Dysfunctional Customer Behavior Severity: An Empirical Examination

Journal of Retailing 85 (3, 2009) 321–335

Dysfunctional Customer Behavior Severity: An Empirical Examination Kate L. Reynolds a,∗ , Lloyd C. Harris b a

Cardiff Business School, Cardiff University, United Kingdom b Warwick Business School, Coventry, United Kingdom

Abstract Although many studies assume that customers monotonically act in both a functional and a good-mannered way during exchange, considerable anecdotal evidence suggests that customers routinely behave negatively and often disrupt otherwise functional encounters. However, to date, rigorous empirical evidence of this phenomenon is lacking. This study synthesizes extant literature from a broad range of areas and advances two alternative conceptions of the factors associated with dysfunctional customer behavior severity. That is, after controlling for a variety of factors, the authors suggest that psychological obstructionism, disaffection with service, and servicescape variables are significantly associated with the severity of deliberate dysfunctional customer acts. The results provide insights for researchers interested in the darker side of service dynamics and generate useful implications for services practitioners charged with reducing the severity and the frequency of episodes of deviant customer behavior. © 2009 New York University. Published by Elsevier Inc. All rights reserved. Keywords: Dysfunctional customer behaviour; Customer misbehaviour; Consumer deviance; Antecedent; Structural equation modelling

The majority of research into customer–firm interactions is founded on the assumption that customers act in both a functional and a good-mannered way (e.g., Ringberg, Odekerken-Schröder, and Christensen 2007). This contrasts with practitioner-oriented research that repetitively alludes to customers behaving badly (e.g., Dube 2003) and with intermittent scholarly studies that typically highlight the prevalence of a single form of customer misbehavior, such as shoplifting (Kallis and Vanier 1985) and illegitimate complaining (Reynolds and Harris 2005). Despite limited academic attention, the pervasiveness of customer dysfunction appears global. Focusing on one individual form of customer misbehavior, Grandey, Dickter, and Sin (2004) reveal that, on average, service employees within the United States fall victim to episodes of customer aggression ten times a day. These findings are comparable to that of a study conducted in the United Kingdom (USDAW 2004), which reveals that front-of-store assistants are subjected to verbal abuse once every 3.75 days, to threatening behavior every 15 days, and to acts of violence every 31 days. Moreover, Bamfield (2006) provides evidence of the ominous rise in thefts by consumers across

∗

Corresponding author. E-mail addresses: [email protected] (K.L. Reynolds), [email protected] (L.C. Harris).

several countries, including the Czech Republic, Japan, Iceland, and New Zealand. This leads Reynolds and Harris (2006) to argue that customer misbehavior is endemic within the service industry. Fullerton and Punj (2004) suggest that norm-violating behavior is pervasive and representative of everyday customer behavior, rather than constituting a segregate faction of society (see also Harris and Reynolds 2004). The disparate focus on individual forms of dysfunctional customer behavior is detrimental to a broader understanding of these issues. Indeed, Fullerton and Punj (1993) argue that there is a need to elucidate the range of antecedents and to provide empirical insights into their dynamics. Thus, research that examines the factors associated with dysfunctional customer behavior is littered with calls for future studies to examine its antecedents more thoroughly (e.g., Al-Rafee and Cronan 2006). Fullerton and Punj (1993, 2004) stress the need for “better” data that captures a more inclusive investigative approach. Echoing this, Harris and Reynolds (2003) call for research to examine the antecedents of dysfunctional customer behavior more holistically. The current research responds to these calls by investigating the factors associated with dysfunctional customer behavior severity, that is the extent to which a customer deliberately behaves in a way that violates the norms and unwritten rules of an individual service setting in a negative fashion. Our aim is to amalgamate previous insights and to examine empirically actual (as opposed to speculative) incidents of customer misbehavior.

0022-4359/$ – see front matter © 2009 New York University. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.jretai.2009.05.005

322

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335

In assuming a norm-breaking perspective, we draw on literature from diverse areas, including: sociology, psychology, criminology, ethics, environmental psychology, marketing, and employee deviance that offer insight into the constructs correlated with dysfunctional behavior. Synthesis of these literatures leads to the forwarding of three main constructs that associate with customer misbehavior severity: psychological obstructionism, disaffection with service, and servicescape variables. By integrating insights from wide-ranging literature streams and drawing on multiple theoretical bases, we develop and test our research model and a rival model. Specifically, our research model is founded upon the propositions of Bitner (1992) and Fullerton and Punj (1993). By contrast, our rival model is inspired by literature that approaches customer dysfunction in a linear and direct fashion (e.g., Phillips, Alexander, and Shaw 2005). Further details of which are explored in the later discussion. The managerial relevance of this study is evident. Our model provides insights into customer dysfunction that may help managers reduce such behaviors. That is, we reveal that managers might manipulate many of the factors that associate with customer misbehavior. Our study is also of interest to marketing theorists. By drawing on diverse research areas, this paper makes a conceptual contribution in deepening the understanding of the factors that relate to customer misbehavior severity. Furthermore, this study makes a methodological contribution through the development and validation of multi-item scales. Finally, this study contributes to the theory of customer deviance by operationalizing and empirically examining conceptual frameworks that depict the constructs associated with dysfunctional customer behavior holistically. Literature review Evidence pertaining to the prevalence of dysfunctional customer behavior has drawn the attention of a small but growing number of academies that have supplied insights into this phenomenon. Such studies often adopt the label “dysfunctional customer behavior,” which refers to behavior by consumers within the exchange setting that deliberately violates the generally accepted norms of conduct in such situations. We use the term “dysfunctional customer behavior” because of its emphasis on the issues of intent and norm infringement. We divide the studies that generate insight into dysfunctional customer behavior into three themes. The first theme focuses on profiling the different forms of customer misbehavior. Possibly the best-known categorization is from the anecdotal work of Lovelock (1994), who identifies six service-based jaycustomers. Contrasting typologies are also offered by Fullerton and Punj (2004), Harris and Reynolds (2004), and most recently, Berry and Seiders (2008). However, although these classifications offer notable insights into the diverse varieties of customer misbehavior, such studies lack empirical support. The second theme of research focuses on the consequences of dysfunctional customer behavior. Specifically, the effects of customer misbehaviors are wide-ranging affecting employees, firms, and fellow customers (Harris and Reynolds 2003). Yet, despite the grave

implications of customer deviance, research in this area is in its infancy and tends to be exploratory. The third theme of research, and indeed, the focus of the current paper, pertains to the drivers of dysfunctional customer behavior. Typically, existing studies are framed within a specific academy, including sociology (Rosenbaum and Kuntze 2003), psychology (Al-Rafee and Cronan 2006), and business ethics (Fukukawa 2002). Here, the main research focus is on exploring the antecedents of individual forms of misbehavior. In particular, shoplifting and consumer fraud have received prolific attention over the past four decades (Harris 2008). However, although the majority of research into the antecedents of dysfunctional customer behavior has centered on shoplifting, sporadic insights into other forms of customer misbehavior include consumer resistance, vandalism, illegitimate complaining, and rage (see Grove, Fisk, and John 2004; Reynolds and Harris 2005). Factors associated with dysfunctional customer behavior severity Before we model the antecedents of dysfunctional customer behavior, the construct of dysfunctional customer behavior severity requires further elaboration. In an attempt to assimilate norm breaking into a conceptual structure and an actionable dependent variable, several studies support the validity of researching the perceived severity of dysfunctional behavior (see Lawrence and Robinson 2007; Vitell and Muncy 1992). Harris and Reynolds (2004) advocate the study of people who knowingly break behavioral norms. Given these arguments, the focus of our study centers on the severity of dysfunctional customer behavior in terms of the extent to which a customer deliberately behaves in a way that violates the norms and unwritten rules of an individual service setting in a negative fashion. The disparate nature of existing research which offers insight into the factors associated with dysfunctional customer behavior results in a multitude of possible constructs of interest, thus representing a challenge to the researchers to identify those most suitable for study within the consumer and services context. Consequently, a number of iterative processes were employed by the authors to identify the most relevant constructs. These stages include reviewing the literature to ascertain the breadth and depth of study of each construct, and the contextual and methodological applicability of each construct. For example, the role of environmental (servicescape) variables in episodes of misbehavior is discussed across a wide range of disciplines including: ethics, environmental psychology, criminology, sociology, and strategic marketing. The cumulative outcome of this process revealed three reflective factors: psychological obstructionism, disaffection with service, and servicescape, as those most worthy of further examination. Psychological obstructionism In terms of the first of our three main associative factors, Fullerton and Punj (1993) draw a link between consumers’ personality traits and predispositions, and dysfunctional customer behavior. Within the context of our research, we utilize

