Do a non-core worker's procedural justice concerns influence their engagement in helping behavior? A multi-method study

Abstract

Purpose

The purpose of this research is to examine how a change in team dynamics impacts an individual's motivation to engage in helping behavior and operational performance.

Design/methodology/approach

An online vignette experiment and a hybrid discrete event and agent-based simulation model are used.

Findings

Study findings demonstrate how a non-core worker's perception of team dynamics influence engagement in helping behavior and system performance.

Originality/value

This study provides a further understanding on how team members react to changes in team processes. This study theorizes on how an individual team member responds to fairness concerns. This study also advances our understanding of the critical importance of helping behavior in a retail logistics setting. This research illustrates how the theory of strategic core and procedural justice literature can be adopted to explain team dynamics in supply chain management.

Citation

Farhan, M., Krejci, C.C. and Cantor, D.E. (2023), "Do a non-core worker's procedural justice concerns influence their engagement in helping behavior? A multi-method study", International Journal of Physical Distribution & Logistics Management, Vol. 53 No. 9, pp. 1015-1042. https://doi.org/10.1108/IJPDLM-02-2022-0044

1. Introduction

The services provided by logistics and supply chain workers in the retail sector can significantly impact a customer's experience. In restaurants, for example, many workers who oversee food and beverage services are responsible for inventory management and replenishment, and hence, their decisions can have immediate consequences for customer service (e.g., stock-outs leading to lost sales and damaged firm reputation). While this holds true across many other industries, including the manufacturing, defense, and public sectors (Swink et al., 2022), it is of particular importance to the retail logistics sector, which was valued at $219.5 billion in 2021 and is expected to increase approximately 12% in 2022 (Global Newswire, 2022). Given this enormous growth, it is critical to recruit and prepare logistics workers for rewarding career opportunities in this and related industries. However, improving the performance of logistics workers remains a serious challenge for several reasons, including high employee turnover (MH&L 2022) and significant problems with team productivity (Haefner, 2011). Instable and, hence, poorly motivated teams in retail and many other industries can have negative ramifications for customer wait times, labor productivity, and overall firm performance.

One step that logistics managers should take is to consider the differences between team members' roles and responsibilities and their impact on team productivity and performance. Indeed, the concept of “core workers” refers to the situation where some employees (e.g., assembly or production line leaders) have a more visible or central responsibility in the completion of logistics tasks than other “non-core” team members (Bartholdi and Eisenstein, 1996; Doerr et al., 2002, 2004). However, non-core workers have important responsibilities as well, including providing information, material, and related resource support to the core workers on their team. For example, non-core “water spiders” monitor and replenish inventory to help the production line avoid delays and interruptions. Other relevant examples are illustrated in Table 1, such as less-than-truckload drivers who have the central responsibility or “core role” of providing same-day transportation services but rely on non-core “driver helpers” for assistance with loading and unloading freight (e.g., Lu et al., 2022). In the context of food service, prep cooks serve as non-core “material replenishers” in the kitchen and are responsible for restocking line cooks' stations with inventory, while non-core bussers provide support to servers in the delivery, replenishment, and clearing of food and beverages at customer tables.

In this study, we consider how policies that are targeted to core workers are perceived by and impact the non-core workers on a team. Organizations naturally value the core worker's career experience (e.g., tenure at the firm), and as a result, non-core workers may perceive that their own procedural justice concerns (e.g., regarding promotion opportunities) are not being addressed since they have less experience (Humphrey et al., 2009). If non-core workers' procedural fairness concerns are not sufficiently attended to by the organization, they may engage in withdrawal behavior and eventually resign from the organization. Thus, a manager may not have sufficient labor capacity to meet customer demand (Cantor et al., 2011). Indeed, it is common for managers to struggle to meet demand because of employees' staffing and scheduling concerns (Lucas, 2022; Alsever, 2022). Furthermore, non-core workers who perceive unfair treatment may not engage in prosocial behavior in the workplace, such as helping behavior.

The purpose of this research study, then, is to theorize about how perceptions of procedural justice regarding the non-core team member's role, team dynamics, and organizational policies impact their motivation to engage in helping behavior, and the subsequent overall impact on operational performance (e.g., customer service time). Helping behavior, defined as one worker's providing of assistance to another worker on a voluntary basis, is critical to worker and team productivity across many supply chain settings, such as warehousing, transportation, manufacturing, and retail (see column 4, Table 1). Our study thus builds-upon prior supply chain research by capitalizing upon the procedural justice and strategic core literature to study how perceptions of organizational policies influences helping behavior (see Table 2 for a brief list of factors relevant to this study) (e.g., Hofer et al., 2012; Griffis et al., 2012; Cantor et al., 2011; Summers et al., 2012; Humphrey et al., 2009). In particular, we extend the work of Humphrey et al. (2009) to the contexts of logistics and service operations management by providing a new theoretical perspective on how team dynamics are influenced by procedural justice and workflow concepts. Specifically, our study extends prior logistics research that has not considered the impact of a non-core team member's procedural fairness concerns about their work environment.

We test our research questions using a multi-method approach. In Study 1, we conduct an online vignette-based experiment to examine how a non-core team member's reaction to an expansion of a core team member's role motivates the non-core worker to provide help to the core worker. Because these behaviors commonly occur in a dynamic setting, our second study builds upon the findings from Study 1 through the development of a simulation model. In Study 2, using hybrid discrete event and agent-based simulation modeling techniques, we further explore how team dynamics influence helping behavior and overall performance.

We empirically test the study's hypotheses in the retail (restaurant) sector for several reasons. First, this service context consists of core workers (e.g., restaurant managers and lead servers with several years of work experience in the restaurant industry) and non-core workers (e.g., prep cooks, bussers, and related wait-staff with relatively less experience), thus offering a rich environment to study team dynamics among employees and their productivity (Swink et al., 2022). We add to prior research that has studied the dynamic interactions of logistics workers (see Table 2: Keller and Ozment, 1999; Cantor et al., 2011). Second, the workers in this setting manage inventory on a daily basis (i.e., food and beverage services to retail customers), and their decisions can have an immediate impact on metrics such as customer wait times and service levels (see Table 2: Schultz et al., 1998, 1999). Third, the restaurant industry is experiencing significant employee turnover: the Bureau of Labor Statistics reports a 57% turnover rate in the restaurant industry, representing 230 million days of lost productivity (DailyPay, 2021), and the National Restaurant Association reported that the industry was unable to fill 750,000 jobs in 2021, representing approximately 6.1% of its workforce. Turnover affects team dynamics, team productivity, and customer service levels (see Table 2: Griffith et al., 2006; Cantor et al., 2011). As a result, many restaurant managers are struggling to provide consistent customer service experiences, with employment shortages resulting in a 23% increase in customer wait times and, in some cases, limited daily operating hours (i.e., reduced productive capacity). Thus, an implication for this study is that managers need to consider strategies to maintain a stable workforce that is well positioned to advance the collective team's and organization's interests.

