1 Introduction

George buys a 75″ television and gathers friends for a barbecue to watch the Super Bowl. The morning after the game, he returns the product to the store for a refund. Olivia visits a store to purchase a coat. Among several coats priced at $200, she discovers one that bears a $20 price tag. Despite recognizing that this is a mistake made by the company, she approaches the cashier, pays the $20, and leaves the store. Within service-related situations, dysfunctional behaviors such as wardrobing, fraud, verbal abuse, shoplifting, sexual harassment, and false complaints are unfortunately common. In the United States, 13.7% of all merchandise returns involve fraud, which results in overall losses of $107 billion for retailers (NRF 2023). Data from the United Kingdom suggest that, over 1 year, 92% of frontline employees (FLE) were verbally abused, 70% faced customer threats, and 14% experienced physical assault (USDAW 2021). Even in digital sales, the situation is not much better: a global report revealed a 64% increase in global payment fraud in 2022 concerning B2C digital goods and services (SIFT 2023). Consumer misbehavior is therefore widespread across various industries and countries, posing a challenge for companies to address.

Dysfunctional customer behavior (also addressed interchangeably in this paper as DCB), commonly referred to as customer dishonesty or customer incivility, gives rise to various consequences, including financial losses for organizations (Wan et al. 2021; Zhao et al. 2020) and increased stress and emotional exhaustion among employees (Lages et al. 2023). The severity of dysfunctional behavior ranges from minor instances, such as unjustified complaints or failure to report an error in a bill, to more serious or illegal activities like shoplifting and payment fraud (Garnefeld et al. 2019; Lages et al. 2023; Wilson et al. 2021). Dysfunctional behaviors also vary according to their nature, being opportunistic (such as Olivia’s case, who notices the retailer’s mistake and seizes the opportunity) or premeditated (like George’s case, who planned the purchase to return the product).

For a long time, literature has sought to identify the motivators behind consumers’ dysfunctional behaviors, and various drivers have been identified (see Lages et al. 2023 and Fisk et al. 2010). However, little emphasis has been given to a specific cause: customer dissatisfaction resulting from a service failure (Wirtz and McColl Kenedy 2010). Service-related problems, for example, can stimulate dysfunctional customer behavior as a means of punishing the company (Gong and Wang 2021).

Researchers and managers aim to find mechanisms capable of inhibiting such undesirable behaviors and their negative consequences, with options ranging from employee training and supervisor support (Xiao et al. 2022) to more stringent consumer controls, such as surveillance (Pieper and Woisetschlager 2024). However, studies using interpersonal relationships as inhibitors of DCB are scarce. The level of psychological proximity between the customer and the service provider is a variable that represents an interpersonal relationship and could potentially mitigate intentions towards engaging in dysfunctional behavior (Wan et al. 2021). Construal-level theory posits that the perceived closeness or distance of an object, situation, or person is contingent upon changes in the level of interpretation (Eyal and Liberman 2012). Consequently, individuals can construct mental representations about things that are psychologically close to or distant from them (Trope et al. 2007). Despite the extensive examination of construal-level theory in consumption decision-making and recommendation situations (Herter et al. 2021; Tan 2018), its application to the assessment of the acceptability of dysfunctional behaviors remains relatively unexplored (Pinto et al. 2020).

Therefore, this research aims to address the following research questions: what is the impact of social distance between customers and service providers on the acceptability of DCB in service failure contexts? What factors influence the perceived acceptability of dysfunctional behavior? To answer these questions, we focused on understanding service failures and consumer dissatisfaction as triggers for DCB acceptability, as well as on how to minimize the occurrence and negative consequences of such behavior.

The construal-level theory serves as an appropriate theoretical framework for this research because different construals can influence human behaviors, judgments, and preferences (Eyal and Liberman 2012). We argue that social distance has a direct effect on the acceptability of DCB, especially because it triggers feelings of trust and anticipated guilt. A psychologically close relationship between a customer and a brand requires commitment and typically involves trust, which can foster mutual support between client and employee, even if the service is deficient (Kim et al. 2019). Trust engenders more positive customer behaviors towards service providers (Dang et al. 2020) and, as such, can alleviate the consequences of unsatisfactory experiences, such as the propensity to accept DCB. Similarly, emotions like guilt have the power to reduce the likelihood of customers engaging in unethical behaviors (Arli et al. 2016).

Furthermore, we consider two boundary conditions capable of moderating the effects of social distance on DCB: the controllability of the failure and the nature of the behavior. When failures happen, consumers tend to attribute blame to some factor (human or not), and attribution theory is commonly used to explain consumers’ responses in these cases (Weiner 2014). While there are studies examining the consequences of attributing service failures to other customers (He et al. 2019), there is limited literature exploring the causal attribution of failures to employees, particularly considering the social distance and the potential for customers to engage in dysfunctional behavior. Moreover, the nature of the behavior can influence its perceived acceptability: opportunistic behavior, for instance, might be mistakenly considered less culpable and punishable compared to premeditated behavior, which means that if the misbehavior stems from passively benefiting from an opportunity, it is perceived as less unethical (Vitell 2003).