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335

the term “psychological obstructionism” to denote the enduring personality traits and predispositions, which impede individual consumers from behaving in a normative fashion, and shape and constrain their interpretations of service encounters. In this regard, psychological obstructionism is present before, during, and after an exchange. Specifically, the dimensions of Machiavellianism, sensation seeking, aggressiveness, and consumer alienation are conceptualized here as reflective dimensions of psychological obstructionism. Within the fields of psychology and criminology, theorists pay considerable attention to the study of personality and individual misbehavior. An early study in criminology was that of Eysenck (1964), who proposed a relationship between hereditary personality traits and criminal behaviors. Subsequently, numerous studies have correlated aspects of psychological obstructionism and acts of deviance (Romero, Luengo, and Sobral 2001; Sutherland and Shepherd 2002). Studies focusing on exchange settings have also recognized the role of personality in driving acts of dysfunctional behavior. Fox and Spector (1999) and Harris and Ogbonna (2006) both advance personality variables to correlate with acts of workplace sabotage by employees, while Rallapalli et al. (1994) and Ross and Robertson (2003) separately forward personality as key in ethical decision-making processes. With reference to dysfunctional customer behavior, several studies draw links between personality dimensions and individual forms of misbehavior, including shoplifting (Kallis and Vanier 1985), cheating behavior (Wirtz and Kum 2004), and digital piracy (Al-Rafee and Cronan 2006). Disaffection with service Consumers’ appraisals of the exchange encounter are also associated with acts of deviant behavior. Mills (1981) draws a direct link between consumer disaffection and dysfunctional customer behavior. Thus, we define “disaffection with service” as customers’ negative cognitive-emotive evaluations of a service provision. Specifically, the dimensions of dissatisfaction and inequity are conceptualized as reflective indicators of disaffection with service. Disaffection with service occurs during service exchange and therefore implicitly comprises perceptions of front-line service employees who are intrinsically linked with service provision and the exchange experience (Bitner 1992). This contrasts with psychological obstructionism, which is both pre-existing and enduring. Indeed, customers’ negative interpretations and reflections of a given situation are prominent antecedents of acts of undesirable behavior (Lee and Allen 2002). Specifically, judgments of dissatisfaction and inequity are key motives for incidents of dysfunctional behavior (Diamond 1997; Fullerton and Punj 2004).

comprising four reflective sub-dimensions: layout and design, atmospheric environment, behavior of fellow customers, and exterior environment. As with disaffection with service, evaluations regarding servicescape variables occur during service. Although prior research has attempted to examine the effect of servicescape variables on consumers’ behavior (Morin, Dubé, and Chebat 2007), only recent advancements within marketing theory note a link between servicescape design and customer misbehavior (e.g., Areni 2003). Nevertheless, support for the relationship between environmental design and misbehavior is evident within the environmental psychology and criminology literature streams (Hopkins 2002; Wilson and Kelling 1982). Therefore, as mentioned previously, this synthesis leads us to advance the three constructs of psychological obstructionism, disaffection with service, and servicescape variables as those that are most consistently linked to dysfunctional behavior. However, there is little consensus as to how these constructs should be modeled. Indeed, different research traditions infer contrasting theoretical models. Thus, we devise two competing models and detail each in turn. Because of the degree of overlap between the hypothesized paths across the two competing models, we present a summary of the theoretical foundation for each model. We follow this with a detailed discussion of each of the five hypotheses. Theoretical models Our favored model, the research model (see Fig. 1), is grounded in Bitner’s (1992) and Fullerton and Punj’s (1993) frameworks and research traditions and centers on the indirect relationship between servicescape variables and dysfunctional customer behavior severity. Drawing on environmental psychology, Bitner (1992) suggests that servicescapes do not directly affect consumer behavior, but rather that a person’s internal responses mediate the relationship between the servicescape and the behavior. Thus, the research model hypothesizes a path between servicescape variables and disaffection with service and a relationship between disaffection with service and the severity of dysfunctional customer behavior. This indirect relationship is investigated extensively in studies of “functional” consumer behavior within environmental psychology and servicescaperelated literature streams (Lin 2004; Williams and Dargel 2004)

Servicescape Research also highlights the role of the characteristics of the exchange setting in influencing deviant customer behavior. Individual’s interpretation of the tangible and intangible characteristics of the exchange setting, which we label “servicescape” for reasons of parsimony, is conceptualized as

323

Fig. 1. Research model.

324

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335

Fig. 2. Rival model.

and is suggested within the domain of customer dysfunction (Fullerton and Punj 1993; Mills 1981). Furthermore, Bitner (1992) acknowledges the role of personality traits in influencing a person’s evaluation of his or her physical surroundings, a view that is upheld within the sphere of deviant behavior by Eysenck and Eysenck (1970) and Fullerton and Punj (1993). Further echoing Fullerton and Punj’s (1993) model of aberrant customer behavior, the research model hypothesizes a link between psychological obstructionism and the severity of dysfunctional customer behavior. In their conceptual model, Fullerton and Punj (1993) propose a relationship between personality traits and aberrant customer behavior. Support for this hypothesis with specific reference to deviant behavior also derives from studies of business ethics, criminology, and psychology (Gottfredson and Hirschi 1990; Rayburn and Rayburn 1996). Given the strength of conceptual support for the research model, we favor it above the rival model. The rival model (see Fig. 2) is distinct in that it depicts only direct linear relationships between each of the three constructs of interest and the severity of dysfunctional customer behavior. This model originates from and thus reflects the disparate nature of prior research into deviant behavior, much of which focuses on the association between a single antecedent and a specific form of misbehavior. Therefore, in addition to the direct paths with severity, as presented in our preferred research model, this model suggests that there is a direct relationship between servicescape and severity. This is in direct contrast to traditions that support a mediated model (Bitner 1992; Mehrabian and Russell 1974). Support for this model comes from multiple research disciplines. For example, Kuo and Sullivan (2001) suggest that the relationship between physical environments and aggressive behavior is strong. Similarly, Phillips, Alexander, and Shaw (2005) highlight a link between the physical design of servicescapes and incidents of consumer theft. Influence of interpretations of the servicescape on disaffection with service The view that an organization’s servicescape is associated with consumer disaffection is widely debated and supported within services literatures (see Bitner 1992). Barnes, King, and Breen (2004) who finds that perceptions of environmental factors contribute to feelings of customer disaffection demonstrate

this. Similarly, Bitner (1992) argues that servicescapes do not directly cause consumers to behave in a certain fashion, but rather, behaviors are mediated by an individuals’ interpretation of the servicescape and their overall evaluations of the store. More specifically, Baker and Cameron (1996) argue that it is the combination of servicescape elements (for example, loud music, repellent decorative colors, uncomfortable furniture, and long queues), that antecede feelings of customer disaffection. Providing complementary findings within the context of the hospitality industry, Schmidt and Sapsford (1995) forward evidence of an empirical relationship between negative perceptions of servicescape and customer disaffection. Moreover, in utilizing critical incident technique, Bäckström and Johansson (2006) reveal a combination of servicescape constructs to give rise to poor service experiences, in that negative interpretations of the servicescape put consumers in a negative frame of mind, which increases the likelihood that the consumers will be disaffected with the overall service encounter. Complementary evidence is also forwarded by a number of studies, which find an association between negative interpretations of servicescapes and increased levels of customer displeasure and thus, disaffection (e.g., d’Astous 2000). Thus: H1 . The greater the negative interpretation of an outlet’s servicescape, the higher is the level of perceived disaffection with service. Influence of disaffection with service on the severity of dysfunctional customer behavior Commentators often position dissatisfaction and inequity (disaffection) as key drivers of misbehavior (see Harris and Reynolds 2004). In addition to research that examines the link between dissatisfaction and inequity, and incidents of dysfunctional behavior in a separate fashion (see Lee and Allen 2002; Mills 1981), prolific support for the relationship between disaffection and deviance is offered within both employee and consumer contexts. For example, Diamond (1997) forwards empirical results that link employee job disaffection with acts of employee sabotage. Robinson and Bennett (1997) who presents evidence of a relationship between worker disaffection and employee misbehavior, also echo this finding. Focusing on the deviant activities of consumers, Lovelock (1994), and Wirtz and Kum (2004) separately note the presence of customer disaffection in increasing the intensity of acts of belligerent and cheating behaviors. Offering corresponding findings, Yi and Gong (2008), provide empirical support for the association between customer disaffection and dysfunctional customer behavior within a student context. Focusing specifically on acts of consumer retaliation, Huefner and Hunt (2000) offer fascinating insight that depicts how consumers engage in misbehaviors of ranging severities, including theft, vandalism and physical violence, as a means of expressing their disaffection with an individual organization or employee. Thus: H2 . The higher the level of disaffection with service, the greater is the severity of dysfunctional customer behavior.