2. Literature review and theoretical background

2.1 Strategic core theory

The primary theoretical basis of this study is the theory of the strategic core. Previous organizational behavior research has proposed the theory of strategic core as a basis to understand the dynamic interaction of team members (e.g., Humphrey et al., 2009; Summers et al., 2012). These team scholars have theorized and provided empirical evidence that a core worker's career experiences and the nature of their job roles and responsibilities in organizations can have a noticeable impact on team performance (Delery and Shaw, 2001; Humphrey et al., 2009; Summers et al., 2012). In their description of the strategic core theory, Humphrey et al. (2009) explain that team members with greater career experience are more knowledgeable on how to perform their job efficiently and effectively. Further, workers with greater experience are more able to successfully react to disruptions in the workplace. Summers et al. (2012) used the theory of the strategic core as a basis to examine how the flux in the coordination construct captures the disruption a team undergoes during team membership replacement. Summers et al. (2012) and Humphrey et al. (2009) comment that the strategic core role is the position on a team that is (a) primarily responsible for solving the problem or task that is presented to the team, (b) required to perform a greater number of certain tasks on the team, and (c) central to the functioning (or workflow) of the team. Moreover, Humphrey et al. (2009) further emphasized that a strategic core job role can be thought of as a continuum: the more a job role meets these criteria, the more “core” the job role is to the team.

While the theory of the strategic core has not previously appeared in the logistics literature, we believe that this perspective about team behavior has significant utility in logistics research. As an example, in the supply chain literature, scholars point out that team members having more career experience may occupy a more central role in the team's operations compared to their colleagues. Indeed, the concept of a core role appears in the warehousing and logistics literature, where some employees are assigned to an assembly line based on their work rates (e.g., slowest to fastest) and experience in performing the task (Bartholdi and Eisenstein, 1996). Workers can obtain additional work by looking to upstream stations and move completed work to downstream stations (Schultz et al., 1998). Supply chain scholars have noted that some workers are more critical to the productivity of teams because they work faster and are more proficient in processing work-in-progress inventory, based on their previous experience (Doerr et al., 2002, 2004; Humphrey et al., 2009). We add to this stream of research by theorizing about how and why the interactions among core and non-core members of the team could impact operational performance metrics studied in prior logistics research (e.g., wait times and customer service performance).

2.2 Procedural justice theory

We enhance our understanding about the social dynamics among core and non-core team members by integrating the procedural justice literature into this study. The procedural justice literature describes that some individuals are more attentive to the firm's organizational policies/practices and supervisor-employee interactions than others (Griffith et al., 2006). In this regard, before an employee becomes more committed and engages in pro-organizational behaviors, the individual will assess how their feedback is being listened to and incorporated into new organizational and supervisory practices. Put simply, employees assess how they are treated by individuals who are in a position of authority. These perceptions could influence the employee's decision to engage in behaviors consistent with the interests of the organization (Griffith et al., 2006).

Procedural justice scholars suggest that a key aspect of the theory is the concept of social exchange. Dynamics of social exchange are present when individuals help others because the helpee was treated fairly. It logically follows that helping behavior is an important outcome of procedural justice. Indeed, helping behavior is a prominent type of prosocial behavior commonly examined in the justice and organizational commitment literature (Konovsky and Pugh, 1994). Individuals are more inclined to help others when they are treated fairly by their supervisor and other employees. In contrast, individuals will engage in withdrawal or retaliatory behavior in situations where the person feels distrust or injury caused by other participants in negative or threatening situations.

The procedural justice literature can be used to examine how the fairness of team policies and procedures influences the interdependent nature of logistics and production activities among core and non-core team members. Previous justice research has studied the attitudes and behaviors of the people influenced by those who occupy positions of authority (Korsgaard et al., 1995; Lind and Tyler, 1988; Moorman, 1991; Moorman et al., 1998). This literature can provide valuable insight into why supervisors and core team members should make decisions in a transparent and fact-based manner. Moreover, core workers should leverage their career experiences to offer reasonable suggestions to the organization, including non-core workers, on how to fairly resolve potential problems such as work scheduling concerns, career advancement strategies, poor worker productivity, and/or employee absenteeism and turnover (Hofer et al., 2012). Procedural justice concepts have been used in supply chain research (e.g., Cantor et al., 2011; Griffith et al., 2006; Hofer et al., 2012; Liu et al., 2012) but not in prior research that has adopted the strategic core theory.

2.3 Research contribution

We seek to contribute to the literature by juxtaposing the strategic core and procedural justice theories to explain how perceptions of fairness influence a non-core worker to provide helping behavior to the core worker on the team and the subsequent impact on operational performance. The fundamental tenet of our model is that non-core workers who perceive that procedural fairness elements are present in the work environment will become more motivated to help other non-core workers and the core worker in the team. Specifically, we examine how a non-core worker's perception of fairness regarding the promotion of the core worker and the assignment of work schedules will impact their decision to provide help. Perceived violations of the non-core worker's norms of reciprocity are also studied, as well as the impact of their assigned workload and their tendency toward collectivism. We then test the degree to which helping behavior by the non-core worker will improve the organization's overall operational performance. A summary of literature most pertinent to this study is presented in Table 3. Figure 1 provides a depiction of our theoretical model.

3. Hypothesis development

3.1 Fairness perceptions

Our first hypothesis focuses on the linkage about a non-core worker's perception of fairness and their decision to engage in helping behavior. In the context of our study, the research participant is the non-core worker, and she/he is told that another employee in the organization was promoted into a more prominent role and provided additional job responsibilities (e.g., stretch work). In particular, we contend that the non-core worker will develop impressions of fairness about the policies, rules, and practices concerning the promotion process (Bies, 1987; Folger and Konovsky, 1989; Moorman, 1991; Moorman et al., 1998; LePine et al., 2008). In many organizations, non-core workers observe how others such as core workers and supervisors create and promulgate new organizational policies and practices (Lind and Tyler, 1988). Because employees want assurances that they will have the opportunity to pursue job promotion opportunities, they are more likely to be satisfied with favorable outcomes if they believe that the procedures used to derive those outcomes are fair and equitable (Folger and Konovsky, 1989; Lind et al., 1990; Lind and Tyler, 1988). However, non-core workers may perceive that core-workers may have not been promoted fairly even though these senior workers have accumulated significant organizational experience and task competency (Humphrey et al., 2009). As a consequence, non-core workers may feel undervalued and form lower feelings of commitment towards the firm, leading to a lower likelihood of pro-social behavior (Cantor et al., 2012; Ilgen et al., 2005) and increased withdrawal behavior (Moorman et al., 1998).

Nevertheless, there are some reasons to believe that an organization's willingness to address fairness concerns regarding job promotion opportunities will not motivate pro-social behavior. Some employees do not want to be burdened with the stress of managing employees or extra responsibilities. Others are more concerned about supplementing their income to cover part of their post-secondary education costs and hence promotion opportunities are not a primary concern. Notwithstanding these and other counter arguments, many workers are concerned about procedural fairness in the workplace because they want to pursue career advancement. Thus, we present the following hypothesis.

H1.

In the situation where there is the perception that the core worker was promoted fairly, the non-core worker is more motivated to provide helping behavior to the core worker than the situation where there is the perception that the core worker was not promoted fairly.

3.2 Work schedules

Work schedules will directly influence fairness perceptions and pro-social behavior among the non-core workers on the team. When an organization's work scheduling practices are in-conflict with the employee's preferences, it can lead to relationship conflict over time. Indeed, Matusik et al. (2019) point out that relationship conflict may result in a breakdown of team processes, even when the employee has significant work experience.

