers respond to employee transgressions in a reparative versus a punitive manner.They called for future research to examine additional factors that might differentiate managers' reparative versus punitive responses.Our research answers this call by examining additional individual influences.
Research on managers' decision to respond to employee misconduct in a reparative versus punitive manner is important because responses convey important signals about organizational values, standards and norms regarding "who matters" (e.g.only punishing the offender versus supporting the victim and third-party stakeholders adversely affected by the wrongdoing) and "what matters" (e.g.only retribution or emphasizing the material and emotional needs and interests of a variety of participants and stakeholders beyond the offender) (Goodstein & Butterfield, 2015).Moreover, as argued by Fehr and Gelfand (2012), the failure of managers to repair harm and restore relationships in the aftermath of workplace wrongdoing can lead to undesirable workplace outcomes.

Theory and hypotheses
Reparative versus punitive approaches Much of the existing literature has examined punishment as the primary response to employee misconduct (Treviño & Weaver, 2010).From a managerial perspective, punishment is the manager's "application of a negative consequence or the removal of a positive consequence following a subordinate's undesirable behavior" (Butterfield et al., 1996(Butterfield et al., , p. 1479)).Punishment can involve a range of responses, from relatively mild actions such as a verbal reprimand or a warning letter to more harsh actions such as suspensions and termination.
Such approaches emphasize three core objectives: restoring victims, reintegrating offenders back into the community and facilitating community healing (Braithwaite, 2002).Compared to solely punitive approaches, reparative approaches are more inclusive, interactive, transparent, forward-looking, empowering, flexible and emphasize mutual responsibility and accountability for repairing harm and relationships (Goodstein & Butterfield, 2015).
We propose two types of individual-level influences on the manager's decision to respond to employee wrongdoing with a more reparative versus a solely punitive approach: cognitive and emotional.These are grounded in research that suggests that managers' use of punishment and/or reparative practices is likely to stem from psychological and affective processes (Butterfield et al., 1996;Gromet & Darley, 2009).We offer and test hypotheses regarding cognitive (managers' justice orientation, moral, social and instrumental considerations and the perceived seriousness of wrongdoing) and emotional factors (anger, moral outrage and sympathy in the aftermath of transgressions).

Managers' justice orientation
Previous theory and research suggest that individuals possess a particular justice orientation (Fehr & Gelfand, 2012;Okimoto, Wenzel, & Feather, 2012;Wenzel, Okimoto, Feather, & Platow, 2008) or "justice mindset" (Gromet & Darley, 2011: 830) that is characterized as reparative or retributive in nature.Punitive orientations are associated with an intuitive impulse to punish offenders (Carlsmith & Darley, 2008).Reparative orientations likely emerge when harm to targets and/or other third parties is salient (Gromet & Darley, 2011) or when affected parties seek to repair harm and renew consensus toward shared values and norms that were violated (Okimoto et al., 2012).Research suggests that punitive versus reparative orientations predispose individuals to an initial preference toward choosing a reparative or punitive response (Okimoto et al., 2012).Managers with a stronger retributive orientation will be motivated to give greater emphasis to the use of punitive Moral repair versus punishment practices, while managers with a stronger reparative orientation will be motivated to give greater emphasis to the use of reparative practices: H1.Managers' retributive orientation will be positively related to a punishment response and negatively related to a reparative response.
H2. Managers' reparative orientation will be positively related to a reparative response and negatively related to a punishment response.

Moral, social and instrumental considerations
In addition to relatively stable justice orientations, other cognitive influences are more specific to the given situation.These include the salience of moral, social and instrumental considerations (Goodstein et al., 2011) and the perceived seriousness of the transgression (Robinson & Bennett, 1995).Goodstein et al. (2011) developed a conceptual framework that emphasizes moral, social and instrumental considerations in the context of reintegrating offenders in organizational contexts.Drawing on this framework, we suggest that managers are motivated to respond to workplace transgressions with either a reparative or punitive response for moral, social and instrumental reasons.
Moral considerations.Moral considerations, such as one's moral identity (Aquino & Reed, 2002), influence what managers view as "the right thing to do" in a particular situation.Managers may view punishment as a morally appropriate response in certain situations, such as when managers perceive that offenders have created an unfair advantage through their transgressions and seek to restore victim status, or when managers anticipate desired outcomes such as respect for authority or maintaining social order (Treviño & Weaver, 2010).However, theorists have also argued that a reparative response may be morally appropriate, particularly when the repair of harm and interpersonal relationships are salient (Goodstein & Butterfield, 2010;Walker, 2006).
Empirical research has begun to examine workplace contexts within which people view a reparative approach as more appropriate, primarily regarding the making of amends, forgiveness and reintegration (e.g.Bradfield & Aquino, 1999;Goodstein, Butterfield, & Neale, 2015;Gromet & Okimoto, 2014;Tripp, Bies & Aquino, 2007).We argue that when moral considerations such as harm experienced by victims and expressed contrition or remorse by offenders are more salient, managers are more likely to respond to wrongdoing in a reparative manner where there is a stronger emphasis on offender reintegration.When such moral considerations are less salient, managers are more likely to choose a punitive response (Gromet & Darley, 2009): H3. Managers are more likely to respond to wrongdoing with a reparative (punitive) approach when moral considerations are more (less) salient.