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335

Influence of psychological obstructionism on the evaluation of the servicescape Support for a link between psychological obstructionism and individual’s evaluations of servicescape can be sourced from multiple literatures including: criminology, environmental psychology, consumer behavior, and organizational deviance (Fox and Spector 1999; Lindsay and Anderson 2000). Here, it is argued that an individual’s personality traits and predispositions will affect and shape how they interpret the world around them. In particular, aspects of psychological obstructionism (Machiavellianism, aggressiveness, sensation seeking and consumer alienation) are commonly argued to foster negative interpretations of environments (e.g., Mudrack 1993; Slater 2003). That is to suggest, individuals’ who are high in psychological obstructionism inherently observe the environmental settings in which they encounter in a negative light, in comparison to persons who exhibit low levels of these traits. Offering insight into the mechanism of this relationship concerning normative consumer behavior within service settings, Bitner (1992) maintains that an individual’s analysis of a servicescape environment is influenced by their personality. Also focusing on consumers, Aylott and Mitchell (1998) uncover evidence to suggest that the manner in which individuals respond to stressors within the environment is shaped by their personality and predispositions. Within the context of unethical and deviant behaviors, in examining the drivers of criminal behaviors, Eysenck and Eysenck (1970) suggest that an individual’s enduring traits are linked to their interpretation of the environment. Focusing on the aberrant acts of consumers, Fullerton and Punj (1993) argue that individuals who are high in psychological obstructionism traits are more prone to viewing servicescapes in a negative fashion owing to their predominantly negative outlook. This, Fullerton and Punj (1993) suggest, may ultimately result in acts of aberrant customer behavior. Aligned with this argument, Robinson and O’Leary-Kelly (1998) draw on the perspective of attraction–selection–attrition (Schneider 1987). In particular, Robinson and O’Leary-Kelly (1998) suggest that persons who possess anti-social tendencies or personality traits are instinctively drawn to certain types of environments that foster such behaviors. Providing an aligned argument, Kenrick et al. (1990) suggest that within certain environmental settings, specific personality characteristics may become more apparent. Concurrently, Williams and Dargel (2004) discuss the proposition that individual personality and predispositions affect the way in which individuals screen, and thus ultimately respond to, environmental cues. Thus: H3 . The greater the level of psychological obstructionism, the greater is the negative interpretation of the outlet’s servicescape environment. Influence of psychological obstructionism on the severity of dysfunctional customer behavior Within the sphere of criminology and indeed, the general psychological study of deviance, it is widely accepted that

325

aspects of psychological obstructionism play an important role in driving acts of criminal behaviors of ranging severities. To detail, evidence to support this association is forwarded in an individual and holistic manner. Regarding individual facets of psychological obstructionism; Machiavellianism, aggressiveness, sensation seeking, and consumer alienation are repeatedly considered significant drivers of ethically questionable behaviors (e.g., d’Acremont and Van der Linden 2005). Support for the broader link between psychological obstructionism and the severity of dysfunctional customer behavior is also evident within the deviance literature. For example, traits and predispositions pertaining to psychological obstructionism are argued to antecede a variety of criminal behaviors, employee misbehaviors, and unethical behaviors (Al-Rafee and Cronan 2006; Harris and Ogbonna 2006). Focusing explicitly on the drivers of customer deviance, Fullerton and Punj (1993), forward that individual personality traits and predispositions are crucial to understanding the antecedents of aberrant customer behavior. Personality traits and predispositions that are considered to obstruct normative behavior are also recognized as important within McGrath and Goulding’s (1996) contemplation of customer misbehavior within public service settings. Thus: H4 . The greater the level of psychological obstructionism, the greater is the severity of dysfunctional customer behavior. Influence of servicescape variables on the severity of dysfunctional customer behavior Research that offers insight into the relationship between the perceived design of servicescape and the severity of dysfunctional customer behavior is varied. To illustrate, Phillips, Alexander, and Shaw (2005) highlight a link between the physical design of self-service servicescapes and varying severities of customer theft. Focusing on more severe acts of misbehavior within inner cities, based on a review of past research, Kuo and Sullivan (2001) argue that the relationship between physical environments and aggressive behaviors is well established and widely accepted. Furthermore, in exploring this relationship, Homel and Clark (1994) find organizations that are perceived to be overcrowded, poorly ventilated, unclean, and noisy, experience higher rates of physical violence, than establishments that possess qualities of ‘good’ physical design. Comparable findings are offered by Graham et al. (1980) who forward evidence of a statistically significant relationship between poorly maintained, dirty, and unattractive service environments and incidents of customer aggression. Also assuming a holistic view of the servicescape, Rose and Neidermeyer (1999), stress that the manipulatable components of service outlets including levels of crowding, background music, ambient temperatures, and color schemes, may influence the severity of aggressive behaviors by consumers. Lawrence (2004) broadens this proposition in arguing that organizations should consider the initial design of servicescape environments at the planning stage of construction, thus avoiding the need to later ‘react’ to acts of misbehavior once the outlet has started to trade. In agreement, Dotter and Roebuck (1988) suggest that the

326

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335

physical design of organizational settings may induce an array of misbehaviors. Indeed, in examining episodes of vandalistic behavior, Allen and Greenberger (1978) propose that the environmental design of an organization plays a role in driving its own mutilation. Thus: H5 . The greater the negative interpretation of an outlet’s servicescape, the greater is the severity of dysfunctional customer behavior. In summary, the research model depicts both a direct and an indirect relationship between psychological obstructionism and the severity of dysfunctional customer behavior. In addition, it denotes an indirect relationship between servicescape variables and the severity of customer misbehavior mediated by disaffection with service. By contrast, the rival model advances direct relationships between all three constructs and severity. However, as indicated previously, we favor the composition of the research model and posit that it yields a significantly better fit with the data than the rival model. Method The hospitality industry is an ideal context for this study because of the sector’s economic importance and features, such as extended and close customer contact (Reynolds and Harris 2006). Specifically, we considered the bar, hotel, and restaurant sectors potentially fruitful contexts of inquiry. Indeed, several studies suggest that this industry is a particularly “potent” environment in which to study the dynamics of customer misbehavior (Harris and Reynolds 2003; Jones and Groenenboom 2002). A total of 1300 customers were approached in a public space (e.g., shopping malls) and asked a screening question to (i) ascertain their suitability (regarding, having deliberately behaved in a dysfunctional manner within a bar, hotel, or restaurant during the past three months), and (ii) so that the researcher could provide confidentiality assurances. Of the customers approached, 696 declined to participate and 220 indicated that they had misbehaved in the past but not within a hospitality-based outlet, or had misbehaved in a services setting but not in the role as a customer. Consequently, 384 questionnaires were completed (four of which were incomplete and removed from the sample). This yielded a response rate of nearly 30 percent. Of the respondents, 53.9 percent were female, the median age was 51 and the largest group comprised respondents who earned between $40,000 and $60,000 annually. Before completing the survey instrument, respondents were required to recall and describe an incident of dysfunctional behavior that they had undertaken. This enabled us to gain a better understanding of the episode and to record both the form and the severity of the behavior. This first stage of data collection assists in stimulating memory, helps respondents complete the questionnaire in a more focused frame of mind, and engenders trust (Podsakoff et al. 2003). Thereafter, respondents individually completed a structured questionnaire that focused on the single incident (see Appendix A).

Measure development The scales used were newly created or modified from existing scales. Following standard psychometric scale development procedures, first, we completed an extensive review of the literature to gain insight into the underlying dimensions of each construct. Second, we consulted 12 consumers, eight frontline employees, three service managers, and four academicians during in-depth interviews. In addition, we employed Q-sort procedures as a means to assess each measure, with a panel of 21 judges (10 consumers, three frontline employees, four service managers, and four academicians). Thereafter, we followed pretesting procedures that entailed two separate pilot studies of the research instrument. During the first pilot study (n = 50), particular attention was devoted to the phrasing of the severity measure and ensuring that only those cases of misbehavior that occurred within the service setting were eligible. The second pilot study (n = 66) trialed the refined measures. We examined the results for reliability and validity and found that they met the standard benchmark criteria. Measures of constructs In addition to demographic and control measures, we used eleven scales, seven of which we adapted from existing measures. We adopted seven-point Likert-type scoring for all items because seven-point scales increase the reliability of data findings. To measure the severity of dysfunctional customer behavior, we developed a four-item scale specifically for the study (see Appendix A). We designed this scale to gauge the extent to which the measured behavior violated the norms of the service outlet. Thus, we developed and used a four-item scale underpinned by the concept of norm violation. To assess the robustness of our dependent measure, we employed three additional measures. First, respondents were required to indicate which form of behavior they had perpetrated. Second, during screening and according to social norms, the researcher recorded (1) the severity of the behavior and (2) the form of behavior performed. Subsequent analysis revealed strong correlations between the four measures (p < .01), indicating a high degree of consistency in terms of perceptions of severity and form, both across the sample and between respondents and the researcher, and support for the self-reported severity scale employed in subsequent analyses. We gauged psychological obstructionism using four separate measures. We measured Machiavellianism using four items from Christie and Geis’s (1970) MACH IV scale. A refined version of Buss and Perry’s (1992) measure was employed to gauge aggressiveness. We assessed sensation seeking propensity using four items adapted from Steenkamp and Baumgartner (1992). We gauged consumer alienation using six items from Singh’s (1990) scale. We gauged disaffection with service using two scales. First, we assessed dissatisfaction with a four-item scale based on the themes and items in the studies of Bloemer and OdekerkenSchröder (2002) and Pizam and Ellis (1999). We used Oliver and

