Not My Circus,Not my Monkeys? Frontline Employee Perceptions of Customer Deviant Behaviors and Service Firms’ Guardianship Policies

Data Collection and Analysis

To shape our qualitative study, we utilized ideas from popular press articles on CDB to guide informal discussions with 25 employees (FLEs and FLMs). These initial exchanges informed the semi-structured interview guide employed in Study 1 (see Web Appendix 3) and allowed the authors to identify primary categories of interest for comprehensive exploration in line with our overarching research questions.

Three authors conducted 26 in-depth interviews over the course of 2 months (average 31 min) with FLEs and FLMs in retail and service roles. Periodic refinements were made to the interview guide following an iterative transcript review process (Chandler et al. 2019; Corbin and Strauss 2015). Purposeful sampling sought to vary respondents in terms of age (21–62 years, average = 38), gender (65% female), experience (average 14.7 years of retail services experience and 7.2 years tenure with current firm), roles (38% FLMs, 58% FLEs, 4% dual job function), and employment status (62% employed full-time). Interviewees were drawn from various organizational frontline contexts, including retail (e.g., consumer electronics, fashion, hardware) and services (e.g., restaurant, grocery, food delivery services). Detailed characteristics are included in Web Appendix 4. Interviews were transcribed (345 single-spaced pages) and imported into NVivo (1.6.1) for analysis by a co-author and trained graduate research assistant.

Qualitative Coding Processes

Coders read the full transcripts to provide an initial understanding of the thematic content (Sandberg et al. 2022). Next, following best practices (Corbin and Strauss 2015; Gioia, Corley, and Hamilton 2013), coders used grounded theory processes of open coding, axial coding, and selective coding for conceptual ordering. Thus, an inductive, descriptive qualitative approach was taken with line-by-line, open coding used for the initial assessment. Guiding questions were used to create categories of interest by labeling phenomena and determining properties within classification parameters to extract conceptual meaning. Findings were subjected to axial coding to identify commonalities and second-order themes. Coders used sensitizing questions—for example, considering issues, reviewing in/consistency across and within actor responses—as well as theoretical questions—for example, relationships across first-order constructs and changes across respondents (Corbin and Strauss 2015). Finally, using selective coding, themes were arranged hierarchically, creating aggregated dimensions across categories. A co-author reviewed themes for consistency; divergent categorizations were resolved by consensus. The iterative interview process continued until theoretical saturation was reached—that is, sufficient repetition of nodes to constitute hierarchical aggregation without new insights emerging (Zeithaml et al. 2020).

Credibility Measures

Attention was paid to quality-enhancing conditions (Corbin and Strauss 2015), including (1) methodological consistency and awareness, (2) qualitative research training, (3) clarity of purpose, and (4) sensitivity to participants and data. Credibility was further enhanced through (1) peer debriefing (unrelated, experienced researcher reviewed the data collection process, methodological consistency, and emergent primary codes and descriptions), (2) triangulation of sources (FLMs and FLEs), and (3) prolonged engagement with contextual material (Lincoln and Guba 1985). Transferability was demonstrated by recruiting respondents across a wide variety of retailing services while dependability was demonstrated by repeated themes and descriptions emerging from both initial discussions with employees and the subsequent formal qualitative inquiry. Attention to confirmability or neutrality was warranted and thus the qualitative research was designed to include three researchers conducting interviews with another two researchers completing the data analysis. Primary insights from Study 1, which provide the foundation for subsequent studies, are explored in detail below (additional support provided in Web Appendix).

Qualitative Findings

Types of guardianship policies (RQ2)

Interviewees highlighted that firms vary in the composition and communication of their guardianship policies. Namely, our analysis identified three primary guardianship types—Required, Authorized, and Prohibited with both active and passive approach styles employed (see Table 2 for supporting quotes). An active approach involves confronting CDB directly, while a passive approach involves service provision or assistance without direct confrontation. Guardianship policies were relatively balanced across primary types (i.e., 19.23% required, 23.08% authorized, 30.77% prohibited), with variance in passive, active, or combined approaches being utilized.

OPEN IN VIEWER

Table 2.

Types of Guardianship Policies with Representative Quotes.

		Guardianship Type
		Required GuardianshipEmployee is required to interact with the customer if CDB is witnessed or suspected	Authorized GuardianshipEmployee is permitted to interact with the customer if CDB is witnessed or suspected	Prohibited GuardianshipEmployee is prohibited from interacting with the customer if CDB is witnessed or suspected
Approach Style	ActiveRegards policy as it relates to whether employee can directly address customers (e.g., confrontation or accusation) if CDB is witnessed or suspected.	Required confrontation: Employee is required to confront about CDB• “So, you have the, the head of loss prevention say do this, do that, do this, to help stop shoplifting. If you have to confront a customer or whatever, do it.” (Winston)• “I didn’t want to confront them, but I knew I was the one that saw. So, I ultimately had to be the one that took control and took action to stop this.” (Melissa)• “We try to defuse the situation through conversation, through taking a step back. It’s part of our leadership training, how to handle confrontation, it’s part of our training (Stacy)	Authorized confrontation: Employee may confront about CDB at their discretion• “Yes. I could engage customers in the parking lot if I wanted to…” (Stan) a • “We can approach customers while they are in-store and be specific when addressing them. For example, we can say “did you misplace something in your purse? No harm no foul if it makes its way back onto the shelf.”” (Cassidy)• “If you can clearly identify a shoplifter… and you can catch, or recover is the word, I guess recover your merchandise, you know there’s like $100 payout, for you know getting your stuff back.” (Christa)	Prohibited confrontation: Employee may NOT confront about CDB• “Well, even I as the store manager I wasn’t procedurally allowed to confront.” (Randy)• “But it is against policy. You are not supposed to confront anyone for that. You’re just to call the police.” (Tiffany)• “Yeah, I mean we know we’re not supposed to chase them, you know. That the most we’re supposed to do, the policy is to get the person’s driver’s license plate.” (Larry)• “Don’t worry about loss, if we take, we take a loss, we take a loss. We’re making enough money they’re not worried about it that thing because it’s always guest satisfaction first.” (Scott)
	PassiveRegards policy as it relates to whether employee can passively address customer (e.g., servicing behaviors) if CDB is witnessed or suspected.	Required service: Employee is required to provide service• “…they really like, pushed on the fact you should approach almost every customer and keep following up because it also helps if you do see someone acting suspicious, they can kind of scare them a little. So, then they would either not steal or probably the store.” (Elizabeth)• “So, the policy was…we’re supposed to call security… Then after we called them, we told them, apprise them of the situation, and then they would, every single time, I was told to keep an eye on them. Keep someone with them all the time, provide the customer service.” (Deb)	Authorized service: Employee may provide service at their discretion• “Yeah, you ask them if they need any help, you try to engage in that regard, and then you report it to a manager.” (Noah)• “It’s mainly like, distract, customer service them while security is determining if they call the authorities or not” (Leon)• “It is not like I have a gun or can arrest the person. Just trying to talk to the customer, try to advise them on style – I am really good at it - and just really hope for the best.” (Della)	Prohibited service: Employee may not service or interact if CDB• “We can’t do anything really.” (Peggy)• “We don’t let our employees engage shoplifters. It is a liability, most of our lower-level employees are young kids. They have to call a manager.” (Stan) a • “And then it’s really frustrating that [company name] and a couple other companies that have the policy where they can’t accuse people of stealing, they can’t really approach them.” (Raquel) b