This research adopted an experimental approach to the phenomenon and its results add to the existing literature in three ways. First, it helps researchers and practitioners to better understand in which situations consumers perceive misbehavior to be acceptable and how service failures and psychological distance contribute to these responses. Second, this research sheds some light on how emotions influence consumers’ perceptions of dysfunctional behaviors. Third, the findings provide insight into the influence that the nature of such behaviors has on consumers’ perceptions.

2 Literature review and hypothesis

2.1 Dysfunctional customer behavior and construal-level theory

Dysfunctional behaviors refer to actions by customers that overtly or covertly disrupt otherwise functional service encounters (Harris and Daunt 2013). Such behavior violates generally accepted norms of conduct in these situations (Fisk et al. 2010). The literature portrays several customers’ behaviors as dysfunctional, including theft, fraud, vandalism, verbal or physical aggression toward employees, price tag manipulation, queue jumping, and many others, both legal and illegal (Wan et al. 2021). These behaviors not only result in direct financial losses for companies but also generate indirect costs, such as increasing workload for the team when addressing dysfunctional behaviors (Garnefeld et al. 2019; Harris and Daunt 2013; Kang and Gong 2019) or influencing other customers to behave similarly, triggering a “contagion” (Danatzis and Möller-Herm 2023), or a “domino effect” (Telli et al. 2020).

In consumption contexts, dysfunctional behaviors (or DCBs) are closely linked to customer moral perceptions (Pieper and Woisetschlager 2024). When an individual evaluates a behavior, one of the considered variables is whether it violates the generally accepted norms of conduct in consumption situations. Norm violation may be related to causing problems for employees or other consumers, or even damaging the company (He et al. 2019; Reynolds and Harris 2009). When a behavior is perceived to violate such social norms, it is usually seen as less acceptable. It is important to understand that an individual’s attitude toward potentially unethical behavior is a range of acceptable positions rather than a single point (Fullerton et al. 1996), which means that a behavior is not completely acceptable or unacceptable; one individual may perceive a certain behavior as more acceptable than the others. Consumers may assess how tolerable they perceive their behavior to be or evaluate a third party’s behavior. The focus of this research is on understanding how consumers perceive their own dysfunctional behavior as acceptable.

There are elements that can influence how much a consumer perceives the idea of misbehaving more or less acceptable (Wan et al. 2021), and social proximity is one of these variables. Previous studies have shown that when there is proximity between the parties, individuals are less likely to behave in a dysfunctional manner (Wan et al. 2021). Likewise, when services do not live up to consumers’ expectations or when individuals perceive a violation of the contract made with the company, they are more prone to adopt dysfunctional behaviors (Wirtz and McColl Kenedy 2010; Gong and Wang 2021).

Construal-level theory, originating from social psychology, posits a connection between psychological distance to objects, events, or individuals and their mental representation (Trope and Liberman 2010). Psychological distance encompasses subjective feelings of being distant from the present (immediate) experience and the ego (Herter et al. 2021). The notion of distance can be understood in terms of spatial, temporal, social, or hypothetical dimensions (Adler and Sarstedt 2021; Eyal and Liberman 2012; Tan 2018). Human cognition involves the construction of high- or low-level mental representations. High-level constructs are abstract representations that focus on the general aspects of an object, event, or individual, while low-level constructs are concrete and detail-oriented representations. Construal-level theory suggests that information is processed at higher levels when it is related to psychologically distant events, as opposed to psychologically close events (Trope et al. 2007; Sordi et al. 2022).

Therefore, the psychological distance from a target influences the level at which individuals construct events and situations (Pinto et al. 2020; Yan et al. 2016). Social proximity, for example, leads individuals to perceive others as having similar opinions, enabling the construction of low-level mental representations that provide a concrete and detailed understanding of the other person’s feelings.

This study adopts the concept of psychological distance and compares failures caused by socially close employees to those caused by socially distant employees, assuming that a customer’s perception of failure can vary based on psychological distance. When the victim is identifiable (known) and perceived to be closer, the offender is more likely to engage in behaviors aimed at reducing harm to the victim (Dootson et al. 2016).

Since previous literature does not relate social distance with DCB, we aim to bridge this gap. We propose that when a failure is attributed to a psychologically distant employee, it is described in relatively more abstract terms compared to when the failure is committed by a socially close employee, whose mental representation is more concrete (Trope et al. 2007). In a situation of social proximity, customers would perceive dysfunctional behaviors that harm the service provider as less acceptable. Therefore, the following hypothesis is proposed:

H1

The perceived social distance between the customer and the service provider influences the acceptability of adopting a dysfunctional behavior, such that a greater (smaller) perceived social distance leads to a higher (lower) acceptability of dysfunctional behavior.

2.2 Trust

One theory that can potentially offer important support in understanding the effects of psychological distance on the acceptability of DCB is trust. Trust is the expectation consumers have that the service provider is dependable and that it will deliver what has been agreed between the parties (Sirdershmukh et al. 2002). High-contact service encounters facilitate the establishment of customer-employee relationships, leading to increased trust between these actors. Perceiving someone as socially close entails recognizing this person in a more concrete manner, which can result in a stronger sense of trust. When customers perceive a similarity with the service provider, they are more likely to identify with the employee, which subsequently reduces interpersonal barriers and increases customer trust. Similarity and identification represent two forms of perceived social proximity, hence closer service providers tend to be perceived as more trustworthy (Nguyen et al. 2020).