The procedural justice literature provides us with further insight into why relationship conflict emerges when an employer does not create a desirable work schedule based on the work experiences and preferences of their employees. Scholarly research about the US motor carrier (trucking) industry contains several good examples in this regard. In the trucking industry, motor carrier employees and independent contractors are responsible for hauling freight across the supply chain. These employees have a desire to perform their work based on their seniority (e.g., career experience) and work scheduling preferences (e.g., Cantor et al., 2011). While more senior workers expect to receive more favorable work schedules, industry participants recognize that it is challenging to recruit new workers into the industry. Thus, there is tension on how to best accommodate both worker demographics. Nevertheless, commercial drivers have expressed serious procedural fairness concerns about how their work scheduling preferences are restricted based on the use of electronic monitoring technology to enforce industry work scheduling rules (e.g., the US DOT hours of service rules). There was anecdotal evidence indicating that some drivers would rather illegally circumvent the policy, thus creating relationship conflict between the drivers, their supervisors (dispatchers), the firms, and government policy makers. Moreover, Cantor et al. (2011) theorize and empirically demonstrate that the policies associated with work scheduling has resulted in increased employee turnover intentions among truck drivers. Likewise, Miller et al. (2013) studied the effects of formal controls on an employee's work hours. Collectively, these and related studies demonstrate that employees are motivated to either evade governmental and/or organizational policies that are perceived to be intrusive or leave a place of employment rather than become subjected to the enforcement of unfair work scheduling rules. This perspective has generated significant animosity among commercial drivers who have accumulated many years of career experience in the industry.

To summarize, the literature suggests that employees will not engage in pro-social behavior if their concerns about work-related rules are not attended to in a fair manner. Put simply, when disagreements occur in the workplace, workers will reduce their helping behavior (e.g., Morgeson et al., 2015). Thus, we theorize that non-core workers who are not treated fairly regarding work scheduling will avoid engaging in pro-social (helping) behavior.

H2.

In the situation where there is the perception that the non-core worker was treated fairly regarding her or his preferred work schedule, it is more likely that the non-core worker will be motivated to provide helping behavior to the core worker, compared to the situation where there is the perception that the non-core worker was not treated fairly regarding her or his preferred work schedule.

3.3 Norms of reciprocity

The concept of norms of reciprocity is an important construct in our model as well. Norms of reciprocity are defined as the assertion that one person is willing to help another individual so long as the helping provider believes that he or she will receive help from the helped person in the future (e.g., Rhoades and Eisenberger, 2002; Shore et al., 1995). The norms of reciprocity concept has received theoretical and empirical attention in the supply chain management literature (e.g., Cantor et al., 2012; Griffith et al., 2006; Pulles et al., 2014). We build on the nascent stream to suggest that supply chain employees will engage in behavior that supports either the organization's or supervisor's goals so long as the organization is willing to reciprocate with appropriate leadership, training, and rewards. This issue is particularly important to non-core workers who have less career experience and are reliant upon their more senior coworkers for resources to become successful in their careers. Scholars have used social exchange theory to support some of their empirical evidence that employees are encouraged to meet organizational obligations and expectations without relying on contractual governance mechanisms (Wang et al., 2022). This is particularly true among employees who have accumulated a wealth of career experiences.

We contend that the non-core worker will provide help to the core worker based on the non-core worker's professional interactions with the core-worker. However, should the core worker regularly need help from the non-core worker, the non-core worker may develop feelings of frustration, stress, and burnout, which will reduce the non-core worker's job performance and the likelihood of providing future help to the core worker (Barnes et al., 2008; Bolino et al., 2015; Cantor and Jin, 2019). We believe that if there are violations in the non-core and core workers' reciprocity expectations in helping behavior, the non-core worker will no longer have the motivation to help the core worker in the future.

H3.

In the situation where there is the perception that the non-core worker's norms of reciprocity were not violated, it is more likely that the non-core worker will be motivated to provide helping behavior to the core worker, compared to the situation where there is the perception that the non-core worker's norms of reciprocity were violated.

3.4 Capacity to provide help (workload)

A non-core worker's capacity to provide help is impacted by the demands placed upon workers in many service settings (e.g., Dimotakis et al., 2012). Insights from equity theory and the social loafing literature are used to support this theoretical relationship. Equity theory suggests that a worker will compare her or his capacity to carry out job responsibilities relative to other workers in the team and/or organization to determine the amount of effort they will put forth into their job responsibilities. If the worker believes that she is not receiving her fair share of resources from other co-workers and/or supervisors, then the worker will engage in withdrawal behavior. Likewise, the social loafing literature is also useful to further our understanding. Social loafing is the tendency for workers to exert less effort when working in teams than when working alone (Li et al., 2014). Cantor and Jin (2019) found that workers in a production line experiment who perceived high social loafing levels engaged in lower levels of helping behavior. Schultz et al. (1999) theorize that social loafing theory can explain why some workers operate machines slower than other workers on an inventory management task. In their study on team absence norms, Ten Brummelhuis et al. (2016) use social exchange theory to argue that, when an employee's workload increases because team members are absent, his/her sense of unfairness may lead to calling in sick more often.

Non-core workers who have limited resources will engage in lower levels of helping behavior because they have developed less expertise due to a lack of career experience. A worker's available resources are determined by their capacity to provide service to others based on their current workload and collective career experiences. A non-core fully utilized worker will monitor their surroundings and recognize that they contribute more to assist the team or organization than other workers including colleagues who have much more career experience. Thus, there is a non-equitable distribution of resources, making it unreasonable for the non-core worker to contribute more assistance to the team. The non-core worker will perceive it as unfair to provide helping behavior when they are currently contributing significantly to the team.

H4.

If the non-core worker's resources are substantially utilized, the lower the likelihood that the non-core worker is motivated to provide helping behavior to the core worker compared to a situation where the non-core worker's resources are significantly underutilized.

3.5 Collectivism

Many teams are affected by a non-core worker's predisposition to act collectively. Collectivism refers to a general desire among individuals for greater equality among society or individuals within an organization (Colquitt et al., 2002; Colquitt et al., 2005). Individuals who espouse the collectivist perspective believe that it is important for employees to have a voice in the workplace. In general, these individuals desire fair action, generosity, kindness to others, and a caring society. This perspective is shared by Colquitt et al. (2002) who suggest that team collectivism could influence perceptions of team climate and team performance.

The collectivism concept has received attention in the buyer-supplier literature as well. Lee et al. (2018) discuss that suppliers who operate in locations characterized as having a collectivist culture exert more effort and shirk less. Eckerd et al. (2016) theorize that individuals from collectivist cultures (e.g., China) are more willing to trust their supply chain partners than people from individualist cultures (e.g., USA). Handley and Angst (2015) find that relational governance is more effective in collectivist societies.

We contend that employees who value collectivism are more likely to engage in helping behavior. Generally speaking, collectivistic individuals are caring people and do not want others to face hardships in the workplace (Colquitt et al., 2002). When other people struggle to complete their work, collectivist employees desire to engage in pro-social behaviors to promote a collegial environment. These individuals attempt to set an example that altruistic behavior is expected.

H5.