^aDifferent employees or managers can have different roles/responsibilities in this process.

^bThere is overlap for the prohibited category in terms of passive versus direct characteristics, but we provide representative quotes to demonstrate these aspects.

Required Guardianship involves policies whereby employees are required to actively confront and/or passively service customers if CDB is witnessed or suspected. Thus, even when CDB is witnessed, FLEs must interact with customers. Multiple interviewees commented on a sense of pride or ownership associated with protecting their store and customers. For example, Winston suggests, “And so I’d say there’s a little bit of pride that comes in, I think, with the store. You’re stealing from my store, my department.” Employees also disclose feeling frustrated, demotivated, fearful, anxious, targeted, helpless, and uncomfortable regarding CDB attempts. Some even feel a sense of duty to thwart these attempts. However, somewhat paradoxically, many FLEs would prefer not to confront customers. For example, some feel that the policy itself is ineffective when firms do not ultimately prosecute deviant customers, that confronting CDB may result in violence or bodily harm, and that it is not their responsibility or is above their pay grade, implying differences across FLEs and FLMs. Thus, despite several firms having required guardianship policies in place, many FLEs are saying, “not my circus, not my monkeys.”

Authorized Guardianship policies permit the employee to interact with customers to actively confront and/or passively service customers if CDB is witnessed or suspected. So, it is at the discretion of the FLE whether and to what degree they engage with deviant customers. Giving employees the authority to determine how and when to handle CDB can be empowering, and these employees are more likely to commit to guarding against CDB. Larry says, “…when it comes down to it, in my opinion, what guy doesn’t want to be that hero right?”

Prohibited Guardianship represents policies where employees are prohibited from interacting with the customer in either actively confronting and/or passively servicing customers if CDB is witnessed or suspected. Many FLEs prohibited from taking action felt frustrated about their policies, making comments such as “I would like to be able to approach a customer” (Tiffany), and “That was one of the biggest things that I hated about it is like ‘cause you want me to prevent loss, but at the end of the day, if I can’t physically stop someone that I know is stealing essentially, I’m not loss prevention. I’m just, I’m here to just watch it” (Noah). Moreover, feelings of helplessness and exasperation are often associated with prohibited guardianship. For example, “[Policies] make it too easy. Shoplifting is more prevalent because people know this. It is not fair. They have restrained our powers. They have tied our hands. The criminals have all the rights, and we don’t” (Cassidy). However, when pressed on prohibition from confronting customers, some interviewees appreciated the restrictiveness for safety reasons.

Finally, many respondents (26.92%) were unaware of their firm’s guardianship policy suggesting, “I really honestly didn’t know like the exact procedures of how to go about it,” (Kendra) and “Not sure. I would say just talk to the manager and get his help” (Mark). Moreover, some firms have a policy in place for some types of CDB but not others. For example, for customer incivility, Penny suggests “We are supposed to de-escalate if customers are rude, but if it doesn’t help right away, the manager will intervene,” but when asked about shoplifting, she responds, “I don’t know. Surely, they don’t want it to happen.”

Employee Guardianship and Policy Fairness

As noted previously, firms’ guardianship policies outline acceptable behaviors for FLEs and FLMs when responding to situations involving CDB. These situations may involve enforcing customer compliance with mask mandates, de-escalation of incivility or aggression, or loss prevention tactics (Berry and Seiders 2008; Fombelle et al. 2020). As retail service providers shift the burden of responsibility for policing customers onto FLEs, employees are unequivocally affected by these growing demands (Voorhees, Fombelle, and Bone 2020).

Research suggests fairness “may be one of the most powerful sources of emotions in organizations” (Howard and Cordes 2010, p.411). Relatedly, organizational justice research has established an integrative approach across justice dimensions (i.e., distributive, procedural, and interactional; Blodgett, Hill, and Tax 1997), with the notion of fairness encompassing all facets of justice (Colquitt and Zipay 2015). When employees evaluate and assess their roles and employment policies, fairness perceptions derive from evaluations of general employment circumstances. Namely, employees assess their responsibilities, time, effort, and working conditions, including safety concerns (Colquitt, Greenberg, and Zapata-Phelan 2005). As such, the presence and type of guardianship may alter policy perceptions.

Thus, a focal variable of interest throughout our quantitative investigations is policy fairness, or employees’ perceptions of the acceptability, reasonableness, or justness of an organization’s policy (Pei, Paswan, and Yan 2014). Employment policies may aim to instill a sense of responsibility in employees to aid in CDB prevention, but these increased expectations of FLEs beyond their typical scope of work and without corresponding remuneration should decrease perceptions of policy fairness (Colquitt, Greenberg, and Zapata-Phelan 2005; Schepers and Van der Borgh 2020). Hence, we suggest the extra-role behavior outlined in guardianship policies will reduce employees’ fairness perceptions.

H1:

The existence (absence) of a guardianship policy is perceived to be less (more) fair.