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335

Swan’s (1989) four-item scale of fairness to measure perceptions of inequity with the service. We gauged servicescape variables using four scales. To assess the atmospheric characteristics of the servicescape, we drew and refined four items from d’Astous’s (2000) classification of ambient variables. Using the scale development procedure outlined previously, we developed five-, five-, and six-item measures, respectively, to gauge consumers’ interpretations of the layout and design of a service outlet, an organization’s exterior environment, and the perceived behavior of fellow customers. We measured four control variables—sex, age, income, and level of intoxication—because each have been shown to influence dysfunctional customer behavior (Harris and Reynolds 2003). We also developed a five-item intoxication scale to measure the degree of alcohol and/or drug intoxication. Finally, we assessed social desirability using four items derived from the study of Reynolds (1982). Appendix A presents all the measures and their reliabilities. Scale assessment We used CFA to assess our measurement model. Other than the dependent construct severity and the multi-item control variable intoxication, we consider all factors in our theoretical model second-order constructs. Here, each first-order factor represents a reflective indicator for the higher-order construct. In theorizing each of the three higher-order constructs, we followed the criterion of Hair et al. (2006). Consequently, we conducted second-order CFA (see Table 1). Because of sample-size constraints in relation to the number of parameters to be estimated (Bentler and Chou 1987), we ran three separate CFA models that contained subsets of the most theoretically aligned variables. We then assessed each measurement model using the elliptical reweighed least squares estimation procedure. To evaluate the fit of the second-order psychological obstructionism construct, Measurement Model 1 (see Table 1) comprises Machiavellianism, aggressiveness, sensation seeking, and consumer alienation. Analysis of the fit indexes suggested good model fit (χ2 /df = 1.82, comparative fit index [CFI] = .98, nonnormed fit index [NNFI] = .98, and root mean square error of approximation [RMSEA] = .05). Measurement Model 2 evaluated the servicescape variables of layout and design, atmospherics, exterior environment, and fellow customer behavior. As Table 1 shows, this second-order CFA measurement model represents a satisfactory fit with the data (χ2 /df = 1.83, CFI = .99, NNFI = .99, and RMSEA = .05). Finally, Measurement Model 3 assessed situation-specific factors of disaffection with service, severity, and intoxication. We consider disaffection with service a second-order construct reflected by the first-order dimensions of dissatisfaction and inequity. We estimate the constructs of severity and intoxication as individual first-order constructs. The measure of disaffection with service breached the three-item rule and therefore is underidentified. To remedy this, an equality constraint is added to the disturbance terms, thus satisfying identification stipulations. As Table 1 shows, the results indicate an acceptable fit (χ2 /df = 1.98, CFI = .98, NNFI = .98, and RMSEA = .05). All three measurement models support our con-

327

ceptualization of psychological obstructionism, servicescape, and disaffection with service as higher-order factors and severity and intoxication as lower-order constructs. Indeed, in line with Hair et al. (2006), in comparing the higher-order structure with a lower-order factor model, the second-order structure shows superior predictive validity. In each measurement model, the CFA results indicate good psychometric properties for all constructs. All loadings and corresponding t-values at both the lower- and higher-order level were significant (t > 2.58), thus indicating convergent validity. Furthermore, we scrutinized Cronbach’s alphas, composite reliability (CR), and average variance extracted (AVE) for each scale. The lowest Cronbach alpha value was .85, and the lowest CR value was .79; all measures exceeded acceptable thresholds. Furthermore, each AVE exceeded Fornell and Larcker’s (1981) suggested minimum value of .50. To examine the reliability of each higher-order construct, we calculated Nunnally’s (1978) formula for the reliability of linear combinations. All three higher-order reliabilities exceeded the cutoff of .70. We used two separate forms of analysis to determine discriminant validity. First, we conducted a series of CFA tests in which we analyzed each possible pair of constructs. For every pair, we supply evidence of discriminant validity through a statistically significant chi-square difference between the constrained and unconstrained model. Second, we assessed discriminant validity using Fornell and Larcker’s (1981) test. We found that each construct’s AVE is greater than the squared correlation between the two constructs. We also compared the AVE of each second-order measure with the squared structural link with other constructs within the model. In all cases, the AVE was greater than the squared structural link, thus providing further evidence that all factors exhibit discriminant validity. Given the possibility of social desirability bias, we followed Podsakoff et al.’s (2003) recommendations. Using a latent measure, we assessed the effect of social desirability at both the measurement model and the structural model stage of analysis using a pair test approach. The results suggest that social desirability did not bias the data (see Graziano and Tobin 2002). In investigating if common method variance biases the data, Podsakoff et al. (2003) recommend conducting Harmann’s single factor test. This analysis was subsequently conducted and satisfied. Finally, in order to assess the scope of misbehaviors captured within the data set, the form perpetrated as indicated by the respondent was analyzed. As summarized in Table 2, the data represents a wide range of dysfunctional customer behaviors, both of unethical legal (36.7 percent) and unethical illegal (63.3 percent) orientation. Furthermore, employees, fellow customers and organizations were the most likely victims of the reported misdemeanors. Hypotheses testing We adopted a parsimonious approach to estimate our two competing structural models to satisfy the five-to-one stipulation of sample size to parameters (Bentler and Chou 1987). For each first-order factor, we used weighted composites, which we derived in part from the first-order scale’s alpha coefficient in

328

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335

Table 1 Measurement models. Measurement Model 1

Measurement Model 2

Psychological obstructionism First-order factors

Standardized

Machiavellianism MACH1 MACH2 MACH3 MACH4

.70b .83 (13.13) .85 (13.29) .71 (11.51)

Aggressiveness AGG1 AGG2 AGG3 AGG4 AGG5 AGG6

.70b .85 (14.63) .87 (15.01) .81 (15.01) .87 (14.99) .71 (12.38)

Sensation seeking SNS1 SNS2 SNS3 SNS4

.66b .81 (12.60) .85 (13.07) .89 (13.44)

Consumer alienation ATB1 ATB2 ATB3 ATB4 ATB5 ATB6

.83b .78 (16.42) .84 (18.17) .84 (18.42) .65 (12.93) .89 (19.92)

Servicescape variables loadingsa

a

Situation-specific variables

First-order factors

Standardized

Layout and design LAY1 LAY2 LAY3 LAY4 LAY5

.64b .77 (11.22) .82 (11.79) .90 (12.56) .91 (12.72)

Atmospherics ATM1 ATM2 ATM3 ATM4

.79b .88 (16.70) .87 (16.38) .83 (15.47)

Exterior environment EXT1 EXT2 EXT3 EXT4 EXT5

.90b .91 (24.12) .81 (18.68) .91 (24.34) .86 (21.24)

Fellow customers CUS1 CUS2 CUS3 CUS4 CUS5 CUS6

75b .81 (14.54) .91 (16.52) .88 (15.93) .92 (16.87) .89 (16.24)

Goodness-of-fit statistics χ2 /df = 1.82 CFI = .98 NNFI = .98 RMSEA = .05 b

Measurement Model 3

loadingsa

Goodness-of-fit statistics χ2 /df = 1.83 CFI = .99 NNFI = .99 RMSEA = .05

First-order factors

Standardized loadingsa

Severity of DCB SEV1 SEV2 SEV3 SEV4

.79b .84 (15.45) .85 (15.48) .89 (16.46)

Level of intoxication TOX1 TOX2 TOX3 TOX4 TOX5

.85b .90 (20.74) .97 (23.85) .68 (13.14) .86 (18.78)

Dissatisfaction DIS1 DIS2 DIS3 DIS4

.94b .96 (35.07) .92 (28.80) .89 (25.81)

Inequity INE1 INE2 INE3 INE4

88b .91 (21.21) .77 (15.86) .82 (17.34)

Goodness-of-fit statistics χ2 /df = 1.98 CFI = .98 NNFI = .98 RMSEA = .05

The t-values from the unstandardized solution are in parentheses. Fixed parameter.

Appendix A. Furthermore, each of the three second-order factors employed first-order composites as indicators. We present the results of the two rival models in Table 3. The results reflect those of trimmed models. That is, on initial analysis of each model, we found that the effects of two of the control variables (income and intoxication) were consistently not significant (p > .05). Subsequent removal of these two control factors improved the overall model fit statistics for both rival models.

tistical support for H2 (β = .15, t = 2.72, p < .01) suggests that the higher the level of disaffection with service, the greater is the severity of dysfunctional customer behavior. We also found support for H3 (β = .54, t = 5.58, p < .001), which focuses on the relationship between psychological obstructionism and the servicescape. Finally, the link between psychological obstructionism and customer misbehavior (β = .37, t = 4.66, p < .001) provides support for H4 . Thus, H1 –H4 are accepted.

