The Psychology of Shoplifting: Development of a New Typology for Repeated Shoplifting

Abstract

Despite the damaging effects of shoplifting on individuals, the current literature offers little guidance for changing shoplifting behavior. One limitation in this area of research has been the failure to use empirically and theoretically sound methodologies to identify individuals’ diverse characteristics and motivations. The present study addressed these limitations by developing an empirically and theoretically supported typology of the varied individuals who shoplift. Participants included 202 community individuals who reported repeated shoplifting and provided information about their shoplifting behavior, motivations, mental health, ethical attitudes, personal histories, and life circumstances. Cluster analyses revealed that the sample could be divided into six discrete groups. These clusters comprise a typology of shoplifting, including Loss-Reactive (28% of the sample), Impulsive (20%), Depressed (18%), Hobbyist (18%), Addictive–Compulsive (9%), and Economically Disadvantaged (7%) types. Each type comprises a unique pattern of shoplifting with unique needs. This research establishes a promising foundation for treating the diverse individuals who shoplift.

Keywords

shoplifting theft treatment typology psychology

The wide prevalence and consequences of shoplifting, defined as wrongfully taking merchandise from a place of business, have been well described in past literature. Negative consequences have been noted for individuals who shoplift as well as businesses, other shoppers, and the economy at large (American Psychiatric Association [APA], 2013; Cupchik, 2002; Forney, Crutsinger, & Forney, 2006; Geyer, 2000; National Retail Federation, 2016). For many people who shoplift, the behavior causes considerable distress.

Using data from more than 43,000 individuals, research from Blanco et al. (2008) revealed that people who had shoplifted before were significantly more likely to have a lifetime Diagnostic and Statistical Manual of Mental Disorders (DSM) axis I disorder (87.14% vs. 47.09%), personality disorder, nicotine dependence, and alcohol use disorder. They also scored lower on social, emotional, and mental health scales and were twice as likely to have sought psychiatric treatment. Individuals who had shoplifted before were five times more likely to have had any psychiatric hospitalization or emergency room visit than their nonshoplifting counterparts.

Other research has found shoplifting is associated with further psychiatric disorders and troublesome behaviors. These include borderline personality symptomatology (Sansone, Lam, & Wiederman, 2011), impulsivity (Sarasalo, Bergman, & Toth, 1997), low self-concept, guilt, loneliness, and shame (Krasnovsky & Lane, 1998). In a study of high school students, all of the students who had stolen something experienced poorer grades, higher alcohol and drug use, more regular smoking, more sadness, more hopelessness, and more antisocial behaviors than those who had not stolen (Grant, Potenza, Krishnan-Sarin, Cavallo, & Desai, 2011).

Individuals who shoplift are also at higher risk of suicide (Cupchik, 2002; Odlaug, Grant, & Kim, 2012). Indeed, one study from Odlaug et al. (2012) found that 24.3% of a sample of individuals diagnosed with kleptomania had at least one suicide attempt, 93% for whom the attempt was “directly or indirectly due to their kleptomania symptoms (e.g., shame over the behavior; legal or personal problems resulting from shoplifting).”

These results are not limited only to individuals who shoplift regularly. Even carrying out a single shoplifting act, Brady (2013) describes, “results in a psychosocial reidentity stamping procedure whereby the shoplifter surrenders his or her positive preshoplifting self-identity and mentally replaces it with a very negative postshoplifting identity.” One instance can shift an individual’s view of himself and lead to the negative outcomes identified in the literature.

To address psychologically harmful shoplifting behavior, it is important to look at specific interventions that can be developed to prevent or reduce such behavior. However, attempts to develop interventions are complicated by the diversity of experiences and motivations of people who shoplift. To date, researchers and clinicians have drawn on different theories to suggest relevant typologies for people who shoplift. Each has suggested between 2 and 16 “categories” for shoplifting (e.g., Brady, 2013; Cameron, 1964). Yet at this point, no consensus has been established, and none of these typologies has been empirically tested or corroborated by other researchers. The trend in the research has been to suggest each typology in a single paper (e.g., McShane & Noonan, 1993) or throughout one clinician’s body of work (e.g., Cupchik, 2002). These typologies are often based on qualitative observations of clients in a limited sample, with their structures untested in the literature. Still, typologies are important in understanding and delineating between individuals who shoplift, a critical first step to providing effective treatment.