In the situation where the non-core worker's collectivism level is high, the greater the likelihood that the non-core worker is motivated to provide helping behavior to the core worker compared to the situation where the non-core worker's collectivism level is low.

3.6 Helping behavior

We suggest that as non-core workers provide help to others, team performance should improve. While there are clear benefits of helping behavior, there isn't unequivocal evidence that team performance will improve. Previous research suggests that helping providers may become fatigued or disgruntled with having to help others. Helping involves communication and coordination activities, which is a significant cost incurred by the person assisting (Barnes et al., 2008). Bachrach et al. (2006) argue that helpees who become too reliant upon the helping provider are dis-incentivized to complete their work. Lastly, Bolino et al. (2015) suggest that helping providers may develop feelings of frustration, stress, and burnout. Notwithstanding the downsides associated with providing helping to team members, we contend that helping behavior will reduce idle time, thus improving team performance, including customer satisfaction (Podsakoff et al., 1997).

H6.

The greater the amount of helping behavior, the greater the team's operational performance.

4. Study 1 – vignette-based experiment

We used an online vignette-based experiment to test Hypotheses 1–5 and simulation experiments to test Hypothesis 6. Vignette experiments are used in supply chain research (e.g., Cantor et al., 2014; Eckerd et al., 2013; Polyviou et al., 2018; Knemeyer and Naylor, 2011; Rungtusanatham et al., 2011) to investigate the make-or-buy decision (Mantel et al., 2006) and supplier switching intentions, (Mir et al., 2017), for example. A vignette experiment provides research participants with contextual information so that the subjects can make an informed judgment or decision (Eckerd et al., 2016). This research method enables scholars to control internal validity threats (Eckerd et al., 2016; Mir et al., 2017; Reimann et al., 2017) because research participants are not asked to provide sensitive information (Rungtusanatham et al., 2011).

The empirical setting of our vignettes is the restaurant industry. As noted earlier, this setting consists of core (e.g., restaurant managers) and non-core workers (e.g., servers and related wait-staff), thus enabling us to effectively study the dynamic interactions of workers (see Table 2: Keller and Ozment, 1999; Cantor et al., 2011). Second, restaurant workers regularly manage inventory (i.e., manage food and beverage services for retail customers), and hence these inventory management decisions have consequence on operational metrics such as customer wait times, out-of-stock rates, and customer service levels (see Table 2: Schultz et al., 1998, 1999). Third, the restaurant industry is experiencing significant employee turnover, thus affecting the nature of team dynamics, team productivity, and customer service levels (see Table 2: Griffith et al., 2006; Cantor et al., 2011) [1].

4.1 Experimental design

To test the hypotheses, we conducted a 2×2×2×2 between-subjects experiment in which we asked the study participants to assume that they are a restaurant server at the fictitious restaurant. In this role, the participant represents the non-core worker as described in the study's hypotheses. Each participant was randomly assigned to one of 16 vignettes, each of which represented a different combination of levels of the four factors described in Hypotheses 1–4: fairness of the core server's promotion (fair/unfair), work schedule assignment of the non-core server (preferred/non-preferred), previous core server reciprocity (assisted/did not assist), and non-core server table occupancy (low/high). To test Hypothesis 5, we also asked each participant behavioral questions that assess a person's level of collectivism, using the IND-COL scaled developed by Cozma (2011) (See Appendix C in e-Companion document).

In the role of a non-core “deciding server,” the research participant was asked: “Given your role on the team, which action would you take, given that the core server has a lot of work to complete.” The participant could either provide help to the core server or not provide help. If the participant agrees to provide help, a follow-up question was asked: “What percent of your time are you willing to allocate to helping the core server?”. We investigated the direct effect of the experimental factors on helping behavior and percent helping time.

4.2 Sample

We distributed our vignette-based experiment task to Amazon Master MTurk participants located in the United States and who are at least 18 years old. Master MTurks are people who have demonstrated excellence in their performance on a wide range of tasks (Sheehan and Pittman, 2016). Each Mturk participant received a small monetary incentive upon completion of the experiment. To ensure that participants read the background and related information, our vignettes contained multiple attention-checking questions (Abbey and Meloy, 2017). We also evaluated the realism of our scenarios using the three-item scale developed by Dabholkar (1994). A sample item is: “The situation described above does occur in the real world.” These items were measured on from 1 (strongly disagree) to 7 (strongly agree).

We received 702 completed surveys. We excluded 71 responses because the participants indicated that they did not have any work experience in the restaurant industry. We then removed observations where participants missed any of the six attention checking questions. Our final dataset consisted of 548 observations (53% were female; 65% of the respondents were between 22 and 40 years of age, and 57% reported having at least an undergraduate degree). Sixty-five percent of the participants reported three years of restaurant-related work experience; 52.7% were kitchen staff, 74.2% were wait staff, 51.2% bussed tables, and 21.3% worked as a restaurant manager. Eighty-seven percent worked in a sit-down restaurant; the remainder worked at a fast-food restaurant. The realism checks yielded a mean score of 6.27 (standard deviation of 0.82). Based upon the results of similar studies, these results indicate a high level of realism as perceived by our participants (Reimann et al., 2017). The average number of participants assigned to each treatment condition (16 treatment conditions) was 34.5 with a standard deviation of 3.02 (min = 29 and max = 39).

4.3 Measures

Based on our between-subjects experimental design, we created four categorical independent variables: Fair promotion of the core server (Fair promotion), fairness of the deciding server's work schedule (Preferred work schedule), the core server's prior helping behavior (Reciprocity), the deciding server's workload (Table occupancy), and the deciding server's level of collectivism (Collectivism). Fair promotion was coded as a “1” if the vignette described that a fair process was followed to promote another restaurant worker to the core server role; this factor was coded with a value of “0” if the promotion process occurred unfairly. Preferred work schedule was code with a value of “1” if the research participant was treated fairly concerning her or his preferred work schedule, “0” otherwise. Reciprocity was coded with a “1” if the recently promoted worker helped (assisted) the research participant in the past; this factor was coded with a value of “0” if the recently promoted worker did not help the research participant. Table occupancy was coded with a value of “1” if the research participant's tables are at 100% capacity of her or his assigned tables; table occupancy is coded with a value of “0” if the research participant's tables are at 25% capacity of her or his assigned tables. Lastly, our model also controls for an individual's predisposition to act collectively. The research participant's collectivism level was coded with a value between “1” to “7”. The collectivism variable was operationalized as an average across all eight survey items (see e-Companion document) [2].

Our two dependent variables are the research participant's decision to help and the percent helping time. Following Cantor and Jin (2019), we coded the decision to help with a value of “1” if the participant provides help, zero otherwise. Percent helping time represents the percentage of the deciding server's total time that he/she is willing to allocate for helping the core server. If the participant decided to provide help, an additional question was asked: “On a scale of 0–100%, how much of your total time would you allocate to helping Alex (core server)?” yielding a value between 0 and 100.