Method and Results

Participants (N = 225) were recruited from an online panel (M_age = 40; 58% Male) and pre-screened for previous experience as FLEs in services retailing. The fairness measure consists of four semantic differential items where 1 = unfair, unreasonable, unjust, unacceptable; 7 = fair, reasonable, just, acceptable (Fennell, Coleman, and Kuo 2020), averaged to form a single variable (α = .93).

Supporting H1, the expanded role of FLEs compared with the control condition led to significantly less favorable fairness perceptions (M_Absent = 5.72, SD = 1.08; M_Present = 4.43, SD = 1.81; F(149) = 7.68, p < .001) using a one-way analysis of variance (ANOVA). The post-hoc Tukey’s test revealed fairness perceptions differed significantly between the control (M = 5.72, SD = 1.08) and required (M = 3.86, SD = 1.89; p = .02, 95% C.I. = [1.27, 0.18]), the control and authorized (M = 4.40, SD = 1.85; p < .001, 95% C.I. = [1.99, 0.64]) but not between the control and prohibited guardianship conditions (M_Prohibit = 5.11; p = .24, 95% C.I. = [-0.22, 1.41]).

Discussion

Study 2 set out to determine if the presence/absence of a policy outlining expanded roles and responsibilities for FLEs would be considered fair. Specifically, the presence of an employment policy requiring confrontation (e.g., Potdar et al. 2021) was perceived as less fair than a policy that did not outline a guardianship role. This provides evidence of the detrimental effect of guardianship roles on FLE perceptions of employment policies (Huang, Fang, and Liu 2021) and echoes findings from our qualitative study. Interestingly, the control and prohibited guardianship conditions, both of which do not outline extra-role behaviors, were perceived similarly.

The findings from Studies 1 and 2 suggest a more complex influence of guardianship policies extending beyond policy type. Specifically, Study 1 revealed that policies may pertain to active (confrontation) or passive (servicing) forms of FLE-shoplifter contact. Therefore, in Study 3, we aim to identify how policy types (required, authorized, prohibited) and approach styles (active, passive) jointly influence FLEs’ perceptions of fairness.

Guardianship Classification and Policy Fairness Policy Type

A firm’s expectations and permissions outlined by its policies affect perceptions of policy fairness (Colquitt, Greenberg, and Zapata-Phelan 2005). We posit differential effects for various policy types identified in Study 1 (prohibited, authorized, required guardianship), which are central to our investigation. When prohibiting employees from interacting with shoplifters the lack of autonomy to determine how best to handle the situation can reduce fairness perceptions (Gong, Yi, and Choi 2014). Interestingly, as found in Study 1 and extant literature (e.g., Pu et al. 2022; Tan et al. 2020), employees can experience anxiety and fear when dealing with CDB. This is especially the case with shoplifting, which presents the opportunity for escalation, potentially resulting in bodily harm, and is seen as not “worth it.” Prohibiting employees from confronting customers removes a sense of duty to defy shoplifters. Conversely, authorizing employees to confront shoplifters empowers them to decide when and if it is appropriate to intervene (Gong, Yi, and Choi 2014). In general, qualitative responses indicated that besides the disadvantages and the unfairness of being required to enforce guardianship policies, a lack of autonomy was perceived as frustrating and upsetting. Thus, in line with self-determination theory, employees are likely to view authorized policies more favorably, as they can act of their own volition (Hur, Shin, and Moon 2022). In fact, prior research suggests that employees’ sense of psychological ownership may exacerbate their likelihood of intervening in CDB given their feelings of personal pride or possessiveness toward their workplace (e.g., Potdar et al. 2018). If a firm requires employees to confront customers exhibiting CDB, employees may be left feeling powerless to make decisions that are in their best interest (Gong and Wang 2019). This is particularly true when employees feel that they are potentially endangering or exerting themselves to follow the policy. Thus, we predict:

H2:

Compared to authorized or prohibited guardianship policies, required guardianship policies will be perceived as less fair.

Approach Style

As noted above and in Table 2, approach styles can be active or passive in nature. Many firms implement active or confrontational approach styles; however, passive approaches may also be effective, as the mere presence or perceived alertness of employees may discourage CDB (Potdar, Garry, Guthrie, and Gnoth 2020). Likewise, social prevention strategies, ranging from employees making eye contact with or greeting customers to servicing and engaging in conversations with them, may be valuable tools for counteracting CDB (Fombelle et al. 2020). Notably, retailers may prohibit employees from confronting deviant customers but simultaneously require, authorize, or prohibit greeting or servicing customers. Thus, passive approach styles can empower employees to interact with customers while decreasing perceptions of endangerment derived from active approaches, allowing employees to feel less frustrated with and emotionally exhausted by their guardianship responsibilities (Song et al. 2021) while simultaneously experiencing greater autonomy (e.g., Potdar et al. 2018). Thus, passive approach styles, when used in lieu of an active approach, may be perceived as more just, suggesting that employees’ perceptions of fairness are impacted less by passive versus active approach styles.

H3:

Perceived policy fairness is less favorable for active vs. passive approach styles.

Method and Results

Participants (N = 302) were recruited from an online panel (M_age = 40; 56% Male) and pre-screened for previous frontline experience in services retailing. The fairness measure (α = .96) mirrored Study 2, and turnover intention (TOI) was measured with a 3-item Likert Scale (1 = Strongly Disagree, 7 = Strongly Agree; α = .96) adapted from Van Jaarsveld et al. 2021. Participants responded to the statement “confronting shoplifters should be left to managers” (1 = Strongly Disagree, 7 = Strongly Agree) followed by manipulation checks and demographic variables. The manipulation check items consisted of recall questions whereby participants were asked to indicate the guardianship policy and the approach style.