When there is a high level of trust, the consumer perceives a low-level of risk and is more willing to believe in the company. Considering that the service environment involves social interactions, trust is an important mechanism for driving relationships between consumers and employees. This happens because familiarity builds trust, which increases mutually supportive attitudes and behaviors (Darke et al. 2016). Trust, therefore, has the potential to enhance the effect of social proximity on behavior. The association between social distance and trust has been investigated in various contexts, including the customer-firm dimension and interpersonal relationships (Nguyen et al. 2020). In work-related settings such as sales departments, trust may be a buffer that reduces the occurrence of dysfunctional behavior (Choi et al. 2004). Similarly, in service contexts, trust increases customers’ propensity to adopt positive behaviors towards the company (Dang et al. 2020), which can be interpreted as a reduced willingness to misbehave.

Lower construal levels are associated with more favorable evaluations following a moral transgression, and trust can contribute to this effect by reducing an individual’s intentions to harm the company through engaging in dysfunctional behavior (Choi et al. 2004). When trust stems from self-identification and a concrete perception of the service provider, individuals are less likely to experience anger in response to a failure (Yagil and Luria 2016). In essence, customers who trust service providers tend to exhibit more positive reactions to transgressions occurring within socially closer contexts.

Therefore, we anticipate that trust will affect the relationship between social distance and the acceptability of DCB, so that higher levels of trust related to individuals perceived as socially closer will account for the reduced acceptability of such behavior. These individuals are inclined to scrutinize the service provider more attentively, considering concrete aspects, and trust serves to amplify this effect by decreasing the acceptability of retaliatory or dysfunctional behaviors that harm the service provider. Thus, we propose the following hypothesis:

H2

The effect of social distance on the acceptability of DCB is mediated by trust.

2.3 Failure controllability

Causal attribution represents a cognitive response to service failures that exerts an influence on emotional and subsequent behavioral reactions (Weiner 2000). Attribution theory encompasses three dimensions that possess distinct properties: locus of control, stability, and controllability. Locus of control pertains to the perceived origin of a cause (internal or external to the individual), stability refers to the relative resistance of the cause over time (stable vs. unstable), while controllability captures the extent to which the cause is subject to voluntary changes (controllable vs. uncontrollable) (Weiner 2014).

Despite being widely used in the consumer behavior field (Hampson et al. 2021), attribution theory has received little attention regarding its relationship with dysfunctional behavior. Previous research has suggested that the locus of attribution of consumer misbehavior (i.e. attributed to either the deviant customer or a service employee) influences perceived discomfort and acceptability of dysfunctional behavior towards the company that caused the issue (Rummelhagen and Benkenstein 2017). However, to the best of our knowledge, no previous research has verified the impact of failure controllability on deviant response acceptability.

When service failures are perceived as challenging for the company to control, customers tend to exhibit forgiving behaviors. Conversely, failures that are perceived as within the company’s control tend to provoke feelings of anger and a sense of betrayal (Li et al. 2023). Given that this dimension of attribution can engender punitive responses from customers (Weiner 2000), it constitutes the focal point of the present study.

Previous research has suggested that individuals may employ a perceived controllable workplace problem as a rationale for engaging in deviant behavior (Harvey et al. 2016). However, social distance has not been considered in this equation. Therefore, we expect that this effect will also extend to service consumption situations and will be influenced by the construal level. Accordingly, we propose the following hypothesis:

H3

When the service failure is perceived as controllable by a socially distant (close) service provider, DCB is considered more (less) acceptable.

2.4 The nature of dysfunctional behavior and consumer emotions

The nature of dysfunctional behavior can be categorized as opportunistic or premeditated. Opportunistic behavior refers to instances where the customer passively benefits from a situation, while premeditated responses involve the customer actively and intentionally creating a situation to his/her advantage (Zhao et al. 2020). For example, a customer who knowingly keeps excess change without returning it exhibits opportunistic behavior, whereas stealing something or tampering with price tags constitutes premeditated actions (Bossuyt et al. 2017).

The classification of behaviors based on their nature (opportunistic vs. premeditated) is closely related to the locus of failure. When a customer takes advantage of a company’s failure, the blame is placed on the company; however, when there is no failure on the company’s part and the customer premeditatedly misbehaves, the locus of blame shifts to the customer (Vitell and Muncy 2005). This locus of failure significantly influences the acceptability of the behavior (Bossuyt et al. 2017). Some authors also classify individuals as advantage seekers, who look for opportunities for personal gain through misbehavior, and agitators, who simply take the opportunities as they appear (Wilson et al. 2021).

Opportunistic conduct is generally and erroneously considered less blameworthy and more acceptable than premeditated actions (Vitell and Muncy 2005; Zhao et al. 2020). This study posits that the social distance between the customer and the employee would amplify this effect. Previous research has already demonstrated that individuals with lower construal levels tend to perceive more negative consequences of unethical behavior on their moral identity (Yang et al. 2022). Hence, we expect that opportunistic DCB would be considered more acceptable when the failure is attributed to a distant employee, that is, someone with whom the customer has a weaker emotional bond. On the other hand, when unethical behavior is premeditated in nature, it is suggested that acceptability would be comparable regardless of social distance (distant vs. close). Therefore, the following hypothesis is proposed:

H4

When the DCB is opportunistic and the service provider is perceived as socially distant (close), DCB is considered more (less) acceptable.