4.4 Empirical study results

Statistical tests were performed using SAS 9.4 software. Because the first dependent variable (i.e., the decision to help or not help) is binary, logistic regression was used (Hosmer et al., 2013). The results of this analysis are shown in Table 4 (Model 1). Of the 548 total observations, 419 participants (76%) decided to provide help to the core server. According to the results from Model 1, the log of the odds of a non-core server providing help to the core server did not have a statistically significant relationship with fair promotion (p > 0.05), but it was positively related to the preferred work schedule (p < 0.05), reciprocity (p < 0.05), and collectivism (p < 0.05), whereas table occupancy (p < 0.05) was negatively related since higher table occupancy was coded as one and lower table occupancy was coded as zero. Thus Hypothesis 1 (fair promotion) was not statistically supported, while Hypotheses 2, 3, 4, and 5 (preferred work schedule, reciprocity, table occupancy, and collectivism, respectively) were supported. The reciprocity factor had a particularly strong effect. Given the same values of fair promotion, preferred work schedule, table occupancy, and collectivism, the odds of a non-core worker helping the core worker were 19.119 times greater if help had been provided to the non-core worker in the past.

The second dependent variable in the study is percent helping time. A mean value of 22.90% was derived for percent helping time with a standard deviation of 20.18. Since the output variable has a lower bound of 0 and an upper bound of 100, Tobit regression was used (McDonald and Moffitt, 1980). The results are given in Table 4 (Model 2). Preferred work schedule, reciprocity, table occupancy, and collectivism were statistically significant (p < 0.05) [3].

5. Study 2 – simulation model and analysis

5.1 Model overview

A limitation of the vignette-based experiment in Study 1 is that we do not observe dynamic operational behaviors that may result from helping behavior. Chandrasekaran et al. (2018) emphasize the inherent limitations of using only cross-sectional research methods to analyze complex supply chain problems. To address these limitations, they recommend using simulation models to complement and strengthen an empirical study. Therefore, to test Hypothesis 6, we created a simulation model using AnyLogic 8.0 that examines how helping behavior influences operational performance in a dynamic setting.

To represent both human behavior (i.e., the decision to help) and operational behaviors (i.e., the flow of customers through a restaurant), our model uses a hybrid combination of agent-based modeling (ABM) and discrete event simulation (DES). Majid et al. (2016) advocate using hybridized ABM-DES models to capture the influence of human behavior in service systems, demonstrating that DES is better-suited to modeling queuing customers, while ABM is more appropriate for modeling human behavior. Accordingly, our study combines ABM and DES to simulate a stylized service model of a restaurant. While DES has been used previously to model the flow of customers in a restaurant setting (Vries et al., 2018), in our study the servers are modeled as autonomous agents capable of decision making and working and interacting as a team. Modeling the servers as agents also allows us to realistically represent them as complex social beings that have frequent interactions with their environment and each other. Complex interactions can be conceptualized as social processes, such as helping behavior or social influence (Badham et al., 2018). Figure 2 reflects our DES model.

The model environment represents a restaurant with 14 tables, each of which can seat one to four customers. Customer entities arrive throughout the day and are seated as tables become available. Each customer entity that arrives at the restaurant represents a group of one to four customers. The model also contains three server agents: the core server, the deciding server, and another non-core server. The core server is assigned six tables, and the deciding server and non-core server are assigned four tables each. The server agents act as resource units to the customer entities, assisting them in filling their orders. It is assumed that the server agents work continuously throughout the day. Apart from serving duties, the deciding server periodically makes decisions about whether to provide help to the core server [4]. The regression outputs from Study 1 (Table 4) were used to create multi-attribute functions that inform the deciding server's decision logic. The values of five decision variables are assigned, according to the experimental scenario that is being tested: whether the core server was fairly promoted (FP; binary), whether the deciding server was given his or her preferred work schedule (PWS; binary), whether the core server had helped the deciding server in the past, thereby triggering reciprocity (R; binary), maximum table occupancy (TO; discrete values ranging from 1 to 4), and the level of collectivism that characterizes the deciding server (C; discrete values ranging from 1 to 7). Each simulation run represents a single 13-h day of a dine-in restaurant's operations, starting at 9:00 AM and ending at 10:00 PM. The model time unit is minutes [5].

5.2 Simulation experimentation

To evaluate the impact of helping behavior on the operational metrics listed in Table 5, the hybrid simulation model was run for two different scenarios: one in which the deciding server is likely to provide help (with fair promotion, preferred work schedule, and reciprocity factors set to a value of 1 and collectivism set to a value of 7), and one in which the deciding server is unlikely to provide help (with collectivism factor set to a value of 1 and all other factors set to zero). To study the effects of table occupancy on operational metrics, both the helping and non-helping scenarios were further tested for two-factor settings: one in which the deciding server is responsible for four tables in its area (such that it is often busy with its own customers), and one in which it is responsible for only one table (such that it is often idle). The idea is that when the deciding server has fewer tables of its own, it is more willing to provide help to the core server. Thus a total of four cases were studied (two scenarios and two factor-levels for the factor Table Occupancy). In all four cases, the core server is responsible for six tables.

5.3 Simulation study results

The two scenarios in which the deciding server is unlikely to provide help yielded zero help time at the end of all simulation runs. Figure 3 shows the distribution of total help time for the other two scenarios, in which the deciding server is likely to provide help, for 30 simulation replications. The scenario in which the deciding server is responsible for four tables yielded 34% less help time (mean: 97.5 min, SD: 65.4) than when the deciding server was responsible for one table (mean: 163.4 min, SD: 85.0) (see Figure 4).

To observe the impact of helping behavior on the operational performance metrics listed in Table 5, the values of these metrics were captured at the end of each simulation run for all four experimental scenarios and then averaged over 30 replications. The means were compared using two-tailed t-tests with alpha equal to 0.05; the results are shown in Table 6. Cohen's d values were also calculated to determine the effect size of the differences in means, where d = 0.2 is considered a “small” effect size, 0.5 represents a “medium” effect size, and 0.8 a “large” effect size. This means that if the two means do not differ by 0.2 standard deviations or more, the difference is trivial, even if it is statistically significant (Cohen, 2013).

The results indicate modest improvements in nearly all operational metrics when helping behavior was present. In particular, the wait times at the core server's tables experienced a 19.2% reduction due to helping when the deciding server has four tables and a 22.4% reduction when the deciding server has one table. However, it should be noted that helping behavior increased the wait times at the deciding server's tables by 22.2% when the deciding server has four tables. The increase in wait times at the deciding server's tables suggests a downside of helping, as the deciding server's time is shared with the core server's tables. However, helping behavior reduced the average total wait time across all tables by 6.1% when the deciding server has four tables and 22.4% when the deciding server has only one table.

6. Discussion and conclusion

The purpose of this research study was to examine how a non-core team member's perceptions of procedural justice regarding their role, team dynamics, and organizational policies impact their motivation to engage in helping behavior, as well as the overall impact on operational performance (e.g., customer service time). In doing so, we build upon prior supply chain research by capitalizing upon the procedural justice and strategic core literature (e.g., Hofer et al., 2012; Griffis et al., 2012; Cantor et al., 2011; Summers et al., 2012). This is the first study in the supply chain literature that has juxtaposed the procedural justice literature with the strategic core theory to examine the dynamics of team member roles, workload, and career experiences and the subsequent impact on logistics performance. We also introduce to the supply chain literature a new way of conceptualizing the roles that logistics workers occupy by differentiating core and non-core workers, the ways in which different types of workers interact with one another, and the resulting impact on customer service and overall productivity. Because these workers have different career experiences, they will have unique perceptions of procedural fairness in the workplace. Based on this assumption, we were motivated to study the dynamic interactions of team members in a logistics retail setting (i.e., a new application of these theories). Reconceptualizing logistics in this way allows us to study how theories and practices from supply chain management and organizational behavior can translate to a service (restaurant) setting.