Both manipulations were successful as 94% of participants correctly identified the guardianship policy (χ²[12] = 185, p < .001) and 82% of participants correctly identified the approach style (χ²[6] = 126, p < .001). H2 considers the three different guardianship policy types, while H3 explores the two approach styles for their varying impact on fairness perceptions. An ANOVA procedure was conducted on the outcome variable, policy fairness, with guardianship policy and approach style as between-subject factors. We conducted an additional ANOVA procedure on TOI as a potential consequence of policy types. The procedures reveal a main effect of guardianship policy, F_fairness(1, 221) = 12.28, p < .001; F_TOI(1, 221) = 22.07, p < .001, a main effect of approach style, F_fairness(2, 221) = 13.83, p < .001; F_TOI(2, 221) = 19.72, p < .001, and a significant interaction between the terms, F_fairness(2, 221) = 4.43, p = .013; F_TOI(2, 221) = 14.91, p < .001 on policy fairness. When the policy outlines an active approach style, perceptions of fairness differ based on whether guardianship is required, authorized, or prohibited. Fairness perceptions and TOI are more favorable when confrontation is prohibited (M_fair = 5.04; M_TOI = 2.68) compared with required (M_fair = 2.59; M_TOI = 5.89), F_fairness(1, 221) = 29.18, p < .001; F_TOI(1, 221) = 62.95, p < .001. Authorized confrontation policies (M_fairness = 4.17; M_TOI = 3.32) were perceived more favorably than policies requiring confrontation, F_fairness(1, 221) = 12.17, p < .001; F_TOI(1, 221) = 40.69, p < .001. Moreover, a required passive approach is fairer (M_fairness = 4.48), resulting in lower TOI (M_TOI = 3.09), than requiring active confrontation.

No significant differences exist in fairness perceptions across approach styles when such actions are prohibited. When authorized, we find passive approach policies are more favorably perceived (M = 5.23) when compared to active confrontation (M = 4.17) F(1, 221) = 5.56, p = .019, but not for TOI. For required guardianship, passive approach styles (M_fairness = 4.48; M_TOI = 3.09) were perceived fairer and resulted in lower TOI than active confrontation F_fairness(1, 221) = 18.43, p < .001; F_TOI(1, 221) = 51.36, p < .001. Thus, in general, H2 and H3 are supported.

A post-hoc analysis was conducted to identify employees’ perceived differences across job roles (FLEs and FLMs). While we did not find differences across FLEs and FLMs for policy fairness and TOI, we did find differing views on who should be responsible for combatting CDB. When responding to the prompt “confronting shoplifters should be left to managers,” FLEs (M = 5.21) expressed a belief that FLMs should be the ones policing customers, a value significantly greater than what FLMs indicated (M_FLM = 4.88; t(225) = 1.61, p = .055). This further supports the “not my circus, not my monkeys” mentality amongst FLEs found in Study 1.

Discussion

Study 3 extends our understanding of how employees perceive and respond to guardianship policies. Echoing Study 2, we find that employees do not appreciate active, required guardianship. In addition to generating the least favorable fairness perceptions, requiring confrontation also leads to the highest level of TOI. Interestingly, we find notable differences when the policy entails a passive rather than active approach. We attribute this effect to the “not my circus, not my monkeys” phenomenon, whereby employees do not wish to place themselves in harm’s way through active confrontation. By explicitly communicating that active confrontation is prohibited, retailers can highlight the protection offered by the policy. While limiting the ability to confront suspected shoplifters may seem detrimental, empowering employees to passively approach and greet suspected shoplifters is perceived most favorably.

In Study 4, we aim to understand if factors outside the guardianship policy influence the perceptions of policy fairness, and as alluded to by Study 3, employees’ TOI. Namely, Study 1 identified employee feelings, such as anxiety, as a recurrent theme when asked about perceptions of CDB and their guardianship duties. Given prior results and the salience of the issues in Study 1, we focus on the active approach style. Additionally, we examine the most extreme or constraining forms of policy types (required vs prohibited) for employees, which resulted in the largest mean differences in Studies 2 and 3, respectively. This allows us to provide a clearer picture of how anxiety may further exacerbate fairness perceptions and turnover.

FLE-related factor: Anxiety

While increasing the scope and expectations of employee responsibilities is perceived as less fair, this relationship may be intensified by their emotional states. Added responsibilities can increase employees’ stress levels and may lead to less favorable fairness perceptions (Gong, Yi, and Choi 2014) and deteriorated emotional states (Grandey et al. 2012). Specifically, policies requiring (vs. prohibiting) confrontation of suspected shoplifters can cause employees to experience heightened levels of anxiety, which can influence fairness perceptions (Luo et al. 2014) and TOI (Vanderpool and Way 2013). Yet, some individuals are inherently more anxious about the idea of confronting customers. This aligns with research in trait activation theory suggesting that personality traits are triggered by contextual cues, such that traits become salient when given an opportunity to be expressed (Zahlquist et al. 2022). In addition to anxiety exacerbated by CDB (Van Jaarsveld et al. 2021), the thought of confrontation may activate feelings of anxiety and fear of social and physical ramifications of doing so. Thus, we predict:

H4:

Compared with prohibited guardianship, required guardianship policies will result in less favorable policy fairness perceptions among employees with high anxiety levels.

Turnover Intentions

Study 1 and Study 3 results also emphasize that policy perceptions have a more wide-ranging impact on employees through the activation of withdrawal behaviors. Particularly, the perception of unfairness in the workplace has been linked to TOI (Rubenstein et al. 2018). The mere existence of CDB can create stressful and frustrating situations, increasing employees’ intentions to leave the firm (Gong and Wang 2019; Pu et al. 2022). As employers assume more responsibilities as with guardianship, the negative feelings associated with CDB, role stress, and the urge to leave may be intensified (Van Jaarsveld et al. 2021). Formally stated:

H5:

The effect of guardianship policies on TOI is mediated by fairness perceptions.

Method and Results

Participants (N = 205; M_age = 37; 57% Male), prescreened for frontline retail or service roles, were recruited from an online panel. The fairness (α = .97) and TOI (α = .92) measures mirrored Study 2. The 3-item anxiety measure (1 = Strongly Agree, 7 = Strongly Disagree), adapted from existing scales (Vanderpool and Way 2013), asked that participants indicate how the prospect of confronting suspected shoplifters would lead them to feel “nervous,” “anxious,” or to “lose sleep” (α = .73).