Despite the growing interest in understanding the influence of emotions on ethical decision-making, there is a lack of research focused on the role played by customers’ emotional reactions in situations involving deviant behaviors (Lages et al. 2023). Although misbehaving can potentially be associated with feelings of guilt, the impact of this emotion on customer perceptions and behaviors related to ethics is still limited (Arli et al. 2016; Escadas et al. 2019). Guilt is a negative emotion that arises from a decision that violates an individual’s values or standards. It can also be characterized as remorse or self-punishment (Izard 1977).

Anticipating emotions enables individuals to predict and evaluate the affective consequences of a decision and adjust their behavior accordingly (Kim et al. 2022). While it is known that anticipatory guilt reduces customer propensity to engage in unethical actions (Kim et al. 2022), previous studies have not explored the influence of the nature of misbehaviors on emotions. Considering that opportunistic dysfunctional behaviors appear to be more readily accepted by customers (Bossuyit et al. 2017), we assume that they will also affect anticipatory guilt and the acceptability of DCB towards the company. Consequently, the following hypothesis is formulated:

H5

When the DCB is opportunistic, a close (distant) social distance to the service provider triggers more (less) anticipated guilt, leading to lower (higher) acceptability of the DCB.

Figure 1 provides an overview of the conceptual model proposed in this study, illustrating the relationships among variables.

Fig. 1
figure 1

Conceptual model

Full size image

3 Methodology

We developed three experiments to test the proposed hypothesis. Data were collected online from Prolific respondents. A summary of sample profiles of the three studies is presented in Table 1. Participants read written scenarios (Appendix I) about a service failure situation with the manipulation of the independent variable (social distance) and the moderators (failure controllability and nature of behavior). The severity of failure, scenario realism, as well as the description of social distance, were pretested with the same population that was later the source of the sample.

Table 1 Summary of descriptive statistics across the 3 studies
Full size table

3.1 Study 1

3.1.1 Procedures

The primary objective of the initial study was to evaluate the acceptability of DCB across different levels of social distance (close vs. distant); therefore, this study employs a single-factor design. Participants were recruited from a digital platform (Prolific) and received a small payment fee in exchange for their participation. The sample was randomly assigned to one of two scenarios, both of which depicted a service failure situation at a restaurant. The choice of this context was motivated by its frequent use in research on service failure and its ability to allow individuals to easily envision themselves in such a consumption scenario.

According to the scenario, after finishing the meal, the customer requested the bill and noticed an error made by the waiter while registering the items ordered. The waiter had overcharged for more items than what was consumed. Upon realizing this mistake, the customer brought it to the waiter’s attention and requested a corrected bill. However, in the revised bill, the customer noticed that they were now undercharged, as a soft drink had not been included in the bill. Despite noticing this error, the customer chose not to point it out and proceeded to pay the bill before leaving the restaurant. In one scenario, the customer had no prior acquaintance with the waiter, while in the other the customer had been frequently served by the same waiter (socially distant vs. socially close). A summary of the manipulations can be found in Appendix I.

We conducted a pretest with 239 respondents recruited on Prolific (Mage = 44 years old, 58% male). There were no inconsistencies or interpretation problems in the scenario, nor differences in the perceived realism between scenarios (F = 0.40, p = 0.848). The manipulation checks confirmed that the scenarios were correctly perceived by respondents (F = 15.03, p < 0.001). Therefore, we proceeded with the use of this scenario in Study 1.

3.1.2 Measures

To measure the acceptability of DCB we used the scale developed by Neale and Fullerton (2010). The perception of social distance was assessed with a 2-item scale from Yan et al. (2016). Trust was measured using a 5-item scale (alpha = 0.86) adapted from Smith (1997). Additionally, the perception of the severity of the failure was assessed with a 2-item scale by Mattila (2001). Failure attribution was assessed as a control variable with 1 item adapted from Dong et al (2016). We also measured scenario realism and demographics. The full measurement list is available on Appendix II.

3.1.3 Results

Out of the 127 participants recruited and randomly assigned to the scenarios, 10 participants were excluded from the analysis due to failing the attention check, resulting in a final sample of 117 individuals (57% women, Mage = 36 years old). The perceived severity of the failure was moderate, with a mean rating of 3.46 (SD 1.66). Statistical analysis revealed no significant differences in the severity according to the scenarios (F (1, 114) = 1.66, p > 0.201). Hence this variable was not included as confound in the model.

The manipulation check yielded the expected results, confirming the successful manipulation of social distance (F (1, 115) = 47.80, p = 0.001). Participants in the socially distant condition recognized that the customer did not know the waiter (Mdistant = 1.72, SD 1.08), while those in the socially close condition acknowledged a pre-existing relationship between the customer and the waiter (Mclose = 3.73, SD 1.93). Furthermore, participants assigned to the socially distant condition perceived the customer and the waiter as socially distant (Mdistant = 1.65, SD 0.94), whereas those in the socially close condition viewed them as such (Mclose = 3.70, SD 1.97), thus confirming the manipulation (F (1, 115) = 50.94, p = 0.001).