Research model

Rival model

The fit statistics indicate that the research model provides a good fit with the data (χ2 /df = 3.02, CFI = .92, NNFI = .91, and RMSEA = .07) and supports the four hypotheses (H1 –H4 ). Table 3 documents the results for H1 (β = .49, t = 7.92, p < .001) and indicates a relationship between negative interpretations of the servicescape and customers’ evaluations of disaffection. Sta-

The goodness-of-fit statistics indicate that the rival model represents a poor fit with the data (χ2 /df = 4.71, CFI = .85, NNFI = .82, and RMSEA = .10; see Table 3). Consistent with our previous findings, analysis of the individual path coefficients in this model indicates support for both H2 (β = .25, t = 5.06, p < .001) and H4 (β = .50, t = 6.04, p < .001). However,

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335

329

Table 2 Illustrative forms of dysfunctional customer behavior reported. Illustrations of forms of dysfunctional customer behavior reported

Percentage of respondents reporting perpetrating such behavior

Indicative percentage of victims of dysfunctional customer behavior reporteda

1. Failing to tell an employee when a mistake had been made in the respondent’s favor

20

Employees = 42 Fellow customers = 10 Organization = 92

2. Complaining without genuine cause

13.2

Employees = 33 Fellow customers = 21 Organization = 84

3. Using/consuming the facilities of a service outlet without intending to pay

13.4

Employees = 47 Fellow customers = 32 Organization = 90

4. Knowingly stealing an item from a service outlet

16.3

Employees = 63 Fellow customers = 36 Organization = 100

5. Arguing with, or being openly rude to a service employee or fellow customer

17.4

Employees = 86 Fellow customers = 44 Organization = 65

6. Knowingly damaging or vandalizing a service outlet’s property

12.4

Employees = 59 Fellow customers = 52 Organization = 100

7. Physically touching/striking a service employee or fellow customer

7.4

Employees = 90 Fellow customers = 43 Organization = 61

Note. Respondents were requested to indicate the form of misbehavior that most closely characterized their performed behavior. Consequently, this table constitutes a reflection of these behaviors, rather than an absolute and exact description of the individual misbehaviors perpetrated. a In many cases the misbehavior performed resulted in perceived consequences for more than one ‘victim’, this is reflected within the statistics shown.

we find no support for H5 (β = .07, t = 1.52, p > .05), which suggests a direct relationship between the servicescape and the severity of dysfunctional customer behavior. Therefore, we reject H5 . Model comparison To compare the goodness-of-fit with the data between the two structural models, in addition to reviewing the standardized coefficients for each of the five hypothesized paths, we drew on the Akaike information criterion fit index. In reviewing the Akaike information criterion statistic across the two models (see Table 3), the research model is favored, thus confirming that it represents the best fit with the data. The chi-square difference statistic is also commonly used to assess rival models that are hierarchical in nature. As is depicted in Table 3, the difference in the chi-square value between the two models is greater than 3.84, thus exhibiting statistical significance and favoring the research model. Furthermore, in order to assess the credence of the espoused mediated relationship between servicescape and severity through disaffection, a mediation analysis was conducted. The results revealed support for the mediated relationship with a statistically nonsignificant chi-square difference (p > .05) and non-significant t-value between servicescape and severity, as depicted within the rival model.

Discussion The aim of this research was to conceptually develop and empirically test a framework of the factors that associate with dysfunctional customer behavior severity. In synthesizing literature from disparate research domains and advancing survey-derived results pertaining to our two competing theoretical models, we offer empirical insight into these issues and provide significant implications for marketing academicians and practitioners.

Theoretical contributions This study makes four main contributions. First, by garnering and synthesizing literature from diverse academies and perspectives, our study contributes to existing knowledge by highlighting the pivotal role of three core concepts—psychological obstructionism, disaffection with service, and servicescape variables—as the primary factors associated with dysfunctional customer behavior severity. Although previous studies have tended to include at least one of these dimensions, to date, empirically based and holistic analyses have been lacking. Our review of the literature uncovers diverse theoretical traditions that differ radically in their modeling of the dynamics between these antecedents. Critical analysis leads us to advance two distinct models founded within diverse conceptual academies. In

330

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335

Table 3 Structural model results. Hypothesized paths

H1 : Servicescape → disaffection H2 : Disaffection → severity of DCB H3 : Psychological obstructionism → servicescape H4 : Psychological obstructionism → severity of DCB H5 : Servicescape → severity of DCB Goodness-of-fit statistics χ2 df χ2 /df Probability CFI NNFI RMSEA Akaike information criterion

Research model

Rival model

β (S.E.)

t-value

β (S.E.)

t-value

.49 .15

(7.92) (2.72)

– .25

(5.06)

.54

(5.58)

–

.37

(4.66)

.50

(6.04)

–

.07

(1.52)

190.50 63 3.02 .001

301.75 64 4.71 .001 .92 .91 .07 64.50

.85 .82 .10 173.75

Note. DCB: dysfunctional customer behavior.

this sense, each of the two competing perspectives provides theoretical insight into the dynamics of dysfunctional customer behavior severity. Data analysis indicates that our preferred model constitutes a significantly better fit with the data and leads to the rejection of the less robust rival model. This leads us to question the orthodox apparent within research domains that stay firmly and narrowly focused within their own literature base. If we want to generate novel insights into phenomena that have a rich but diverse research traditions, such as dysfunctional customer behavior, it is both prudent and enriching to delve into such literature and immerse ourselves in the varied perspectives and positions that exist. Thus, we believe that our amalgamated conceptual approach can be fruitfully applied to other areas of research interest within marketing. Second, by undertaking the first holistic, survey-based study of the factors associated with dysfunctional customer behavior, our research also makes an empirical contribution. In this regard, our study was motivated in part by a desire to respond to the plethora of calls for empirical research into these issues (see Bitner, Booms, and Mohr 1994; Fullerton and Punj 2004; Harris and Reynolds 2003; Wirtz and Kum 2004). Moreover, we make a contribution through the study of actual incidents of customer misbehavior. Previous research is weakened by an overemphasis on the employment of experimentation techniques to study artificial scenarios, typically in relation to a single determining factor and a single form of customer misbehavior (e.g., propensity to shoplift). Although the study of hypothetical situations can garner useful insights, our view is that such research is complemented by concurrent studies that focus on real people in real situations. Paper-based or computer-generated contexts are unable to replicate perfectly a multiplicity of crucial, determining factors (e.g., the ambient conditions of a service setting) that are associated with customer misbehavior. In addi-

tion, the research design we employed contributes insights into the real behavior of customers, which laboratory-based studies are ethically and morally constrained from inducing (e.g., acts of violence). The supply of grounded data regarding customer misbehavior constitutes an important step in developing a greater understanding of the darker, less salubrious side of service encounters and highlights the need for a broader perspective on customer dynamics that extends beyond the currently dominant (albeit understandable) emphasis on managerially prescriptive issues. Third, our conception of dysfunctional customer behavior posits that such actions are centered on societal, cultural, and contextual norm breaking. Our research reveals multiple factors across diverse customer behaviors that vary considerably in severity. This suggests that the broadening of emphasis to norm-breaking issues is likely to be both more insightful and more generalizable than narrow forms. In this regard, normbreaking issues are important. We are not suggesting that people prone to less severe forms of dysfunctional customer behavior are equally prone to severe acts. However, our findings indicate that the antecedents of multiple forms of customer misbehavior are commonly shared (albeit at differing degrees). Thus, people with extreme psychological obstructionism are associated with extreme forms of dysfunctional customer behavior, particularly when faced with extreme contexts and situations. Our findings strongly indicate that dysfunctional customer behavior severity cannot be reliably attributed to a single stimuli, but rather is triggered by individual, situational, and contextual factors that amalgamate to elicit episodes of misbehavior (for a similar conception, see Fullerton and Punj 1993). Fourth, our study offers methodological contributions by developing, operationalizing, and testing several new scales. In particular, the development of a novel, robust, multi-item measure of the severity of dysfunctional customer behavior is worthy of comment. Although researchers have previously alluded to the notion of severity in conceptual treaties (e.g., Robinson and Bennett 1997; Vitell and Muncy 1992), to date, research that focuses on deviant customer behavior has overly and narrowly concentrated on, individual forms of misbehavior. In addition, although other studies have previously emphasized the severity of customers’ misbehavior, they did not attempt to operationalize this construct. We contend that to advance our comprehension of incidents of dysfunctional customer behavior, new scale development is essential. In developing, evaluating, and validating a measure, we provide the methodological tools for researchers in this area to progress and extend the understanding of customer deviance dynamics. Managerial implications Although many practitioner-oriented commentaries dismiss incidents of dysfunctional customer acts as random, irrational events that are endemic to some services contexts, the findings of this study refute such notions by highlighting that some factors are subject to managerial control and manipulation. In this regard, the impact of customer evaluations of disaffection and servicescapes is especially important to consider. First,