In consideration of these limitations, the goal of this study was to develop a typology of individuals who shoplift based on dimensions of shoplifting behavior. To do so, the researchers examined participants’ shoplifting behaviors, motivations, and precipitating factors as they relate to a range of psychosocial factors considered throughout past literature. In this process, we hoped to consider all of the factors cited in the extant literature and provide a thorough analysis of the behavior. We hoped to categorize individuals according to these dimensions to allow for a better understanding of the range of shoplifting behavior.

Method

Participants

An initial sample of 695 participants was recruited for the screening survey on the Mechanical Turk and Qualtrics platforms. The screened participants were all above the age of 18 years and were located in the United States. Three-hundred ninety-four of these individuals (56.5%) endorsed having ever shoplifted, “or taken property from a store without paying for it.” Of these individuals, 91 (23.3%) indicated they had shoplifted exactly one time before. The individuals who indicated they had shoplifted two or more times were invited to complete the full study.

Of those who qualified for the study, 245 individuals completed the full set of measures. The 245 individuals scored a mean of 13.40 (SD = 6.24) on the Marlowe–Crowne Social Desirability Scale (MC-SDS), similar to the mean score in the literature of 13.72 (SD = 5.78; Crowne & Marlowe, 1960). Respondents who were assessed to be “High Scorers” (i.e., scores equal to or greater than 20) were excluded, resulting in a smaller sample of 202.

The final sample of 202 participants was composed of 112 women and 90 men, and the mean age was 35.56 (SD = 10.82). The majority of participants were single, never married (52.0%), whereas 36.1% were married and 9.4% were divorced. About 80.2% of participants endorsed that they are heterosexual or straight, 5.4% identified as gay or lesbian, and 14.4% endorsed that they are bisexual. A majority of participants self-identified as White/Caucasian (76.7%), with other participants identifying as Hispanic or Latino (6.4%), Black or African American (6.4%), Native American or American Indian (2.5%), or Asian/Pacific Islander (7.4%). The mean annual income was US$36,591.71 (SD = US$54,964.16), and most participants identified as either middle (36.6%) or lower-middle class (32.2%). All participants were located in the United States, and 33.2% indicated that they lived in an urban area, 43.1% in a suburban area, and 23.8% in a rural area.

Measures

Developing a typology necessitated that individuals complete measures of the numerous factors related to shoplifting cited in past research. Previously, various hypotheses have been proposed to explain shoplifting behavior. Including a large number of measures, each of which aligned with previous theory, allowed for a comprehensive study based on the current literature.

Shoplifting behavior

Participants were asked to characterize their shoplifting behavior. The use of self-report measures of criminal behavior is well supported (Thornberry & Krohn, 2000). Participants were queried about their last shoplifting incident, shoplifting frequency, and history of arrest for shoplifting. These items drew on the National Epidemiologic Survey on Alcohol and Related Conditions (Blanco et al., 2008).

Emotions before and after shoplifting

Participants’ feelings before and after shoplifting were assessed using items from the Kleptomania Symptom Assessment Scale (K-SAS; Grant & Kim, 2002). These items were administered because the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5; APA, 2013) notes that kleptomania involves feelings of tension before theft and feelings of pleasure after theft (APA, 2013). The scale is an 11-item self-rated measure that examines urges to steal, thoughts of stealing, and emotions related to theft. The K-SAS demonstrates good internal consistency with a Cronbach’s alpha of .90 (95% CI = [.77, .97]; Grant & Kim, 2002) in past research and Cronbach’s alpha of .85 (95% CI = [.82, .88]) in the present study.

Motivations to shoplift

To assess motivations for shoplifting, the researchers administered several items based on past literature, which suggests that shoplifting can be due to economic motivation (reselling, trading, or professional shoplifting), peer pressure, opportunism, thrill-seeking, psychosocial stress, depression, impulsivity, cognitive impairment, anger, addiction, compulsivity, and loss or trauma history (Brady, 2013; Cupchik, 1997; McShane & Noonan, 1993; Schulueter, O’Neal, Hickey, & Sieler, 1989; Shulman, 2004; Tyminski, 2014). Some of those motivations have existing scales and are described below. Other factors, such as the presence of economic motivation in shoplifting, have no established scale. Therefore, the study implemented the following procedures.