To examine the direct effect, an ANOVA was conducted with social distance as the independent variable and the acceptability of the DCB as the dependent variable, and the analysis revealed a significant effect (F (1, 115) = 23.33, p = 0.001, η2 = 0.169). In the socially distant situation, the DCB was perceived as more acceptable (M = 4.32, SD 1.96) compared to the socially close situation (M = 2.72, SD 1.62).

Finally, the mediation analysis was performed using the Macro Process (Hayes 2013—Model 4). It was hypothesized that the psychological proximity between the parties would influence trust and, in turn, affect the acceptability of perceived dysfunctional behavior. The results indicated partial mediation, as there was an indirect effect (β = − 0.43, CI from − 0.71 to − 0.18); however, the direct effect remained significant, as depicted in Fig. 2.

Fig. 2
figure 2

Mediation effect of trust

Full size image

3.1.4 Discussion

Study 1 provided empirical evidence supporting the notion that when a service failure is attributed to a socially distant agent, customers are more likely to find dysfunctional behavior towards the company acceptable, compared to a scenario where the agent is socially close. The mechanism that helps to understand this result is trust, which mediates the direct effect, confirming hypotheses H1 and H2.

However, the impact of trust on the acceptability of DCB does not fully account for the observed variation in the dependent variable. The service literature suggests that the controllability of a failure can also play a significant role in shaping customer reactions following service recovery (Weiner 2014). When customers perceive that the company could control the situation and prevent failures, it can provide a justification for engaging in deviant behavior (Harvey et al. 2016). Consequently, study 2 was conducted to examine the effects of a failure perceived as controllable by the service provider.

3.2 Study 2

3.2.1 Procedures

The second study employed a 2 (socially distant vs. socially close) × 2 (controllable vs. uncontrollable failure) design. The scenario used in Study 1 was replicated, with the addition of a manipulation related to the controllability of the failure. Following the description of the failure, participants were informed either that the waiter acted imprudently and was not paying attention due to talking on a mobile phone (controllable failure), or that the restaurant was exceptionally busy (uncontrollable failure). All other procedures remained consistent with Study 1.

3.2.2 Measures

The measures used in Study 2 were identical to those employed in Study 1, with the inclusion of a 3-item scale to assess the controllability of the failure (Russell 1982). Similar to the previous study, the scenario was pretested, and participants were also recruited from the prolific platform.

3.2.3 Results

Out of the initial 247 completed questionnaires, 31 were deemed ineligible for analysis due to not passing the attention check, resulting in a final sample of 216 individuals (67% women, Mage = 35 years old). The manipulation check for the independent variable yielded the expected results (F (1, 215) = 65.16, p = 0.001). Participants in the socially distant situation correctly perceived that the customer did not know the waiter (Mdistant = 1.77, SD 1.02), while those in the socially close situation recognized a previous acquaintance between the customer and the waiter (Mclose = 3.55, SD 2.05). Additionally, participants accurately identified the social distance between the customer and the waiter, with those in the socially distant condition perceiving greater distance (Mdistant = 1.85, SD 1.14) compared to those in the socially close condition (Mclose = 3.62, SD 1.83), confirming the manipulation worked as expected (F (1, 215) = 72.36, p = 0.001). The manipulation check regarding the controllability of the failure was also confirmed (F (1, 215) = 75.77, p = 0.001). Participants perceived the failure as controllable (M = 6.72, SD 0.79) when the waiter was distracted, whereas they considered the failure as uncontrollable in the situation where the restaurant was crowded (M = 5.33, SD 1.47).

The failure was perceived as moderately severe (M = 3.79, SD 1.65). While respondents perceived varying levels of severity across scenarios (F (1, 215) = 8.16, p = 0.005), there were no differences in the main effects when we considered severity as a covariate in the model. Therefore, to maintain consistency with Study 1, we did not include perceived severity in the model. The direct effect was examined through an ANOVA, with social distance as the independent variable and the acceptability of the DCB as the dependent variable. The analysis revealed a significant result (F (1, 215) = 25.08, p = 0.001, η2 = 0.11), indicating that the acceptability of the DCB differed based on the level of social distance. Specifically, the DCB was considered more acceptable in the socially distant situation (M = 4.31, SD 1.88) compared to the socially close situation (M = 3.02, SD 1.91).

To examine the moderation effect of the controllability of the failure, we used the Macro Process (Model 1). The analysis revealed a significant interaction (β = − 1.62, t (213) =  − 3.23, p = 0.001), indicating that the perceived controllability of the failure influences the acceptability of DCB (β = − 2.09, CI from − 2.79 to − 1.40). Specifically, when the failure was perceived as controllable, there was a significant difference in the acceptability of the DCB between the socially distant group (M = 4.57, SD 1.90) and the socially close group (M = 2.47, SD 1.81), with higher acceptability observed in the socially distant group. However, when the failure was perceived as uncontrollable, the moderation effect was not significant (β = − 0.47, CI from − 1.17 to 0.23), and both the socially distant group (Mdistant = 4.06, SD 1.84) and the socially close group (Mclose = 3.58, SD 1.84) showed similar levels of acceptability of DCB.