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335

through the design and monitoring of the quality of service provision, customer service, and complaint and feedback structures, practitioners can actively manage customers’ evaluations of dissatisfaction and inequity. Indeed, our study provides evidence to suggest that rather than being unreasonable and illogical (at least from the consumer’s perspective); the seemingly “rational” motive of disaffection with service directly affects the severity of dysfunctional customer behavior performed. Second, while the findings of our study do not support a direct relationship between servicescape and the severity of dysfunctional customer behavior, an indirect relationship, mediated by disaffection with service is championed. Further, psychological obstructionism is shown to have an indirect effect on severity through servicescape and disaffection with service. Given these results, through the careful design or redesign of servicescape, managers should be able to reduce the severity of dysfunctional customer behavior by creating environments that are satisfactory and exhibit a degree of ‘fit’ with patrons. While potentially fraught with difficulties owing to diverse consumer segments possessing varied levels of psychological obstructionist traits, this suggests that those charged with service environment design should contemplate the target audience of the servicescape with deviance in mind during embryonic design stages. Furthermore, the conceivable compatibility between patrons should also be acknowledged. Indeed, ensuring a degree of congruency between consumers, who owing to the nature of service provision may have to spend extended periods in close proximity to one another, may reduce incidents and the severity of dysfunctional customer behavior brought on by inter-client conflict and subsequent disaffection. In this sense, although dysfunctional customer behaviors are unlikely to cease entirely, a key managerial implication of this study is that persons responsible for the physical design of service environments and customer care strategies are far from powerless in the proactive management of such behaviors. The finding of a direct link between psychological obstructionism and dysfunctional customer behavior severity also indicates interesting and important implications for practice. While our finding that enduring traits are linked to the severity of dysfunctional customer behavior may be interpreted by managers as unhelpful (as such factors are not subject to managerial control), such an interpretation is imprudent. Although the findings of this study suggest that dysfunctional customer behavior cannot be completely eradicated by judicious servicescape design and improvements to service standards, an understanding of the psychological factors linked to customer misbehavior can provide insights into the reduction of such events. First, managers can develop training schemes to improve the abilities of frontline employees in recognizing obstructionist traits and managing their subsequent interactions with consumers exhibiting levels of psychological obstructionism. In this regard, induction and training efforts should stress that, while consumers with such traits may be difficult to recognize, identifying potential offenders and adjusting interaction styles and surveillance practices may reduce (but not eliminate) incidents. Through educating customer-contact employees about the prevalence and triggers of customer deviance, firms can better position them to

331

prevent problem customers from disrupting service encounters, potentially saving firms considerable costs in time and money. Second, in the vein of Berry and Seiders (2008), managers might consider ‘firing’ reoffending misbehaving customers exhibiting obstructionist tendencies, in order to prevent them becoming ingrained within the framework of the servicescape and consequently having a negative effect on other patrons. Thus, although the complete eradiation of customer deviance seems unfeasible, tactical maneuvers are likely to reduce the severity of the misbehavior. For executives and senior managers, this study highlights the merit of a strategic approach to the phenomenon of dysfunctional customer behavior. Given an increased understanding of the factors linked to customer misbehavior, managers should be able to shape and develop systems, structures, and design priorities calculated to monitor, minimize, and manage misbehavior. For example, customers’ evaluations of dissatisfaction and inequity can be reduced through the implementation of effective service delivery, service failure, and customer feedback mechanisms. Furthermore, the development of databases to record, track, and scrutinize such incidents should allow managers to analyze patterns, trends, and the frequencies of different forms of deviant acts. The insights gained should feed into company procedures and policies, as well as both redesigned systems and servicescapes. Limitations and avenues for further research The findings and contributions of our study are bounded by limitations that, in turn, highlight potentially fruitful avenues for further research. In particular, four limitations are especially worthy of further discussion. First, the context of our study limits the extent to which we can universally generalize the results and implications. Although we deemed the hospitality industry as an appropriate setting, its idiosyncrasies (e.g., extended customer contact) are far from universal. Thus, research should explore dysfunctional customer behavior in different and contrasting contexts (varying servicescapes being a potentially fruitful avenue for future studies). We believe that such research could build on the conceptions and measures we employed herein and not only gauge the reliability and validity of the developed measure of the severity of dysfunctional customer behavior but also further explore the critical role of customer disaffection. Second, as the first study to conceptualize and then empirically test a range of factors associated with the severity of dysfunctional customer behavior, our focus was on the principal linkages. However, it would be naive to claim that these factors constitute an exhaustive list. Therefore, further research could identify additional variables and extend understanding of the dynamics between such factors. For example, future research is needed to develop an understanding of the motives for such behaviors and to explore further, how third parties influence these behaviors. Third, discussions with practitioners revealed a widespread assumption that the level of intoxication contributes to customer misbehavior. However, in our study, we found no statistically

332

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335

significant link. Research should explore this issue further. For example, does intoxication link to the frequency rather than the severity of dysfunctional customer behaviors? Are certain personality types more inclined to become intoxicated than others? Fourth, our approach focuses on the retrospective analysis of customer interpretations of real events. As such, similar to most survey-based approaches, we base our study on the assumption that consumers truthfully and accurately recall events. Further, our focus has been on actual behavior rather than future intentions or motives. Further research should explore the demographic and psychographic factors linked to both behavioral intentions to misbehave as well as the motives for such acts. Appendix A. Construct and measurement items Severity of dysfunctional customer behaviora (α = .91, CR = .84, AVE = .57) 1. (SEV1) If others had witnessed my behavior, they would have thought it was inappropriate behavior within that specific outlet. 2. (SEV2) In hindsight, I acknowledge that my behavior is not what is expected of customers within that service outlet. 3. (SEV3) I believe that others would generally view my behavior as acceptable in today’s society. (reverse scored) 4. (SEV4) If others had witnessed my behavior, they would have thought it was acceptable behavior within that specific outlet. (reverse scored) Psychological obstructionism

3. (SNS3) I would like to try an ‘extreme’ sport such as bungee jumping. 4. (SNS4) I like to have new and exciting experiences and sensations even if they are a little frightening, unconventional, or illegal. Consumer alienationa (α = .92, CR = .87, AVE = .54) 1. (ATB1) In general, the customer is usually the least important consideration to most companies. 2. (ATB2) In general, shopping is usually an unpleasant experience. 3. (ATB3) In general, people must be willing to tolerate poor service from most businesses. 4. (ATB4) In general, companies are dishonest in their dealings with customers. 5. (ATB5) In general, businesses who offer product and service guarantees will honor them. (reverse scored) 6. (ATB6) In general, most companies care nothing about the customer. Disaffection with service Dissatisfactiona (α = .96, CR = .90, AVE = .70) 1. (DIS1) I was dissatisfied with the level of service that I received from the outlet. 2. (DIS2) My expectations were not met. 3. (DIS3) I was dissatisfied with the quality of service that I received. 4. (DIS4) I was very satisfied with the outlet. (reverse scored)

Machiavellianisma (α = .85, CR = .79, AVE = .52) Inequitya (α = .92, CR = .85, AVE = .59) 1. (MACH1) Honesty is always the best policy. (reverse scored) 2. (MACH2) The majority of people are basically good and kind. (reverse scored) 3. (MACH3) Most people who get ahead in the world lead good and honest lives. (reverse scored) 4. (MACH4) A white lie is often a good thing. Aggressivenessa (α = .92, CR = .88, AVE = .56) 1. (AGG1) Given enough provocation, I may hit another person. 2. (AGG2) I rarely find myself disagreeing with other people. (reverse scored) 3. (AGG3) When people annoy me, I tell them what I think. 4. (AGG4) When frustrated, I let my irritation show. 5. (AGG5) Some of my friends think that I am hot-headed. 6. (AGG6) When people are especially nice, I wonder what they want.

1. 2. 3. 4.

(INE1) The outlet treated me fairly. (reverse scored) (INE2) I was not treated right by the outlet. (INE3) I felt that the outlet was taking advantage of me. (INE4) I felt that the outlet behaved in an unfair way towards me.

Servicescape variables Layout and designa (α = .91, CR = .85, AVE = .55) 1. (LAY1) The interior of the outlet was designed to my taste. (reverse scored) 2. (LAY2) It was very crowded inside of the outlet. 3. (LAY3) The interior design of the outlet was unpleasant. 4. (LAY4) It was very cramped inside of the outlet. 5. (LAY5) It was easy to move around the outlet. (reverse scored)

Sensation seekinga (α = .88, CR = .82, AVE = .54) Atmosphericsa (α = .91, CR = .84, AVE = .57) 1. (SNS1) I do not like to try new foods that I have never tasted before. (reverse scored) 2. (SNS2) I prefer friends who are exciting and unpredictable.