To test H4 and H5, we conducted mean comparisons and a PROCESS mediation model (model 7; Hayes 2017), whereby the guardianship condition is the independent variable (0 = prohibited, 1 = required), TOI is the dependent variable, and fairness is the mediator. In addition, anxiety was included as a continuous moderator. As expected, when the retailer required guardianship, fairness perceptions were significantly lower (M_Prohibited = 4.96, SD = 1.44; M_Required = 3.55, SD = 1.89), t = 4.07, p < .01) and TOI increased (M_Prohibited = 3.59, SD = 1.58; M_Required = 4.65, SD = 1.49), t = 3.38, p < .01). Furthermore, and as shown in Web Appendix 10, the indirect effect of Guardianship → Fairness → TOI is significant for those individuals higher in anxiety (indirect effect = .65, Boot SE = .20; 95% bias-corrected confidence interval (CI95) [.32 to 1.00], but not for those individuals lower in anxiety (CI95) [−.29 to .31]. The index of moderated mediation (IMM) supports that the indirect effects are significantly different from one another (IMM = −.21(.08), (CI95) [−.37 to −.07]. Hence, H4 and H5 are supported.

Discussion

Results of Study 4 replicated findings regarding the impact of guardianship policies on fairness perceptions from Studies 2 and 3, further suggesting that anxiety is an important factor for organizations to consider in setting their guardianship policies as they can affect policy fairness perceptions. Specifically, employees anxious about confronting shoplifters are likely to perceive required guardianship policies as less fair than prohibited policies. Echoing Studies 1 and 3, less favorable fairness perceptions due to anxiety can increase turnover intentions. Uncovering propensity for anxiety in the hiring process may also be prudent in employee selection.

Study 1 highlighted that employees often perceive that guardianship responsibilities extend beyond their job role, and post-hoc analysis in Study 3 highlights FLEs perceive guardianship to be more of an FLM- (vs. FLE-) level responsibility. Expanding on this, Study 5 explores how job role influences fairness and includes a self-report measure of actual TOI.

Job Role factor: FLE versus FLM and CDB Frequency

Based on our interviews, organizational FLEs may perceive their working conditions to be a fruitless ordeal, where deviant customers and specifically shoplifters appear to have the upper hand. Compounding this, FLM roles are already considered to be more complex and demanding with greater external pressures than roles of FLEs (Lu et al. 2016). While organizational frontlines literature suggests FLEs are more likely to express greater TOI than FLMs (Lu et al. 2016), a high frequency of CDB can be difficult for FLMs, who may experience greater emotional exhaustion and decreased perceptions of justice (Gong et al. 2014), given they are called to step in for FLEs and are frequently charged by upper management with handling CDB. This perception of CDB frequency can exacerbate perceptions of unfairness, encouraging turnover (van Jaarsveld et al. 2021). For FLEs, who may hold a “not my circus” mentality, the effect of CDB and the prevalence on turnover is likely not as strong. But for FLMs, who cannot easily escape their roles policing customers, the adverse consequences of CDB frequency will be the pronounced. Following extant research (Gong et al. 2014), and in line with our qualitative and experimental support that FLEs expect to be helped by management when CDB or customer incivility warrants intervention, we predict:

H6:

Job role moderates the mediating relationship between CDB Frequency → Fairness → TOI, such that the effect is more pronounced for FLMs than for FLEs.

Method and Results

Practitioners (N = 173; 68.2% full-time, 53.2% female, M_age = 40.4 years; 45.1% FLEs and 54.9% FLMs) representing 17 different retail industries completed the survey. The most represented industries include clothing/apparel (24.7%), food/grocery (21.6%), hardware/home improvement (14.4%), convenience (15.5%), discount (6.2%), and consumer electronics (6.2%). Commonly reported job titles included sales (31.8%), sales managers (20.2%), customer service (19.1%), and store manager (12.7%). On average, respondents had 11.6 years of retail or service experience and FLMs held supervisory roles for over 6 years.

A Likert question (1 = Strongly Disagree, 7 = Strongly Agree) measuring shoplifting frequency was presented, stating “customers frequently take things without paying.” A single-item measure of perceived fairness was presented, followed by a measure of turnover intention—whether participants applied to other jobs in the past year as a result of CDB/related policies (1-never to 7-frequently). Additionally, participants were asked to indicate their CDB policy, perceptions of organizational support in the forms of formal training on policy specifics, as well as rewards or sanctions stemming from CDB policy compliance, all measured on a binary scale.

Overall, FLEs and FLMs agreed regarding shoplifting frequency (M = 4.17, SD = 1.91). Interestingly, however, the majority (54.7%) of FLMs indicated intentions to leave employers due to CDB and related policies, an amount notably greater than FLEs’ TOI (38.5%). To test for H6, we used the PROCESS model (Model 7; Hayes 2017), where CDB frequency was included as the independent variable, TOI as the dependent variable, perceived fairness as a mediator, and job classification as the moderator (1 = FLM; 2 = FLE). Supporting H6, the indirect effect of CDB frequency → Fairness → TOI is significant for FLMs (indirect effect = .26, Boot SE = .08; 95% bias-corrected confidence interval (CI95) [.11 to .42]), but not for FLEs (indirect effect = .07, Boot SE = .05; 95% bias-corrected confidence interval (CI95) [-.40 to .01]).

Post-hoc, we explored the data for further insights regarding the differences between FLEs and FLM indicated by our studies but not hypothesized: (1) guardianship policy types and organizational support and (2) policy enforcement training, rewards, and sanctions. First, nearly half (46.4%) of participants indicated their employer authorized them to confront suspected shoplifters. However, significant policy variations were present, with notable differences between FLEs and FLMs (χ2 [3] = 18, p < .001). Most (nearly 60%) of FLMs indicated they are authorized, whereas over 40% of FLEs indicated they are prohibited from confronting shoplifters (see Web Appendix 11: Figure 11.1), highlighting the differing expectations for FLEs and FLMs. Second, policy enforcement training, rewards, and sanctions offer further contextual insights into how firms influence employees to handle CDB (see Web Appendix 11: Figure 11.2). Less than 65% of FLEs, and 82% of FLMs, indicated they were offered training on how to handle CDB. The differences in organizational support also include financial incentives tied to shoplifting prevention. Most FLEs (82%) indicated there was no reward, whereas 40% of FLMs were incentivized to combat shoplifting. While training and rewards differ across job roles, over 80% of both FLEs and FLMs indicated there were penalties for breaking the organization’s policy.