3.2.4 Discussion

Study 2 corroborated H1, which was already confirmed in Study 1, and also provided support for H3. It confirmed that when a failure is attributed to an agent who is perceived as socially distant, the customer’s acceptance of dysfunctional behavior towards the company increases when the failure is perceived as controllable by the service provider. However, the moderation effect of controllability is not significant when the failure is perceived as uncontrollable.

These findings provide support for hypothesis H3 and shed light on the boundary condition that helps to explain why DCB can be perceived as acceptable in the context of service failures. However, another factor that may contribute to this perception is the nature of the behavior itself, specifically whether it is opportunistic or premeditated. In the case of opportunistic behavior, a failure on the part of the restaurant inadvertently enables customers to misbehave, such as issuing a bill with a missing item consumed. Conversely, in the case of premeditated behavior, customers proactively engage in unethical actions, such as falsely claiming that a dish was served cold (Vitell and Muncy 2005). Study 3 was carried out to investigate the role of the type of DCB in a failure context and to examine the potential moderating effect of anticipatory guilt.

3.3 Study 3

3.3.1 Procedures

When customers take advantage of a company’s failure, engaging in opportunistic behavior, blame is usually placed on the company. Conversely, when clients premeditatedly misbehave, blame shifts to the customer (Vitell and Muncy 2005). As opportunistic behavior is often viewed as less censurable than premeditated actions (Zhao et al. 2020), Study 3 aims to investigate whether the social distance between the customer and the employee can amplify this effect. To do so, Study 3 employed a 2 (socially distant vs. socially close) × 2 (opportunistic vs. premeditated behavior) design. The same scenario used in previous studies was utilized, with the only difference being the manipulation of the nature of the DCB. In the opportunistic behavior scenario, the mistake in the bill involved the waiter neglecting to include a soft drink, and the customer choosing not to draw attention to this error, simply paying the bill and leaving. In the premeditated behavior scenario, the bill was correct, but the customer dishonesty claimed that the quality of the chips was unsatisfactory, leading the restaurant to waive the charge for this item. Prior to the main study, a pretest was carried out to verify which types of dysfunctional behaviors were perceived by respondents as premeditated or opportunistic.

3.3.2 Measures

In addition to the measures used in previous studies, we included a single question to check respondent’s perceptions about the nature of the dysfunctional behavior (Vitell and Muncy 2005) and a 3-item scale to measure guilt (Izard 1977). Data were collected on prolific platform.

3.3.3 Results

Out of the total 238 completed questionnaires, 16 were excluded because they did not pass the attention check, resulting in a final sample of 222 individuals (63% women, Mage = 35 years old). The manipulation worked as expected (F (1, 219) = 90.05, p = 0.001), and participants assigned to the socially distant scenario perceived the waiter as socially distant compared to those assigned to the socially close scenario (Mdistant = 1.91, SD 1.15; Mclose = 3.81, SD 1.75). The manipulation of the nature of the dysfunctional behaviour also yielded significant results (F (1, 219) = 18.99, p = 0.005), and respondents in the opportunistic behaviour scenario perceived the situation as such (M = 2.22, SD 0.75), whereas those assigned to the premeditated scenario perceived the customer as the main actor in the deviant conduct (M = 2.66, SD 0.75).

Once again, the failure was perceived as moderately severe (M = 3.45, SD 1.50), and despite different levels of severity associated with the scenarios (F (1, 220) = 11.33, p < 0.001), there were no relevant differences on the acceptability of dysfunctional behavior. The direct effect was confirmed (F (1,220) = 9.82, p = 0.002, η2 = 0.04), indicating that individuals in the socially distant condition perceived DCB as more acceptable compared to those in socially close condition (Mdistant = 2.94, SD 1.91; Mclose = 2.23, SD 1.49). Model 8 of the Macro Process was used for analyzing moderated mediation, with social distance as the independent variable and the acceptability of DCB as the dependent variable. To test the indirect effect, guilt was the mediator and the nature of the DCB was the moderator.

In terms of mediation, the direct effect between social distance and acceptability of DCB was not significant (β = − 0.12; p = 0.063). However, the indirect path through anticipatory guilt was significant (β = 0.19; p = 0.008) as well as the path between guilt and acceptability of DCB (β = − 0.50; p < 0.001). An interaction between the nature of the DCB and guilt was also observed (β = − 1.82, t (217) =  − 3.84, p = 0.001), confirming the presence of moderated mediation (β = 0.76, CI from 0.34 to 1.26).

The indirect effect was found to exist when the DCB was perceived as opportunistic (β = 1.78, CI from 1.13 to 2.44), but not when it was perceived as premeditated (β = − 0.03, CI from − 0.70 to 0.63). When the behavior was opportunistic, a greater social distance between the customer and the service provider led to a higher acceptability of DCB (Mdistant = 4.00, SD 1.88; Mclose = 2.58, SD 1.61). In other words, when the customer is perceived as benefiting passively from the employee’s mistake, higher levels of guilt are associated with a socially close service provider, resulting in lower acceptability of adopting DCB (Fig. 3). Detailed results are available in Appendix III.