1. (ATM1) The temperature inside of the outlet was pleasant. (reverse scored)

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335

2. (ATM2) The music inside of the outlet was too loud. 3. (ATM3) The air quality inside of the outlet was poor. 4. (ATM4) The outlet was very clean. (reverse scored) Exterior environmenta (α = .94, CR = .89, AVE = .62) 1. (EXT1) The exterior of the outlet was unappealing. 2. (EXT2) The outlet was located in a nice area. (reverse scored) 3. (EXT3) The outside of the outlet did not look well maintained. 4. (EXT4) The exterior of the outlet looked run down. 5. (EXT5) The exterior of the outlet looked attractive. (reverse scored) Fellow customersa (α = .94, CR = .90, AVE = .60) 1. (CUS1) Fellow customers behaved in a pleasant manner. (reverse scored) 2. (CUS2) Fellow customers behaved in a way that I was not expecting. 3. (CUS3) I enjoyed being around the other customers in the outlet. (reverse scored) 4. (CUS4) Fellow customers conducted themselves in a manner that I did not find appropriate. 5. (CUS5) Fellow customers behaved in a way that I found to be unpleasant. 6. (CUS6) Fellow customers behaved in a way that I did not agree with. Multi-item control variables Intoxicationa (α = .93, CR = .88, AVE = .60) 1. (TOX1) I had consumed an intoxicating substance prior to visiting the outlet. 2. (TOX2) I believe that I was under the influence at the time of the incident. 3. (TOX3) I consider myself to have been intoxicated at the time of the incident. 4. (TOX4) I consumed an intoxicating substance during my time in the outlet. 5. (TOX5) I was not intoxicated at the time of the incident. (reverse scored) Social desirabilitya (α = .89, CR = .82, AVE = .54) 1. (SD1) It is sometimes hard for me to go on with my work if I am not encouraged. 2. (SD2) There have been times when I was quite jealous of the good fortune of others. 3. (SD3) No matter who I am talking to, I am always a good listener. 4. (SD4) I am sometimes irritated by people who ask favors of me. (reverse scored)

333

a Seven-point scale (1 = “strongly disagree,” and 7 = “strongly

agree”).

References Allen, Vernon L. and David B. Greenberger (1978), “An Aesthetic Theory of Vandalism,” Crime and Delinquency, 24 (2), 309–32. Al-Rafee, Sulaiman and Timothy Paul Cronan (2006), “Digital Piracy: Factors that Influence Attitude Towards Behavior,” Journal of Business Ethics, 63 (3), 237–59. Areni, Charles S. (2003), “Exploring Managers’ Implicit Theories of Atmospheric Music: Comparing Academic Analysis to Industry Insight,” Journal of Services Marketing, 17 (2), 161–84. Aylott, Russell and Vincent-Wayne Mitchell (1998), “An Exploratory Study of Grocery Shopping Stressors,” International Journal of Retail and Distribution Management, 26 (9), 362–73. Bäckström, Kristina and Ulf Johansson (2006), “Creating and Consuming Experiences in Retail Store Environments: Comparing Retailer and Consumer Perspectives,” Journal of Retailing and Consumer Services, 13 (6), 417–30. Baker, Julie and Michaelle Cameron (1996), “The Effects of the Service Environment on Affect and Consumer Perception of Waiting Time: An Integrative Review and Research Propositions,” Journal of the Academy of Marketing Science, 24 (4), 338–49. Bamfield, Joshua (2006), European Retail Theft Barometer: Monitoring the Costs of Shrinkage and Crime for Europe’s Retailers. Nottingham, UK: Centre for Retail Research, [www.retailresearch.org]. Barnes, James G., Brian R. King and Gordon A. Breen (2004), “The Almost Customer: A Missed Opportunity to Enhance Corporate Success,” Managing Service Quality, 14 (2–3), 134–46. Bentler, Peter M. and Chih-Ping Chou (1987), “Practical Issues in Structural Modeling,” Sociological Methods and Research, 16 (1), 78–117. Berry, Leonard L. and Kathleen Seiders (2008), “Serving Unfair Customers,” Business Horizons, 51, 29–37. Bitner, Mary Jo (1992), “Servicescapes: The Impact of Physical Surroundings on Customers and Employees,” Journal of Marketing, 2 (April), 57– 71. Bitner, Mary Jo, Bernard H. Booms and Lois A. Mohr (1994), “Critical Service Encounters: The Employee’s Viewpoint,” Journal of Marketing, 58 (4), 95–106. Bloemer, Josée and Gaby Odekerken-Schröder (2002), “Store Satisfaction and Store Loyalty Explained by Customer- and Store-Related Factors,” Journal of Consumer Satisfaction, Dissatisfaction and Complaining Behavior, 15, 68–80. Buss, Arnold H. and Mark Perry (1992), “The Aggression Questionnaire,” Journal of Personality and Social Psychology, 63 (3), 452–9. Christie, Richard and Florence L. Geis (1970), “Studies in Machiavellianism,” London: Academic Press. d’Acremont, Mathieu and Martial Van der Linden (2005), “Adolescent Impulsivity: Findings from a Community Sample,” Journal of Youth and Adolescence, 34 (5), 427–35. d’Astous, Alain (2000), “Irritating Aspects of the Shopping Environment,” Journal of Business Research, 49 (2), 149–56. Diamond, Michael A. (1997), “Administrative Assault: A Contemporary Psychoanalytic View of Violence and Aggression in the Workplace,” American Review of Public Administrative, 27 (3), 228–47. Dotter, Daniel L. and Julian B. Roebuck (1988), “The Labeling Approach Re-Examined: Interactionism and the Components of Deviance,” Deviant Behavior, 9 (1), 19–32. Dube, Jonathan (2003), “Office Wars: Clients Attack Employees When They Feel Wronged”, ABC News (April 16). Eysenck, Hans J. (1964), “Crime and Personality,” London: Routledge and Kegan Paul. Eysenck, Sybil B.G. and Hans J. Eysenck (1970), “Crime and Personality: An Empirical Study of the Three-Factor Theory,” British Journal of Criminology, 10 (3), 225–39.

334

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335

Fornell, Claes and David F. Larcker (1981), “Evaluating Structural Equation Models with Unobservable Variables and Measurement Error,” Journal of Marketing Research, 18 (February), 39–50. Fox, Suzy and Paul E. Spector (1999), “A Model of Work Frustration: Aggression,” Journal of Organizational Behavior, 20 (6), 915–31. Fukukawa, Kyoko (2002), “Developing a Framework for Ethically Questionable Behavior in Consumption,” Journal of Business Ethics, 41 (1–2), 99–119. Fullerton, Ronald A. and Girish Punj (1993), “Choosing to Misbehave: A Structural Model of Aberrant Consumer Behavior,” Advances in Consumer Research, 20, 570–4. and (2004), “Repercussions of Promoting an Ideology of Consumption: Consumer Misbehavior,” Journal of Business Research, 57 (11), 1239–4. Gottfredson, Michael R. and Travis Hirschi (1990), “A General Theory of Crime,” Stanford, CT: Stanford University Press. Graham, Kathryn, Linda L. LaRocque, Rhoda R. Yetman, T. James Ross and Encico Guistra (1980), “Aggression and Barroom Environments,” Journal of Studies on Alcohol, 41 (3), 277–92. Grandey, Alicia A., David N. Dickter and Hock-Pen Sin (2004), “The Customer Is ‘Not’ Always Right: Customer Aggression and Emotion Regulation of Service Employees,” Journal of Organizational Behavior, 25 (3), 397–418. Graziano, William G. and Renée M. Tobin (2002), “Agreeableness: Dimension of Personality or Social Desirability Bias?,” Journal of Personality, 70 (5), 695–728. Grove, Stephen J., Raymond P. Fisk and Jacoby John (2004), “Surviving in the Age of Rage,” Marketing Management, 13 (2), 41–6. Hair, Joseph F., William C. Black, Barry J. Babin, Rolph E. Anderson and Ronald L. Tatham (2006), “Multivariate Data Analysis,” 6th ed. Upper Saddle River, NJ: Prentice Hall. Harris, Lloyd C. (2008), “Fraudulent Return Proclivity: An Empirical Analysis,” Journal of Retailing, 84 (4), 461–76. Harris, Lloyd C. and Emmanuel Ogbonna (2006), “Service Sabotage: A Study of Antecedents and Consequences,” Journal of the Academy of Marketing Science, 34 (4), 543–58. Harris, Lloyd C. and Kate L. Reynolds (2003), “The Consequences of Dysfunctional Customer Behavior,” Journal of Service Research, 6 (2), 144–61. and (2004), “Jaycustomer Behavior: An Exploration of Types and Motives in the Hospitality Industry,” Journal of Services Marketing, 18 (5), 339–57. Homel, Ross and Jeff Clark (1994), “The Prediction and Prevention of Violence in Pubs and Clubs,” In Crime Prevention Studies, vol. 3, Clarke Ronald V. ed. New York: Criminal Justice Press Hopkins, Matt (2002), “Crimes Against Businesses: The Way Forward for Future Research,” British Journal of Criminology, 42 (4), 782–97. Huefner, Jonathan C. and H. Keith Hunt (2000), “Consumer Retaliation as a Response to Dissatisfaction,” Journal of Consumer Satisfaction, Dissatisfaction, and Complaining Behavior, 13, 61–82. Jones, Peter and Karen Groenenboom (2002), “Crime in London Hotels,” Tourism and Hospitality Research, 4 (1), 21–35. Kallis, M. and Dinoo J. Vanier (1985), “Consumer Shoplifting: Orientations and Deterrents,” Journal of Criminal Justice, 13 (5), 459–73. Kenrick, Douglas T., Heather E. McCreath, Robert King and Jeffrey Bordin (1990), “Person-Environment Intersections,” Journal of Personality and Social Psychology, 58 (4), 685–98. Kuo, Frances E. and William C. Sullivan (2001), “Aggression and Violence in the Inner City: Effects of Environment via Mental Fatigue,” Environment and Behavior, 33 (4), 543–71. Lawrence, Greg (2004), “Designing Out Crime: The Retail Perspective,” International Journal of Retail and Distribution Management, 32 (12), 572–6. Lawrence, Thomas B. and Sandra L. Robinson (2007), “Ain’t Misbehavin: Workplace Deviance as Organizational Resistance,” Journal of Management, 33 (3), 378–94. Lee, Kibeom and Natalie J. Allen (2002), “Organizational Citizenship Behavior and Workplace Deviance: The Role of Affect and Cognitions,” Journal of Applied Psychology, 87 (1), 131–42.