Discussion

While FLEs and FLMs agreed on the frequency of CDB occurrence, the results highlight differences in perceptions and behaviors between FLEs and FLMs. Specifically, the results of Study 5 build upon the qualitative themes of Study 1 and experimental findings of Studies 2–4 and provide insight into how perceptions of fairness impact TOI, specifically for FLMs. The findings also reveal significant differences in organizational support with regards to training and reward structures across FLMs and non-managerial FLEs providing robust support for our conceptual model (Figure 1, Panel D).

Theoretical Implications of our Research

The new service reality requires firms to operate under considerable labor strain, heightened customer demands, and changing behaviors. Both employees and customers are increasingly apprehensive about this retail environment, which frames the landscape facing firms and highlights recent research calls for greater understanding of how dynamic service environments impact frontlines. Our work responds to these calls, including requests to identify contextual factors of policy impacts (Northington et al. 2021), changing employee roles in routine service encounters (Voorhees, Fombelle, and Bone 2020; Fombelle et al. 2020), and how firms require FLEs to cope with structural and managerial changes (Ostrom et al. 2021; Potdar et al. 2021).

Anecdotal evidence and popular press stories on CDB suggest that, especially post-pandemic, CDB is pervasive in retail services. Our work provides and reinforces examples of CDB with which employees must now contend. Specifically, we identify the five most common types of CDB with shoplifting recognized as the most prevalent and concerning for employees. By and large, FLEs do not believe they were hired to police CDB and specifically shoplifting, recalling the old Polish saying: not my circus, not my monkeys! This is a challenge for firms and their FLMs in retail services environments who lead that proverbial “circus” as ringmaster and must develop and implement effective policies to lead it.

Our research also develops and tests the first typology of guardianship policies in retail services. Research on firm policies focuses on reducing the prevalence of CDB through CSR-related aspects (e.g., Potdar et al. 2021) as opposed to guardianship. Other research examines ways to reduce CDB (e.g., Harris and Daunt 2013), prescribing employees perform guardianship duties, yet failing to outline policy specifics such as those presented in our novel typology or considering how these policies affect those asked to protect the store. Thus, our work applies organizational justice theory to the retail services context by considering the expansion of the FLE role and how fairness perceptions of guardianship policies are driven by context and expectations. Required guardianship, or mandatory policing of CDB as part of employees’ roles, is perceived as less fair than policies prohibiting such guardianship actions. Corresponding with self-determination theory, authorized guardianship increases autonomy and generates more favorable perceptions than required guardianship (Hur, Shin, and Moon 2022). Moreover, anxiety is activated when employees are expected to respond to CDB, and this anxiety can lead employees to consider alternative employment, further stressing already tight staffing strategies.

Our work generally contributes to the body of research exploring changing organizational frontlines. More specifically, we explore an often-overlooked dark side of retail services, namely, CDB, and the impact guardianship policies have on employees called to partner in a coherent response to this evolving challenge. While it is more expected that FLMs shoulder responsibility, as frontline roles expand in the face of growing CDB and declining labor resources, firms must confront the reality that most FLEs simply do not want this added responsibility.

Managerial Implications

Our findings yield prescriptions for managerially actionable strategies concerning guardianship for organizational frontlines struggling to reclaim control of the retail floor. Our results address an important question: should retailers expect employees to police CDB and therefore set guardianship policies in the first place? In general, employees feel that guardianship policies requiring or simply authorizing shoplifting enforcement are less fair than those prohibiting it. However, given the prevalence of CDB, as with shoplifting, as well as increasing labor and budgetary constraints, such guardianship policies may well be unavoidable. Thus, we provide insights for practitioners as to whether and how they should prohibit, require, or authorize employees to actively confront CDB or passively service such customers.

Our findings show that guardianship can engender strong, negative emotions for FLEs and FLMs, leading to perceptions of unfairness and increased TOI. Managing this is critical for talent retention (Gong and Wang 2019), particularly for FLMs who feel more responsible for handling CDB and who are more likely to consider alternative employment due to guardianship policies. Interestingly, our work reveals that FLEs begrudge requirements to actively confront shoplifters but appreciate the latitude and empowerment associated with being authorized to decide to (passively) service those they suspect of shoplifting. Hence, suggesting a combinatory approach may be most favorably received. Our findings suggest that training employees and framing such encounters to promote flexibility and employee safety may offer a middle ground between no guardianship role and full employee responsibility.

Moving beyond the perception of fairness, we also uncover that FLEs particularly predisposed to anxiety have the most aversion to guardianship policies requiring active enforcement. This suggests that firms should not only work to better identify FLE candidates with anxiety issues at hire but should employ training to help all employees reduce and manage the anxiety that comes with guardianship (Harris and Daunt 2013). Because our work also reveals that employees with heightened anxiety are more likely to view guardianship policies as unfair, and are more likely to leave, prioritizing candidate selection and training efforts will yield retention dividends later. Thus, acknowledging the full spectrum of CDB types, developing appropriate guardianship policies to address each, and preparing employees to succeed are imperatives that practitioners must carefully manage.

However, as CDB evolves and further challenges retail frontlines, firms should not view one singular solution for all types of CDB or all job roles. For example, while our findings indicate that both FLEs and FLMs generally agree that required guardianship is best left to loss prevention and security personnel, FLEs do feel pride in their roles and stores, and widely view as unfair any notion that shoplifters can “take” from them with impunity. However, both FLEs and FLMs largely feel that FLMs have a greater responsibility (than FLEs) when it comes to guardianship enforcement, particularly when active approaches are mandated. Such FLMs are most often authorized to confront shoplifters and are often incentivized to do so but are quick to express turnover intentions when they feel such requirements are unfair. Firms should employ a “train the trainer” approach so that FLMs tasked with leading FLEs feel more comfortable with and ownership of the guardianship policies they promote. Toward that end, the following approaches seek to further translate our findings into managerially actionable strategies to combat CDB with limited negative impact on firms’ most important resource, employees.