Fig. 3
figure 3

Moderation of the opportunistic nature in the acceptability of DCB

Full size image

3.3.4 Discussion

Study 3 provided evidence for the mediating role of guilt in reducing the acceptability of the DCB. Additionally, the study confirmed that the nature of DCB moderates the relationship between social distance and the acceptability of DCB. The moderated mediation effect supported the notion that these variables collectively influence customers’ perception of the acceptability of dysfunctional behavior. Notably, the effects of guilt and social distance were found to be contingent on the opportunistic nature of the action, thus confirming hypothesis H5.

Opportunistic DCB is likely to evoke higher levels of anticipated guilt and potentially decrease the likelihood of engaging in deviant actions towards a psychologically close employee. Nonetheless, when the victim is psychologically distant from the customer, the same DCB is perceived as more acceptable. This effect can be explained by considering the nature of unethical behavior and construal-level theory, as suggested by previous research (Bossuyt et al. 2017; Zhao et al. 2020).

Studies 1 and 2 offered support for hypotheses H1, H2 and H3, while Study 3 confirmed hypotheses H4 and H5. These three experimental studies have provided insights into the factors influencing the acceptability of dysfunctional behavior in service failure situations involving both socially close and socially distant actors.

4 Discussion and conclusions

This article focuses on examining the acceptability of dysfunctional customer behavior concerning psychological distance, as derived from construal-level theory. Specifically, it investigates how the perceived social distance between the individual responsible for a fault and the customer influences the acceptability of dysfunctional customer behavior. The findings indicate that psychologically close individuals are less likely to perceive engaging in deviant actions as acceptable. The underlying condition that helps to explain this effect is the trust consumers have in socially close service providers, besides perceptions of failure controllability. Furthermore, anticipated guilt is a variable that decreases the likelihood of perceiving DCB as acceptable, especially when the deviant behavior is opportunistic.

4.1 Theoretical implications

Research findings confirm that consumers who perceive themselves as socially close to service providers are less prone to adopt deviant behaviors. While literature has suggested that levels of proximity could influence consumer propensity to misbehave (Wan et al. 2021), no previous research has specifically focused on dysfunctional behavior as a potential response to service failure contexts. Therefore, this research contributes to the fields of service failure and recovery and deviant consumer behaviors using the construal level as the framework of analysis.

Moreover, the results confirm trust as an underlying mechanism that explains reduced DCB acceptability regarding close service providers. Socially close individuals are more relatable and identifiable, leading to greater trust. Consequently, the offender is more inclined to carefully assess the situation and engage in behaviors aimed at minimizing harm to the socially close subject (Dootson et al. 2016; Nguyen et al. 2020). This finding shed light on the role of employees during service encounters and the influence of social distance on trust (Nguyen et al. 2020). When customers have a higher level of confidence in the employee, they develop a deeper understanding of the circumstances that led to the failure (Yagil and Luria 2016; Kim et al. 2019). Additionally, when there is a strong sense of trust, customers perceive lower levels of risk and are more willing to believe in the company, ultimately reducing the acceptability of dysfunctional customer behavior.

Employees, for instance, may not always elicit positive reactions from customers. The perception of a lack of control regarding a service failure can trigger retaliatory responses, particularly when customers believe that the person responsible for the problem could have prevented it (Gong and Wang 2021). This study contributes to this discussion by confirming that when a socially distant service provider is perceived to have control over a failure, it leads to a higher acceptability of dishonest actions. The explanation for this effect lies in attribution theory, which serves as a justification for behavior that would otherwise be deemed unacceptable (Harvey et al. 2016). This research differs from existing studies as it specifically focuses on customers’ dysfunctional behavior and considers social proximity as a relevant factor in this outcome.

Furthermore, emotions such as guilt play a role in explaining the effect of psychological distance on uncivil behavior (Kim et al. 2022; Lages et al. 2023). Academics and practitioners should have a better understanding not only about how dysfunctional responses impact employees and observing customers, but also how core emotions influence consumers adoption of deviant behaviors (Lages et al. 2023). Additionally, a better understanding of the impact of the nature of these behaviors on perceptions can allow progress towards curbing such actions. This research is also aligned with the agenda proposed by Wilson et al. (2021) to gain a deeper understanding of the extent to which customers perceive subversive or dysfunctional behaviors as positive or acceptable, and how failures contribute to this perception.

Apart from examining attribution theory, this article provides other significant contributions. One such contribution is the comprehensive exploration of ethical behaviors, with a specific emphasis on the acceptability of dysfunctional customer behaviors and the role of failures in shaping this perception, which is prevalent across various service contexts and still evolving in marketing and consumption literature (Jin et al. 2022; Wilson et al. 2021). Previous research has suggested a relationship between construal levels and ethical behavior (Pinto et al. 2020), but with no specific focus on service failure and its potential behavioral responses. This article’s unique contribution to the topic lies in its examination of the effects of psychological distance, as it integrates the drivers of dysfunctional behavior with psychological, cognitive, and emotional aspects within the framework of construal-level theory (Adler and Sarstedt 2021).