Social considerations.
A variety of social considerations may also influence the manager's decision to respond with a reparative versus a punitive response.Prior research suggests that managers are influenced by various characteristics of the social environment as they decide whether and how to punish employees, including the expectations of the subordinate who committed the infraction, the expectations of observers and promoting vicarious learning (Butterfield et al., 1996).Social factors potentially influence managers' decisions to engage in reparative practices, including perceptions of shared fate between manager and offender, shared social identity and shared values (Goodstein et al., 2011).Wenzel et al. (2008) suggest that wrongdoing can threaten managers' sense of shared identity and values, which can motivate a reparative response.Consistent with this prior research, we argue that managers will be motivated to respond with a reparative approach when they perceive that it is important to repair relationships and maintain social bonds.This is likely when social considerations such as high task or goal interdependence, shared social identity, shared values and observers' expectations are salient.When social considerations are less salient, managers are more likely to respond with a punitive response (Gromet & Darley, 2009): H4. Managers are more likely to respond to wrongdoing with a reparative (punitive) approach when social considerations are more (less) salient.
Instrumental considerations.Instrumental refers to factors that "serve as a means to some end" (Butterfield et al., 1996(Butterfield et al., , 1486)), and are grounded in the self-interest of the manager or organization.Managers' instrumental goals can include avoiding legal complications, maintaining image or reputation, earning respect, achieving goals, increasing job satisfaction and reducing turnover (Butterfield et al., 1996).Instrumental considerations are also oriented toward deterring and modifying employees' inappropriate attitudes and behaviors (Carlsmith et al., 2002).Consistent with prior literature, we argue that managers will be motivated to respond to wrongdoing in a reparative manner when they perceive that an offender adds value to the work unit and organization.When the offender is perceived as less valuable, the response is more likely to be punitive in nature: H5. Managers are more likely to respond to wrongdoing with a reparative (punitive) approach when instrumental considerations that are associated with a higher value of the offender to the organization are more (less) salient.
Perceived seriousness of the transgression.Perceptions of employee wrongdoing vary in terms of the severity or magnitude of transgressions (Robinson & Bennett, 1995).Some offenses, such as intentionally working slowly, inappropriately blaming or withholding information from others and starting or continuing a damaging rumor, may be perceived as more minor in nature.Other offenses such as embezzlement, sexual harassment and verbal abuse of, stealing from or physical intimidation of other employees are more likely to be perceived and treated as more serious or severe transgressions (Robinson & Bennett, 1995).Previous research has shown that, when confronted with wrongdoing, retribution tends to be people's default mindset; they want and expect offenders to be punished, particularly when offenses increase in perceived severity (Gromet & Darley, 2009).For offenses perceived as less serious in nature, individuals are more open to a reparative response (Gromet & Darley, 2009;Okimoto, Konradt, Krys, & Dawson, 2022): H6. Managers' perceptions of the seriousness of employee transgressions will be positively related to a punishment response and negatively related to a reparative response.
Emotional influences.Emotional reactions to misconduct, such as feelings of anger, moral outrage and sympathy directed toward the offender (Gromet, 2012), likely vary across situations.As discussed by Gromet (2012), evidence shows that when observers have their attention drawn to the harm caused, they experience two primary emotional reactions, anger or moral outrage, toward the offender for the harm they caused and elevated feelings of sympathy for the target.Previous research provides insight into the role of emotion in reparative versus punitive decisions, suggesting that punishment tends to be associated with feelings of anger and moral outrage, and sympathy is more associated with reparative responses (Gault & Sabini, 2000): H7. Managers' feelings of anger will be positively related to a punishment response and negatively related to a reparative response.
H8. Managers' feelings of moral outrage will be positively related to a punishment response and negatively related to a reparative response.
H9. Managers' feelings of sympathy will be positively related to a reparative response and negatively related to a punishment response.