One of our first most unexpected findings, given the seemingly ubiquitous nature of CDB, was that more than one-quarter of respondents (26.92%) in our qualitative study were unaware of their firm’s guardianship policies. Thus, setting and communicating guardianship expectations to all employees is paramount. If employees are expected to combat CDB, it is incumbent upon firms and FLMs to train and lead FLEs by setting expectations early. This should include prioritization and emphasis on the most salient and pressing CDB issues and associated policies they might expect to confront in their roles. Beyond managerial interventions set forth by Gong, Yi, and Choi (2014) to help employees manage what they term “dysfunctional customers,” we offer some additional suggestions related to training, FLE selection, onboarding, and retail floor management: (1) Gamification of shoplifting enforcement in training tools; (2) Short periodical company training videos; (3) Role-playing and improvisation; (4) Coworker storytelling; (5) Customer conversation starters; (6) Personality screening for signs of heightened anxiety; (7) In-store signage and videos discouraging shoplifting; and (8) FLE training on how to avoid profiling while confidently/effectively addressing clear instances of shoplifting.

Thus, effectively implementing and managing guardianship policies and requirements from date of hire, setting expectations, and facilitating FLE partnership with FLMs likely offers the best and most comprehensive approach to combatting CDB and enhancing retention. Training should be conducted for both FLEs and FLMs and must include how to cope with, de-escalate, and handle CDB to increase employee self-efficacy and decrease burnout. Moreover, communication regarding why guardianship policies are in place (e.g., prohibited for employee safety), and what role FLMs versus FLEs can and should play, is also important to keep employees from intentionally or unknowingly breaking policies. Extant work suggests that “service firms may sometimes need to act as ‘police officers’ to ensure that their customers behave appropriately,” implementing policies for managing deviant behavior (Huang, Lin and Wen 2010, p. 159). Our research shows that, in short, it begins with a managerial strategy for FLEs to understand, appreciate, and embrace guardianship policies as a mechanism for maintaining control of the retail floor and retention of FLEs who want to play a positive role in driving store success. It ends with ensuring well-trained and committed FLMs view their responsibilities as being fair, and with their demonstration of guardianship for FLEs looking to them for guidance.

Future Research Agenda

While our multi-method research approach offers insights to academia and practitioners, it also raises additional questions, sometimes in light of our limitations, that warrant future research exploration in this domain. Therefore, we propose an extensive research agenda, detailed in Table 3 and discussed below, which offers research questions and supporting cites across seven focal categories pertinent to expanding development of this research stream.

OPEN IN VIEWER

Table 3.

Related Future Research Agenda.

Research Topics- Subthemes	Key Research Questions	Relevant Support
Guardianship typology- Policy types versus approach styles- Empirical validation of guardianship	• What other mediating mechanisms or boundary conditions (see Web Appendix 8) might be tested as it relates to the typology introduced here?• Are there other dimensions of guardianship that might warrant typification?• How might different types of CDB in retailing services (see Web Appendix 5) benefit from different types of guardianship policies? Should conditional or combinatory policies be used to account for different types of CDB?• How might longitudinal studies offer insights into FLE policy perceptions (see Web Appendix 7) over time/with increased CDB frequency?• Which policies increase FLEs’ satisfaction in their roles?• How might attitudes toward and preferences regarding types of guardianship policies differ across various national or cultural contexts?	See Table 2 and Web Appendices 6, 7, and 8 (this research);Field et al. 2021;Ostrom et al. 2021;Korgaonkar et al. 2020; Potdar et al. 2021
Policy conveyance and enforcement - Employee involvement - Training - Changing perceptions of policy fairness - Rewards/financial incentives - Sanctions - Firm reputation - Message framing	• How might providing rewards or sanctions for CDB intervention impact employee actions, attitudes, and behaviors? Might rewards create unnecessary competition among FLEs? How might sanctions alter FLE policy perceptions?• Are formal or informal policies more effective? How does this impact enforcement?• How might firms convey the value of their guardianship policies to FLEs/FLMs? Can message framing enhance receptivity, fairness perceptions, and buy in?• How might firms take FLE dignity, wellbeing, and fairness perceptions into consideration when designing and conveying guardianship policies?• Does increasing FLE/FLM “voice” in designing guardianship policies improve employees’ willingness to adhere to and/or implement policies?• Can measures aimed at increasing FLE engagement help to increase acceptance of and adherence to guardianship policies?• Do employees expect additional compensation commensurate with new responsibilities required by firms’ guardianship policies?• How might intraorganizational social capital (e.g., collaborative relationships between FLEs within firms) or cross-level relationships (e.g., leader-member exchange) impact FLEs’ willingness to accept and adhere to guardianship policies?• Who is ultimately responsible for ensuring employees understand the guardianship policy (e.g., FLEs, FLMs, HR, etc.)? Likewise, who should be accountable for monitoring compliance and enforcing guardianship policies?• Do store reputation and policies impact whether overall CDB increases or decreases (e.g., if customers are aware of guardianship policies and if they are enforced)?• How might external mandates (e.g., mask wearing and social distancing policies during COVID) impact firms’ ability to effectively convey and enforce policies?	See Web Appendix 8 (this research);Kalra et al. 2021;Field et al. 2021; Potdar et al. 2018
Job role and tenure with firm- FLEs versus FLMs responsibilities, roles, etc.- Tenure with firm	• How might employees respond to different guardianship policies based on different job levels/roles? Should different policy types be used contingent on job role?• If FLEs are seen as the “eyes and ears” of the organization, how might FLMs/firms use this competitive intelligence to address and reduce instances of CDB and better design firm policies?• Do perceptions of job autonomy differentially impact FLEs versus FLMs willingness to engage in guardianship for the firm?• How can FLMs instill a sense of shared purpose with FLEs in combatting CDB, implementing guardianship strategies, and reducing turnover intent?• How might factors such as organizational identification impact fairness perceptions and turnover intentions following the implementation of guardianship policies?• Does tenure of employment impact perceived fairness of policies? Moreover, are there differences for FLEs who experienced several policy changes versus new hires or those with a stable policy?• Does job role (FLM vs FLE) impact expectations of compensation or sanctions for guardianship?	Kalra et al. 2021;Field et al. 2021;Van Jaarsveld et al. 2021;Rubenstein et al. 2018
Legal ramifications and protection of vulnerable populations - Legal issues - Protection and safety of employees - Stereotyping and biases - False accusations - Fairness for all stakeholders	• How are legal issues driving the implementation of guardianship policies?• How can governmental infrastructure help reduce CDB in retailing services?• How is a(n) un/willingness by firms and police to prosecute deviant customers impacting FLEs’ willingness to implement guardianship policies? Does it increase the likelihood of FLEs’ engaging in deviant behaviors either through retaliation, providing personal justification, or otherwise?• What are the financial and social costs for firms when false or wrongful accusations are made regarding CDB? How can such occurrences be mitigated, particularly for vulnerable and disadvantaged populations?• How are personal prejudices/biases impacting FLEs accusations of CDB and how might firms mitigate the negative effects of biases (e.g., stereotyping) in effectively implementing guardianship policies?• Should firms call on customers to engage in guardianship practices (i.e., “see something, say something”)? What are potential legal issues facing such practices?• Do employees react differently to government versus firm mandated polices (e.g., mask wearing vs. shoplifting)?	See Web Appendix 8 (this research);Field et al. 2021;Potdar et al. 2020; Korgaonkar et al. 2020
Social implications and contextual factors - Personal values - Moral attributes - Social pressures	• How might some of the social pressures identified in Study 1 be mitigated?• How might FLE reactions, affect, and guardianship roles be impacted if the deviant customer is known by the FLE or a loyal customer of the firm?• Do FLEs respond differently to instances of CDB by similar others (i.e., those with perceived similarities with the FLE)?• How might other customers’ requests to intervene or observations on CDB play a role in FLEs’ willingness to protect or guard the store?• How might inaction by FLEs (including FLMs) following instances of CDB positively or negatively impact the attitudes and experiences of other witnessing and/or complaining customers?• How do FLEs and customers impact each other’s service experiences?• How does the customer orientation of FLEs impact their enforcement of guardianship policies? How do FLEs deal with the potential tension between customer and firm interests?	See Web Appendix 8 (this research);Potdar et al. 2017; Korgaonkar et al. 2020;Subramony et al. 2021
Frontlines in change- Policies changes over time- Influence of COVID-19- Temporary versus permanent changes to service	• What are ways employees are coping with the implementation of or changes to guardianship policies over time and/or following a global pandemic?• Likewise, how might other market shocks (e.g., COVID-19, wars, inflation, etc.) influence FLEs’ responses to CDB?• How are other market shocks exacerbating FLE anxiety and negative affect regarding confronting CDB?• For shoplifting, might the nature and value of the offering that is stolen impact FLEs’ willingness to follow through with intervention (e.g., if customers are seen as being desperate or not having means to pay for necessary vs. frivolous items)?• How might implementing guardianship policies help firms to improve profitability during an era of store closings, rising inflation, and supply chain/distribution issues?• How can guardianship policies and training potentially increase agility at both the FLE and firm levels?	See Web Appendix 8 (this research);Field et al. 2021;Vorhees et al. 2020
Other considerations - CDB Frequency - Technology integration, usage, and innovation	• How might the frequency of CDB occurrences impact FLE perceptions of fairness and intentions to remain with the firm?• Does the size of the firm affect FLEs’ willingness to follow guardianship policies?• Are and should there be different policies for corporate firms versus privately owned or small businesses?• Are new pricing models needed to offset increasing losses incurred due to CDB?• What are new applications for CDB and guardianship in digital/online settings?• With the increase of access-based consumption and customers taking on FLE roles, do fairness perceptions of customers matter? And, if so, how does it influence their service experience?	Potdar et al. 2020; Voorhees et al. 2020; Field et al. 2021; Ostrom et al. 2021Hazée et al. (2017)