These findings also suggest that, even in cases where the service provided falls short of expectations, the customer-employee relationship can serve as a protective factor against the negative effects of such failures. While studies examining transactional relationships between customers and companies are plentiful in literature (Dootson et al. 2016; Gong and Wang 2021), there is a dearth of research supported by construal-level theory that specifically investigates dysfunctional customer behavior within the context of the customer-employee relationship. Trust, for instance, is extensively studied in relationship marketing literature, known to increase customers’ propensity to adopt positive behaviors towards the company (Dang et al. 2020). However, previous research has not explored the influence of trust on the acceptance of DCB considering different levels of psychological proximity between actors. This research demonstrates that customers exhibit a higher level of trust in employees with whom they share a closer social connection compared to those perceived with a greater social distance. As a result, customers who identify with employees are more likely to trust them, potentially mitigating the impact of negative experiences during their interactions, such as service failures.

Moreover, this study establishes a connection between the acceptability of dysfunctional behavior and a theoretical framework that represents an evolution of the theories encompassing the field of causal attribution (Weiner 2014). One of the key dimensions of this theory, namely the ability to control failure, emerges as a moderator in the relationship between social distance and the acceptability of DCB. Aligning with the tenets of attribution theory, the ability to control failure acts as a moderating factor, influencing how customers assimilate and respond to a failure event. Notably, even in cases where the failure is deemed controllable, customers who identify with employees can exert an influence on how the failure is perceived and subsequently impact the acceptability of DCB.

4.2 Practical implications

In addition to its academic contributions, this research holds significant implications for management practices. The findings underscore the fact that customers consider dysfunctional behaviors acceptable in certain circumstances. Consequently, organizations must develop management strategies that proactively address and mitigate such detrimental customer behavior. Emphasizing the customer’s sovereignty becomes contradictory when their dysfunctional actions can cause harm to an organization. Hence, adopting a management approach that effectively addresses dysfunctional behavior is essential not only for the direct victims of such actions (i.e., employees and companies) but also for other customers. Consumer misbehavior not only leads to financial losses but also undermines the existing relationship between customers and service providers, particularly in the context of service failure and subsequent recovery efforts. Given that dysfunctional customer behavior tends to be more acceptable in situations characterized by significant social distance between the parties involved, especially when it occurs passively, companies can devise management strategies to curb this behavior.

Service companies can invest in training and development programs for their frontline employees to enhance their interpersonal and communication skills. This can help employees build better rapport with customers and create a sense of social closeness. Moreover, companies can stimulate employees to personalize their interactions with customers. For example, using the customer’s name or remembering past interactions, employees can make customers feel socially closer to the service provider and build trust.

Considering that attribution also plays a relevant role regarding customer responses, when service failures occur, service companies should take a transparent and accountable approach. If the service provider is perceived to have control over the failure, acknowledge the issue, communicate the steps being taken to rectify it, and provide solutions promptly may be interesting ideas to offer consumers a sense that the issue will not happen again, therefore, mitigating potential customer dishonest behavior.

Furthermore, empowering frontline employees to make decisions and resolve issues without needing to escalate problems to higher levels of management offers consumers a sense that employees have control over the situation. Likewise, when customers feel their feedback is valued and acted upon, they are more likely to trust the service provider. Therefore, collecting customer feedback and using it to improve services continuously has the potential not only to enhance closer relationships but also to develop trust.

Fostering customer engagement with the brand can also be done in online channels, using customer relationship management (CRM) software to track and personalize customer interactions, hence establishing closer and more enduring connections. Moreover, another way of developing lower construal-level perceptions can be to utilize customer data for preference-based segmentation. By understanding individual customers better, service companies can tailor their approaches to create a sense of social closeness with each customer segment.

By nurturing such relational exchanges, companies not only have the potential to promote customer loyalty but also induce lower construal levels which imply a perception of closeness between customers and employees, ultimately serving as a means to indirectly reduce the occurrence of DCB.

4.3 Limitations and future research

This study has some limitations that warrant consideration. Firstly, it focuses solely on one service context, and the effects observed in this context may differ in services characterized by higher levels of customer involvement. Future research could explore different service contexts to assess whether the effects are consistent across various settings or if there are mitigating or exacerbating factors at play.

Secondly, this study only examined the perception of dysfunctional behavior, rather than directly measuring the adoption of such behavior. It would be valuable for future studies to include measures of actual behavioral outcomes and explore other dependent variables associated with retaliatory responses, such as negative word of mouth or the posting of false reviews. Furthermore, previous literature suggested that consumers from different cultures react differently to service failures (Rohden and Matos 2022). Hence, new research could adopt a comparative lens aiming to analyze if the perceptions of deviant behavior acceptability diverge considering distinct backgrounds.

Additionally, future research could delve into underlying mechanisms that help explain the observed effects. Previous studies have suggested that anger is a variable that can help to understand how the attribution process justifies deviant behaviors. Therefore, investigating the role of anger in this context could provide further insights (Harvey et al. 2016).

Moreover, future research could explore additional factors that contribute to individuals engaging in dysfunctional behavior (Jin et al. 2022). Personal traits such as materialism and perceptions of the service provider’s competence have been found to influence engagement in dysfunctional behaviors. Recent research has also suggested that in a service failure situation, consumer´s perceptions that the treatment received was in some way discriminatory can also potentially influence responses such as retaliatory behaviors (Rohden Pizzutti 2023). It would be interesting to examine how social proximity, through the lens of construal-level theory, would interact with these variables.