We use these complementary theories to conceptualize how and why the non-core team member would react to a simulated situation using an online vignette experiment and a hybrid discrete-event/agent-based simulation. To test our non-core team member hypotheses, we demonstrate the utility of leveraging the advantages of both analytical and numerical methods of analysis. Put simply, the hybrid simulation method enables us to investigate the causal assumptions initially explored in the vignette experiments in a dynamic setting. Our use of hybrid simulation modeling provides us with the opportunity to explore and test the strategic core and procedural justice theories about how customer throughput and wait times could be influenced when workers choose to help their team members. Through this integrated modeling approach, we examine the effects of helping behavior on several system performance metrics in the dynamic setting of a virtual restaurant. Importantly, we demonstrate the value that these theoretical constructs bring to logistics (i.e., helping behavior in production/distribution teams, throughput, and customer wait times).

We now discuss the theoretical and managerial implications of our findings, which are examined in the context of our six hypotheses.

6.1 Study findings and theoretical implications

In Hypothesis 1, we theorized that non-core workers are attentive to how core workers such as supervisors create and promulgate new organizational policies and practices (Lind and Tyler, 1988). When non-core workers observe that the formation and implementation of promotion policies are unjust, non-core workers could react negatively and reduce their amount of helping behavior. However, we did not find support for Hypothesis 1. To gain insight into the reasons for this unexpected finding, we reviewed the qualitative data that we collected from our online experiment. We examined responses to the prompt “Please explain your reasoning behind your decision to help” from participants who decided to help the core server in spite of the unfair promotion. Several participants pointed out that the core server (Alex) was not to blame for his unfair promotion, as exemplified by this comment: “Even though I felt Alex received his promotion unfairly, it may not have been all Alex's fault. He may have been presented with the opportunity and given the option to take the position or quit/be fired.” Moreover, many participants noted that unfair organizational policies were not the customers' fault, either, and they should not have to suffer from poor service as a consequence. These participants indicated that providing help to ensure a positive customer experience was an implicit expectation of restaurant workers. Indeed, one participant commented: “No matter how I feel about the promotion, waiters always help out other waiters. The reputation of the restaurant depends on excellent service to maintain customers.” Some went even further, stating that they decided to help because that is what a “good person” or “team player” should do, regardless of the context. Finally, some participants were motivated by personal gain, indicating their belief that helping the core server would be viewed positively by supervisors (i.e., it would demonstrate good organizational citizenship), and this could yield future professional benefits. As an example, one participant commented: “Despite the unfair promotion, I want to stay on the good side of management.” The variety of reasoning that emerged from the participants' responses suggests multiple potentially fruitful avenues for future research. For example, scholars could study how employees who have a desire to see themselves being promoted fairly may have a direct incentive to help to core workers.

By contrast, our study provided empirical support for Hypothesis 2, in which we theorized that the non-core worker's fairness perceptions and pro-social behavior are directly impacted by the assignment of work schedules. Previous research has found that logistics workers who believe that their work scheduling preferences are violated may exhibit counter-productive behaviors such as quitting their job (Cantor et al., 2011; Miller et al., 2013). Our results extend these findings, suggesting that when organizations disregard work schedule preferences, non-core workers could perceive that the organization is not committed to providing a rewarding work environment, causing them to engage in withdrawal behavior such as lower levels of helping. Additional support for this finding was observed in the research participants' responses to the prompt “How did your work schedule influence your decision to help or not to help?”. Many participants who were told that their work scheduling preferences had been routinely ignored indicated that this significantly influenced their decision to decline to provide help. Most viewed it as an issue of fairness, as exemplified in this comment: “I do not have a 40-h work week and yet I'm supposed to be on call 24/7. I'm just a widget.” Others made an explicit connection between work scheduling and the organization's regard for its non-core workers: “I've been in that situation, and it's the first indication that your employer has no respect for you as a human being.” The theoretical implication is that supply chain scholars should recognize that team composition and the resources provided to core and non-core workers (e.g., varying levels of work schedule flexibility) have important implications on logistics performance. Furthermore, while the lack of support for Hypothesis 1 implies that a non-core worker's helping behavior may not be significantly impacted by strategic human resource (HR) decisions such as the firm's promotion practices, the support we found for Hypothesis 2 indicates that the team leader's implementation of operational HR policies such as daily or weekly scheduling policies may have a direct impact.

We also found strong support for Hypothesis 3, in which we theorized that norms of reciprocity influence a non-core worker's decision to engage in helping behavior. As with Hypothesis 2, participants' explanations of why they did/did not help the core server made it clear that they viewed this as an issue of fairness, as exemplified by this typical response: “If I am taking on extra work but not getting the extra reward I am not going to be bothered with helping someone who has not helped me in the past.” Many participants emphasized that this expectation of quid pro quo is prevalent among workers in the service sector: “Server law is you get what you give. Alex helped me in the past, so I will help Alex. It is mutually beneficial to continue that relationship.” Indeed, these comments suggest that non-core workers may view reciprocity as not only a motivator for pro-social behavior, but also a requirement for good organizational citizenship.

We next theorized that when the non-core worker had greater resource capacity (i.e., they were not fully utilized), they would be more likely to provide help to the core server. We found support for this relationship (Hypothesis 4), which is consistent with previous literature indicating that demands on workers affect their decision to engage in helping behavior (Dimotakis et al., 2012). When asked about their decision to not help the core server, many participants who were told that all of their tables were occupied (i.e., they were fully utilized) indicated that they were unwilling to sacrifice their own productivity, their customers' satisfaction, and their ability to earn tips. Some responses implied that it would be unfair and/or impractical to expect them to do so. For example, one participant emphasized that helping the core server at the expense of their own customers would actually go against the organization's interests and would likely (and justifiably) result in termination: “I have a commitment, first and foremost, as assigned by the restaurant manager, to take care of my four tables ….I would not dare risk one of my customers having to wait unnecessarily for me to service them and possibly give me a bad review, or a bad tip for that matter, because I was trying to help a coworker … It's not worth me losing my job not following the restaurant manager's orders.” This finding is aligned with the support we found for Hypothesis 2 (i.e., undesirable work schedules lead to less helping): when insufficient labor capacity burdens non-core workers with extra work and/or inflexible schedules, they are likely to view the arrangement as unfair and are unlikely to extend themselves further to help the core worker. An important implication of this finding is that an organization will be more likely to benefit from employee helping behavior if it provides a supportive work environment, in which non-core workers do not feel that they are constantly being asked to shoulder the burden of the organization's limited resources.