Sample
We initially recruited 1,328 sales managers and 549 financial services managers via Amazon Mechanical Turk (M-Turk).We required respondents to have a 95% approval rating.
Following best practices for quality control (Lu, Neale, Line, & Bonn, 2022), we collected data at various times of day over numerous days and weeks, including pre-screening questions, randomized questions and carefully designed attention checks.We screened out those who failed attention checks and those who did not confirm that they are sales or financial service managers.The final sample consisted of 654 sales managers and 240 finance service managers.Respondents' mean age was 34.64, 46% were female and the average education level was a baccalaureate degree.They averaged 6.92 years of management experience and represented a wide range of industries.All respondents live and work in the USA.

Survey procedure
To increase the generalizability of our findings, we constructed 12 scenarios in two contexts: a sales context and a financial services context.The scenarios manipulated moral, social and instrumental considerations as high and low.These manipulations were also intended to prompt variances in the perceived seriousness of the transgression and emotional reactions to the situation.For example, a scenario that is high in moral (e.g.ethics-related issues are at stake), social (e.g. the company's culture might be affected) and instrumental considerations (e.g. the company's success and reputation might be affected) is more likely to elicit perceptions of seriousness and emotional reactions (e.g.anger) than a scenario that is low in such considerations.Similar to past scenario studies (Weber, 1992), our goal was to create scenarios that were realistic and engaging for managers without priming either a reparative or punitive response.Both scenarios used gender-neutral names and were designed to be as similar as possible, differing only in the focal transgression and the language used to manipulate the moral, social and instrumental variables.Respondents were randomly placed into one of 12 versions that matched their actual employment (sales versus financial services).
The sales scenario involved a salesperson stealing a list of prospects from a co-worker.The financial services scenario involved an employee verbally abusing a coworker.Chisquare results showed no significant difference between the sales and financial services scenarios regarding managers' reparative versus punitive responses [x 2 (1) ¼ 2.06, p ¼ 0.15], so the data were combined.We used a between subjects' design where respondents read their scenario (sales or financial services) and were then presented with descriptions of a reparative approach and a punitive approach, which were presented in random order.Respondents then responded to a series of questions that measured our independent, dependent and control variables.
Measures Managers' response.The dependent variable is the manager's decision to respond to the situation presented in the scenario in a reparative (coded 0) versus a punishment (coded 1) manner.The two options were presented in random order.
Managers' justice orientation.We used previously validated retributive and reparative orientation scales (Okimoto et al., 2012).The retributive orientation scale includes six items (e.g."As a matter of fairness, an offender should be penalized"), and the reparative orientation scale also includes six items (e.g."for justice to be reinstated, the affected parties need to achieve agreement about the values violated by an incident").Cronbach's alpha was 0.82 for retributive orientation and 0.72 for reparative orientation.
Moral, social and instrumental considerations.Moral, social and instrumental considerations were manipulated in the scenarios as dichotomous variables (coded low ¼ 0, high ¼ 1).
Perceived seriousness of the transgression.After reading the scenario, respondents were asked two questions about "how serious" and "how severe" the transgression was, with responses ranging from 1 ¼ not at all serious to 5 ¼ extremely serious.Cronbach's alpha for this measure was.83.
Emotions.To measure anger, moral outrage and sympathy, we asked respondents to "rate the degree to which these words describe your feelings about the incident," from (1) clearly does not describe my feelings to (5) clearly describes my feelings.
Control variables.We included three control variables that are commonly used in similar research: age, gender and years of management experience.Age and years of management experience were continuous variables.Gender was coded as (0) female and (1) male.