Lin, Ingrid Y. (2004), “Evaluating a Servicescape: The Effect of Cognition and Emotion,” Hospitality Management, 23 (2), 163–78. Lindsay, James J. and Craig A. Anderson (2000), “From Antecedent Conditions to Violent Actions: A General Affective Aggression Model,” Personality and Social Psychology Bulletin, 26 (5), 533–47. Lovelock, Christopher H. (1994), “Product Plus: How Product and Service = Competitive Advantage,” New York: McGraw-Hill. McGrath, Hannah and Anne Goulding (1996), “Part of the Job: Violence in Public Libraries,” New Library World, 97 (1127), 4–13. Mehrabian, Albert and James A. Russell (1974), “An Approach to Environment Psychology,” Cambridge, MA: MIT Press. Mills, Michael K. (1981), “Deviance and Dissatisfaction: An Exploration Study,” Advances in Consumer Research, 8, 682–6. Morin, Sylvie, Laurette Dubé and Jean-Charles Chebat (2007), “The Role of Pleasant Music in Servicescapes: A Test of the Dual Model of Environment Perception,” Journal of Retailing, 83 (1), 115–30. Mudrack, Peter E. (1993), “An Investigation into the Acceptability of Workplace Behaviors of Dubious Ethical Nature,” Journal of Business Ethics, 12 (7), 517–24. Nunnally, Jum C. (1978), “Psychometric Theory,” 2nd ed. New York: McGrawHill. Oliver, Richard L. and John E. Swan (1989), “Consumer Perceptions of Interpersonal Equity and Satisfaction in Transactions: A Field Survey Approach,” Journal of Marketing, 53 (April), 21–35. Phillips, Simon, Andrew Alexander and Gareth Shaw (2005), “Consumer Misbehavior: The Rise of Self-Service Grocery Retailing and Shoplifting in the UK,” Journal of Macromarketing, 25 (1), 66–75. Pizam, Abraham and Taylor Ellis (1999), “Customer Satisfaction and its Measurement in Hospitality Enterprises,” International Journal of Contemporary Hospitality Management, 11 (7), 326–39. Podsakoff, Philip M., Scott B. Mackenzie, Jeong-Yeon Lee and Nathan P. Podsakoff (2003), “Common Method Biases in Behavioral Research: A Critical Review of the Literature and Recommended Remedies,” Journal of Applied Psychology, 88 (5), 879–903. Rallapalli, Kumar C., Scott J. Vitell, Frank Wiebe and James Barnes (1994), “Consumer Ethical Beliefs and Personality Traits: An Exploratory Analysis,” Journal of Business Ethics, 13 (7), 487–95. Rayburn, J. Michael and L. Gayle Rayburn (1996), “Relationship between Machiavellianism and Type A Personality and Ethical-Orientation,” Journal of Business Ethics, 15 (11), 1209–1. Reynolds, Kate L. and Lloyd C. Harris (2005), “When Service Failure is Not Service Failure: An Exploration of the Forms and Motives of “Illegitimate” Customer Complaining,” Journal of Services Marketing, 19 (5), 321–35. and (2006), “Deviant Customer Behavior: An Exploration of Frontline Employee Tactics,” Journal of Marketing Theory and Practice, 14 (2), 95–111. Reynolds, William M. (1982), “Development of Reliable and Valid Short Forms of the Marlowe-Crowne Social Desirability Scale,” Journal of Clinical Psychology, 38 (1), 119–25. Ringberg, Torsten, Gaby Odekerken-Schröder and Christensen (2007), “A Cultural Models Approach to Service Recovery,” Journal of Marketing, 71 (July), 194–21. Robinson, Sandra and Rebecca J. Bennett (1997), “Workplace Deviance: Its Definitions, Its Manifestations and Its Causes,” Research on Negotiation in Organizations, 6, 3–27. Robinson, Sandra L. and Anne M. O’Leary-Kelly (1998), “The Influence of Work Groups on the Antisocial Behavior of Employees,” Academy of Management Journal, 41 (6), 658–72. Romero, Estrella M., Angeles Luengo and Jorge Sobral (2001), “Personality and Antisocial Behaviour: Study of Temperamental Dimensions,” Personality and Individual Differences, 31 (3), 329–48. Rose, Randall L. and Mandy Neidermeyer (1999), “From Rudeness to Roadrage: The Antecedents and Consequences of Consumer Aggression,” Advances in Consumer Research, 26, 12–7. Rosenbaum, Mark S. and Ronald Kuntze (2003), “The Relationship Between Anomie and Unethical Retail Disposition,” Psychology and Marketing, 20 (12), 1067–93.

K.L. Reynolds, L.C. Harris / Journal of Retailing 85 (3, 2009) 321–335 Ross, William T. and Diana C. Robertson (2003), “A Typology of Situational Factors: Impact on Salesperson Decision-Making about Ethical Issues,” Journal of Business Ethics, 46 (3), 213–34. Schmidt, Ruth A. and Roger Sapsford (1995), “Issues of Gender and Servicescape,” International Journal of Retailing and Distribution Management, 23 (3), 34–40. Schneider, Benjamin (1987), “The People Make the Place,” Personnel Psychology, 40 (1), 437–54. Singh, Jagdip (1990), “A Typology of Consumer Dissatisfaction Response Styles,” Journal of Retailing, 66 (1), 57–99. Slater, Michael D. (2003), “Alienation, Aggression, and Sensation Seeking as Predictors of Adolescent Use of Violent Film, Computer and Website Content,” Journal of Communication, 53 (1), 105–21. Steenkamp, Jan-Benedict E.M. and Hans Baumgartner (1992), “The Role of Optimum Stimulation Level in Exploratory Consumer Behavior,” Journal of Consumer Research, 19 (December), 434–48. Sutherland, Ian and Jonathan P. Shepherd (2002), “A Personality-Based Model of Adolescent Violence,” British Journal of Criminology, 42 (2), 433–41.

335

USDAW (2004), Life on the Frontline: A Report on Shopworker’s Experience of Work-Related Violence and Abuse, (June), Manchester, UK: Union of Shop, Distributive and Allied Workers. Vitell, Scott J. and James Muncy (1992), “Consumer Ethics: An Empirical Investigation of Factors Influencing Ethical Judgments of the Final Customer,” Journal of Business Ethics, 11 (8), 585–97. Williams, Russell and Miriam Dargel (2004), “From Servicescape to ‘Cyberscape’,” Marketing Intelligence and Planning, 22 (3), 310–2. Wilson, James Q. and George Kelling (1982), “Broken Windows: The Police and Neighborhood Safety,” Atlantic Monthly, 249 (3), 29–38. Wirtz, Jochen and Doreen Kum (2004), “Consumer Cheating on Service Guarantees,” Journal of the Academy of Marketing Science, 32 (2), 112–26. Yi, Youjae and Taeshik Gong (2008), “The Effects of Customer Justice Perception and Affect on Customer Citizenship Behavior and Customer Dysfunctional Behavior,” Industrial Marketing Management, 37 (7), 767–83.