We then hypothesized and found support about the role of a non-core worker's predisposition to act collectively (Hypothesis 5). There is management (e.g., Colquitt et al., 2002; Colquitt et al., 2005) and supply chain research that has begun to examine how collectivism, a personality trait, enhances individual and team performance (Cannon et al., 2010; Eckerd et al., 2016; Handley and Angst, 2015; Lee et al., 2018). While participants' responses in support of Hypotheses 2, 3, and 4 indicate that self-interest and perceived fairness are primary motivators for a non-core worker's decision to help, the relationship we found between collectivism and helping suggests that there are other reasons. For example, one participant justified helping the core server by pointing to service-sector workers' common objective of collectively providing high-quality customer service, regardless of the worker's level of career experience: “We are all a team … working a restaurant is a group effort and everything needs to run smoothly, from the tables being bussed, to the kitchen staff putting out good food on time, and the wait staff doing a good job taking care of the customers … everyone needs to pitch in.” Others mentioned that working collectively, rather than competitively, with co-workers made the work environment more pleasant and less stressful. Several participants who decided to provide help, despite their ambivalence toward the core worker and/or the organization due to perceived unfairness, explained that they were concerned about the collective welfare of the organization and its employees: “I would decide to help Alex because … customers will view us as a whole and not always as individuals. It is each team member's responsibility to provide great service to keep returning customers. This helps everyone's pay, not just our own. It keeps our restaurant running for the future and helps with job security as well.” While our study increases our understanding of how a person's collectivism values influence helping behavior in a team-based retail logistics setting, future research could investigate the tension between a non-core worker's self-interest and their desire to support their team and how this influences their decisions to engage in pro-social behavior.

Our final hypothesis evaluates how helping behavior can improve operational performance. We expected that helping behavior would yield improved performance, and indeed, our simulation results appear to support Hypothesis 6: helping behavior significantly increased overall customer throughput and decreased overall average customer wait time, with very large effects sizes (i.e., Cohen's d values = 1.2 and 21.2, respectively) when the deciding server agent has a reduced workload in its own service zone (i.e., one table). This finding suggests that managers should encourage helping behavior to improve operational performance. At the same time, however, our simulation results indicate that individual-level performance suffered with helping behavior: average customer wait time at tables in the deciding server agent's own service zone increased. Thus improvements to overall system/team-level performance do not provide a complete picture of the impact of helping behavior; in fact, relying upon these metrics alone may be misleading, especially in terms of long run operational performance. While the simulation model did not include a feedback loop connecting service performance with customer behavior, in real-world restaurants, customers experiencing longer wait times might offer lower tips to their server, thereby disincentivizing future helping behavior by that server. Disappointed customers might also be less likely to return to the restaurant in the future.

According to the simulation results for this specific system, the team-level benefits of helping behavior by a non-core server would likely outweigh the negative impact on the individual server's performance. The average wait times in the deciding server's zone are relatively low in all cases (0.2 min/2.7 min without helping for low/high table occupancy; 0.7 min/3.3 min with helping for low/high table occupancy), compared with the core server's wait times (7.8 min/6.3 min without helping for low/high table occupancy; 8.3 min/5.8 min with helping for low/high table occupancy). Thus, the small absolute increase in wait time for customers in the deciding server's service zone due to helping might not have a noticeable adverse impact on customer experience. However, this finding would not necessarily hold true for different system or parameter settings. Clearly, if helping behavior detracts substantially from individual-level worker performance, managers should not support it, even if it yields improved overall system/team-level performance in the short term.

6.2 Managerial implications

The services provided by logistics and supply chain workers in the retail sector can have a significant impact on a firm's customer experience (e.g., Swink et al., 2022). However, improving the performance of logistics workers remains a serious challenge because several sectors of the economy are either experiencing severe worker shortages or significant employee turnover, thus affecting the nature of team dynamics, team productivity, and customer service levels (see Table 2: Griffith et al., 2006; Cantor et al., 2011). The results of our study offer practical implications for supply chain managers in other logistics contexts. In particular, the results suggest that managers should study helping behavior in more traditional logistics settings because our study offers initial evidence that helping can lead to improved overall operational performance (e.g., reduced customer wait times). The implication to supply chain management and operations practice is that agile work structures can significantly improve operational performance and reduce loafing behaviors. To encourage helping, managers should be mindful of perceptions of fairness in incentivizing individual and team performance. In particular, managers should collect feedback concerning the formation and implementation of employee promotion and work scheduling policies. If these policies are viewed as unfair, non-core workers could react negatively and reduce their amount of helping behavior. Managers should also pay attention to the capacity of their team members. If a team member is already fully occupied and provides help to another team member, it may come at a cost to their own individual performance. Thus, managers should consider both individual and overall team performance when evaluating team members for incentives such as bonuses. Furthermore, the results of this research indicate that reciprocity is a strong motivator for helping, and it may mitigate feelings of ill-will or unfairness that can emerge over time when one team member is frequently providing help. To encourage reciprocity, managers might consider rotating team members to ensure that it is not the same individual who is in the position to provide help all the time. Finally, managers can also consider hiring and placing individuals with collectivist tendencies on teams in which roles, responsibilities, and workload are likely to be dynamic.

Figures

Figure 1

Theoretical model

Figure 2

DES flowchart

Figure 3

Total help time for different number of tables assigned to the deciding server

Figure 4

Deciding server agent decision flowchart

Table 1

Examples of core and non-core workers

Setting	Core worker (supervisor)	Non-core worker	Examples of non-core helper providing help
Warehousing	Assembly-line workers	Material Replenishers (MR)	MR restocks inventory at assembly line
Transportation	Less-than Truck Load Driver	Driver Helper (DH)	DH loads/unloads trucks
Retail	Restaurant Manager	Wait Staff	Wait Staff can help other waiters provide meals, clean tables to guest
Manufacturing	Senior Buyer	Buyer	Buyer provides help to senior buyer with negotiating contract terms with potential suppliers
Health Care	Doctors	Nurses	Nurses provide help to physicians with screening patients, providing routine and specialized care, and so on

Source(s): Table 1 by authors

Table 2

Importance of theoretical factors present in restaurant context and logistics literature