Results
Means, standard deviations and correlations among all study variables are presented in Table 1.
We conducted manipulation checks to test the effectiveness of the scenario manipulations.All t-test results were significant at the p < 0.01 level, suggesting that moral, social and instrumental considerations were effectively manipulated.As the survey involved a betweensubjects design across three manipulations (moral, social and instrumental considerations), we separately tested H3-H5 using Chi-square analysis.Results show significant differences for social considerations [x 2 (2) ¼ 7.80, p < 0.05] and instrumental considerations [x 2 (2) ¼ 7.09, p < 0.05], but results were not significant for moral considerations [x 2 (2) ¼ 1.34, p ¼ 0.51].Thus, H4 and H5 were supported, but H3 was not.
We then tested the remaining hypotheses using hierarchical logistic regression (see Table 2).Results indicated that the overall model was statistically significant (x 2 (9) ¼ 200.97, p < 0.001).Control variables (age, gender and years of management experience) were entered in the first step, which was not significant [x 2 (3) ¼ 3.82, p ¼ 0.28].Independent variables were entered in the second step and were significant [x 2 (6) ¼ 197.15, p < 0.001], explaining an additional 27.7% of the variance.Further analysis revealed significant results in the expected direction for retributive orientation (b  needs-based perspective.This research underscores the importance of influences that are relatively stable across contexts (managers' justice orientation) as well as cognitive and emotional influences that are more variable and situation-specific in nature.
We found strong evidence that the manager's justice orientation is an important influence on managerial responses to subordinate transgressions; managers with more retributive orientations tended to adopt punishment responses, while managers with a more reparative orientation were more likely to adopt a reparative approach.While consistent with existing research, our findings are unique in that they offer initial evidence of retributive and reparative orientations among practicing managers.This is also the first study, to the best of the authors' knowledge, to explore the influence of managers' justice orientations within the context of workplace scenarios, specifically how managers' retributive and reparative orientations predict responses to workplace transgressions.
Another contribution of this research is its adaptation of the Goodstein et al. (2011) conceptual framework, which emphasizes the importance of situational cognitive motives (moral, social and instrumental) as influences on co-worker responses.Results showed that two of the three considerationssocial and instrumentalwere significant predictors of the managers' responses.The finding regarding social considerations supports prior literature (Neale et al., 2020) and has important practical implications because it suggests that managers are more likely to be punitive when employees are relatively independent and more likely to be reparative when employees are relatively interdependent.Regarding instrumental considerations, it is possible that results are partly a function of the emphasis of short-term results related to sales and finance contexts.Managers may believe that a reparative response helps them fulfill their duties (Neale et al., 2020).The lack of significant findings regarding moral considerations may suggest that managers are as likely to find punishment morally acceptable as reparative responses or that managers' decisions to respond to wrongdoing in a reparative versus punitive manner are independent of moral considerations.However, it is also possible that a different or stronger manipulation of moral language or studying other roles or industries might show a relationship between these variables.We recommend that future research continue to explore this interesting topic.

Moral repair versus punishment
We also found that managerial emotions can influence the likelihood of engaging in punitive or reparative responses.Results indicated that anger had a significant effect on managers' responses, suggesting that managers who were more angered by the incident were more likely to punish.How anger is expressed and interpreted by others is likely to be a function of organizational norms and culture (Geddes & Callister, 2007).Therefore, strong forgiveness norms at the unit and organizational levels (Fehr & Gelfand, 2012) might influence how managerial anger in the aftermath of a transgression is channeled into more reparative rather than punitive responses.

Study limitations
A limitation of this study is that the data are cross-sectional, so causality inferences should be approached with caution.Another potential limitation is common method bias because of single-source and single-wave data.To lessen the potential for common method bias, we used procedural safeguards recommended by Podsakoff, MacKenzie, Lee, and Podsakoff, (2003), such as randomizing questions and ensuring anonymity.
Implications for future research and practice Overall, these findings are promising with respect to better understanding the role that relatively stable and situation-specific cognitive and emotional factors play in influencing responses to workplace transgressions.Future studies might consider other cognitive and emotional influences.One potentially important set of cognitions are those associated with perceptions of fairness.Neale et al. (2020) showed that managers' responses to transgressions can be influenced by a desire for procedural and interactional justice.The punishment literature (Treviño & Weaver, 2010) also demonstrates how perceptions of distributive, procedural and interactional fairness influence employee outcomes.Future research may explore the extent to which managers adopt responses to transgressions that they believe will be perceived as fair.Although we emphasized managers' emotional responses to misconduct, the emotions of others, such as an offender's remorse (Radzik, 2009) or a victim's or co-workers' desire for revenge or forgiveness (Aquino et al., 2001), may also play an important role.Though managers' justice orientations are relatively stable, examining how situational factors (e.g.transgression severity) influence or moderate justice orientations may be beneficial.We encourage researchers to move beyond individual differences by exploring contextual influences on the manager's decision to respond in a more punitive or reparative manner.
Important practical implications also emerge from this research.Our research demonstrates that when practicing managers are given the option of adopting a reparative response to a workplace transgression, many opt for a reparative response instead of a default punitive response.The mean for our dependent variable reflecting the manager's response was 0.35, with 0 representing a reparative approach and 1 representing a punitive approach, suggesting that managers' responses were somewhat more reparative than punitive in nature.However, little is known about important issues such as the prevalence and relative effectiveness of reparative responses versus punitive responses and whether certain workplace situations might call for a combination of these approaches.Additionally, because we constrained respondents to punitive and reparative responses, other options (e.g.ignoring or postponing) were unavailable.While we have extended inquiry into managerial responses to transgressions that consider punishment and restoration, we invite management scholars to pursue this new and important domain of research.

Table 2 .
Notes: N ¼ 894; the values shown within each model are B(log odds); a the dependent variable is managers' restorative versus punishment response; *p < 0.05; **p < 0.01 Source: Tableby authors