First, given we introduce and test the first guardianship typology in retail services, work exploring other dimensions of guardianship and conducting empirical tests considering additional mediating mechanisms and boundary conditions related to typifying guardianship seems pertinent. Second, research exploring policy conveyance and enforcement, for example, considering the framing of messaging related to guardianship policies, is worthwhile. For example, if phrased as necessary to “benefit the greater good” or as to “protect employees” versus to “protect the bottom line,” it is possible that employee perceptions of guardianship may differ. Additionally, Study 1 introduced varying views regarding the use of incentives and sanctions for policy enforcement; thus, it warrants consideration of how these can be used for intervention and if it can lead to overzealous employees, false accusations, or ignoring CDB altogether. Third, given insights introduced in Study 1 and tested in Study 5, we show there are different responses, expectations, and even policies depending upon job roles. So, additional research into impacts for and differences between FLMs versus FLEs, as well as tenure with the firm, could be considered.

Fourth, there are several legal ramifications for firms and individual employees when combatting CDB. Therefore, investigating how legal and safety issues are driving the formation and implementation of guardianship policies is of considerable interest. Fifth, overall retail service evaluations and fairness perceptions are inherently subject to the constraints of social impact theory, such that the presence of customers and other contextual factors can impact both employee and customer perceptions. Factors identified by Study 1 that may influence responses to CDB and guardianship expectations include: (1) value of items stolen, (2) FLMs’ stressing importance of preventing CDB, (3) presence of other customers (4) attitudes of offending customers, (5) other FLEs’ attitudes and actions, (6) perceived racial bias/stereotyping, (7) FLE pride, and (8) use of video surveillance monitoring. Hence, there are numerous social and contextual factors in relation to CDB and guardianship which can be explored (Subramony et al. 2021). Sixth, we identified numerous contextual factors related to changing organizational frontlines, particularly related to COVID-19, which provide inspiration for future research. For example, scholars can explore the effects of various market shocks on CDB, policy and guardianship interventions, and ways to mitigate negative effects of these changes. Seventh, as frontline guardianship policy research is in its infancy, factors related to the frequency of CDB as well as technological innovations and usage that enable CDB or assists in its prevention and policy implementation appear fruitful. Finally, retailers continue to exist, evolve, and confront problems inherent in the dynamic retail services environment. Consequently, online retailing serves as another context to investigate CDB and guardianship policies to combat it.

In sum, whether confronting shoplifting, enforcing mask policies, or policing relatively innocuous or inadvertent CDB, the role of FLEs and FLMs as the first lines of defense for both firms and other customers against disruptive infractions is unlikely to end. As understanding how to manage that evolution effectively is the joint responsibility of researchers and practitioners, with the goal of protecting all stakeholders, we provide extensive research into this domain alongside ample suggestions for future exploration.