Theoretical factor	This study (restaurant setting)	Examples from logistics/Operations management literature	Common themes in logistics, OM, and restaurant operations	Practical importance of theoretical factor
System Performance (Throughput Time)	Average, Customer Throughput Rate; Average Customer Wait Time	Helping behavior influences throughput time (Cantor and Jin, 2019). Work-sharing fixed assignment systems influence throughput times (Doerr et al., 2002, 2004). Physical arrangement of workers impacts throughput time (Bartholdi and Eisenstein, 1996)	It is critical that both fields improve service performance. Downstream customers do not want to experience significant wait times	Optimizing throughput time can have serious repercussions on customer satisfaction and retention. Longer throughput time may result in highly impatient customers potentially resulting in lost sales
Helping Behavior	Non-core server decision to help core server	Emotions of logistics workers influence their intention to help others in the workplace (Keller et al., 2020). Stability of production orders and material handling complexity influences helping behavior (Cantor et al., 2019). Supplier helping behavior influences retailer helping behavior (Esper et al., 2015). Focal firms attempt to provide volitional help to their suppliers (Autry and Griffis, 2008). Managers can incentivize workers in helping behaviors (Siemsen et al., 2007). Helping behavior influences group performance based on level of task interdependence (Bachrach et al., 2006)	Both fields have some workers that have the capacity to complete tasks faster than others. Faster workers can reduce wait time caused by slower workers	Worker motivation continues to be a persistent problem in many organizations. Managers are interested in understanding how to motivate less productive (slower) workers by providing nonfinancial incentives (organizational support) to stronger workers to help their coworkers. Increased helping behavior can have direct implications to both logistics and retail performance
Fair Treatment of Workers	Fair process was followed to promote another restaurant worker to the core server role	The procedural fairness literature has been adopted in logistics to study how the treatment of truck drivers could influence their decision to leave the industry (e.g., Cantor et al., 2011). Hofer et al. (2012) use the procedural justice literature to study how justice perceptions influence logistics outsourcing decisions. Griffith et al. (2006) find that perceived justice has a positive effect on logistics performance	It is critical to understand factors that influence procedural justice concerns in both settings. Otherwise, workers will become less committed to their organization and have exit (turnover) intentions	In both logistics and retail settings, core and non-core workers interact with the end customer. Thus, workers who may engage in counterproductive behaviors towards customers if their fairness concerns are not attended
Work Scheduling	Non-core server was treated fairly (or unfairly) concerning her or his preferred work schedule	Work scheduling is a fundamental principle studied in many production, distribution, and transportation problems (e.g., Miller and De Matta, 2003). Employer scheduling practices can influence employee fatigue (Crum and Morrow, 2002). Work scheduling can influence the availability of capacity for logistics and transportation services, for example	Both fields have machines and people that need to be scheduled to meet downstream customer demand. Restaurants and more broadly logistics operations have production and people constraints that must be scheduled so that the firm can provide high levels of customer service	Many logistics and retail workers operate on highly structured schedules. Disruptions to these schedules can have a direct bearing on the performance of the systems that the workers operate. It is thus of paramount importance to minimize worker fairness concerns about work scheduling to reduce worker absenteeism and potential turnover issues
Norms of Reciprocity	Did the recently promoted core server help (assist) the non-core server in the past?	Employees will engage in norms of reciprocity when the organization provides organizational and supervisory support (Cantor et al., 2012). Supply chain employees will act altruistically to each other so long as customer managers perceive that supplier personnel are engaging in helping behaviors toward them. Employees will then likely reciprocate by engaging in customer helping behaviors (Esper et al., 2015). Norms of reciprocity has received some attention at different levels of an organizations including line-level employees such as truck drivers in the logistics and operations literature	Workers in both fields are likely to positively engage in helping behavior so long as their colleagues have previously provided them with help	Team work is very much present in logistics and retail settings. Managers desire practices that can encourage collaboration among team members
Workload	The non-core server's assigned tables are either at 100% capacity or at 25% capacity	Inventory levels impact work group performance (Schultz et al., 1998, 1999). When an employee's workload increases because team members are absent, his/her sense of unfairness may lead to calling in sick more often (ten Brummelhuis et al., 2016)	Managers in both fields seek to optimize workload allocation to their employees to maximize performance outcomes	Managers in logistics and retail sectors seek to optimize full utilization of machines, systems, and service capacity. However, some operators of these systems may need the help of others who work at faster rates
Collectivism (Attitude to Teamwork More Broadly)	Surveyed perceptual measures on participant's proclivity to working as a team	SCM scholars are studying how individual differences affect operational performance (e.g., Bendoly et al., 2010; Boudreau et al., 2003; Moritz et al., 2013; Moritz et al., 2014)	Managers across disciplines and industries struggle to leverage nonfinancial incentives to motivate their employees to engage in team-level behavior	Human resource (HR) professionals commonly use personality surveys to assess the potential employee's fit with the organization. Since there is significant turnover in logistics and retail settings, HR managers could find these surveys valuable to efficiently identify talent
Structure of Work	Service-sharing zones influence server helping behavior and productivity	Work-sharing and fixed assignment systems impact performance (Schultz et al., 1998, 1999; Doerr et al., 2002, 2004)	Both restaurants and logistics operations are organized in work zones	Logistics and restaurant managers often evaluate how the structure of work is conducted with a particular interest in optimizing the design of the systems
Core and Non-Core Workers	The core server has increased workload (i.e., responsible for more tables) compared with the non-core server	These dyadic relationships are present in transportation (drivers and driver helpers), warehousing (assemblers and material replenishers), health care (doctors and nurses)	There is a core/non-core worker dynamic present in both settings that can have a direct bearing on customer service outcomes	Many logistics and retail settings have different types of workers including core and non-core roles

Source(s): Table 2 by authors

Table 3

Relevant literature

Table 4

Empirical study results

	Model 1	Model 2
	Logistic regression model	Tobit regression model
Dependent variable	Helping behavior	Percent helping
	Estimate (Std. Err)	Estimate (Std. Err)
Constant	−2.99*(0.69)	−12.57**(5.63)
Fair promotion	0.33(0.2)	1.16(1.93)
Preferred work schedule	0.84*(0.25)	5.98*(1.94)
Reciprocity	2.95*(0.32)	21.56*(1.99)
Table occupancy	−0.57**(0.24)	−12.37*(1.93)
Collectivism	0.57*(0.12)	4.41*(0.99)
_Sigma		21.72*(0.79)

Note(s): *p < 0.01, **p < 0.05, ***p < 0.1

Source(s): Table 4 by authors

Table 5

Operational performance metrics

Metric	Description
Total help time	Total time help was provided by deciding server to core server
Successful customers	Total number of customers receiving service
Wait time at core server's tables	The average time a customer seated in the core server's area waits on the core server
Wait time at deciding server's tables	The average time a customer seated in deciding server's area waits on the deciding server
Total wait time across all tables	The average time a customer waits on any server

Source(s): Table 5 by authors

Table 6

T-test and Cohen D results for Study 2

	Deciding server has four tables							Deciding server has one table
Output metric	No helping behavior			Helping behavior			D	No helping behavior			Helping behavior			d
	Mean	SD	N	Mean	SD	N		Mean	SD	N	Mean	SD	N
Successful customers	80.7	2.0	210	81.2	1.8	510	0.28*	64.2	1.7	180	66.1	1.9	540	1.2*
Average wait time on core server's tables	7.8*	0.8	500	6.3	0.8	500	1.72*	8.3	0.7	500	5.8	0.6	500	3.83*
Average wait time on deciding server's tables	2.7*	0.4	500	3.3	0.4	500	1.24*	0.2	0	500	0.7	0.2	500	8.72*
Average total wait time across all tables	4.9*	0.4	500	4.6	0.4	500	0.81*	5.8	0.2	500	4.5	0.3	500	21.2*

Note(s): *p < 0.01

Source(s): Table 6 by authors

Notes

1. Detailed information about the vignettes can be found in the e-companion.

2. Following a reviewer's suggestion, for robustness purposes, we also operationalize the collectivism variable using factor analysis (see Calantone et al., 2017 and McNeish and Wolf, 2020). All factor scores were above 0.65 or greater. Our regression results using the factor score measure for collectivism are consistent with the original results.

3. Following a reviewer's advice, we verified our results by estimating models 1 and 2 using several control variables such as gender, participant age, education, work experience, and number of restaurants previously employed at. Except for age of participant, all control variables are not statistically significant. The results are consistent with the original models.

4. Details about the framework used to design the simulation can be found in the e-companion (Appendix E).

5. A description of the discrete event simulation (DES) can be found in the e-companion (Appendix F).