The researchers asked participants about their motivation to shoplift using items rated on a Likert-type scale. To assess for economic motivation, the survey asked about the participants’ endorsement of the statements, “I knew I could resell the item(s) later” and “I needed to provide for my family,” which are cited as economic motivators (Brady, 2013; Cameron, 1964). To evaluate peer pressure (Brady, 2013), the survey assessed the participants’ endorsement of “I wanted to show my friends I could” and other statements about peer influence. Participants’ level of opportunism (Brady, 2013) was measured through their endorsement of items including “The opportunity suddenly presented itself” and “I didn’t think I would have the chance again.” Shoplifting as thrill-seeking (Brady, 2013; Shulman, 2004) was considered by asking about participants’ endorsement of “I was bored and wanted something to do” and “I like taking risks.”

Depression

To indicate their depression at the time they shoplifted, participants completed the 10-item Center for Epidemiological Studies Depression Scale (CES-D 10; Andresen, Malmgren, Carter, & Patrick, 1994; Radloff, 1977). The CES-D is used to screen for depression in the general population. The measure exhibits good internal consistency with a Cronbach’s alpha of .89 (95% CI = [.88, .90]) in past research and Cronbach’s alpha of .88 (95% CI = [.85, .90]) in this study.

Impulsivity

Participants’ levels of impulsiveness were measured using the Barratt Impulsivity Scale, Version 11 (BIS-11; Patton, Stanford, & Barratt, 1995). The BIS-11 is a 30-item measure of aspects of impulsivity: attention, cognitive instability, motor, perseverance, self-control, and cognitive complexity. The BIS asks respondents to indicate how often they exhibit behaviors such as “get easily bored when solving thought problems” and “act on the spur of the moment.” In the last major study of the BIS-11, which included 1,577 participants (Stanford et al., 2009), the BIS-11 showed strong internal reliability with a Cronbach’s alpha of .83 (95% CI = [.82, .84.]). The internal consistency of the BIS-11 in this study was α = .82 (95% CI = [.78, .85]).

Mental impairment

To gauge mental impairment at the time of shoplifting, the researchers administered the Cognitive Function Instrument (Amariglio et al., 2015). This 14-item self-report scale is a measure of cognitive functional decline over time. The instrument asks respondents to indicate their perceived change in various cognitive abilities over the last year, although in this study, participants were instructed to consider changes from the year leading up to the last time they shoplifted. The internal consistency of the scale is strong; Amariglio et al. (2015) found that Cronbach’s alpha for the scale was .78 (95% CI = [.75, .81]). In the present study, Cronbach’s alpha for the scale was .88 (95% CI = [.85, .90]).

Presence of anger

Anger was evaluated with the Dimensions of Anger Reactions–5 (DAR-5) scale (Forbes et al., 2014) because the DSM-5 (APA, 2013) categorizes kleptomania as specifically excluding theft committed that is related to “anger or vengeance.” The DAR-5 includes five items in which respondents indicate the presence of anger symptoms over the last 4 weeks on a Likert-type scale. The internal consistency of the scale is strong, α = .90 (95% CI = [.89, .91]) in past literature and α = .90 (95% CI = [.88, .92]) in this study.

Presence of addiction

The role of addiction was assessed using the Bergen Shopping Addiction Scale (Andreassen et al., 2015). Some research conceptualizes shoplifting as a behavioral addiction (Grant, 2006; McElroy, Pope, Hudson, Keck, & White, 1991; Shulman, 2004). To test the theory that shoplifting can be an addictive behavior, we adapted the Bergen Shopping Addiction Scale, a measure of shopping addiction that assesses the core components of addiction. The measure includes seven statements to which respondents indicate their agreement on a Likert-type scale. The initial validation of the scale, which was conducted with 23,537 individuals, yielded an internal consistency estimate of .87 (95% CI = [.87, .87]; Andreassen et al., 2015), and the estimate in the present study was α = .91 (95% CI = [.89, .93]).

Compulsivity

Some research characterizes shoplifting as a compulsive behavior, fitting into the obsessive compulsive realm of psychopathology (e.g., Grant & Potenza, 2006). To assess for the role of compulsion in shoplifting, the researchers administered an adapted version of the Yale–Brown Obsessive Compulsive Scale (Y-BOCS; Goodman et al., 1989). The Y-BOCS weighs the severity of obsessive compulsive symptoms in 10 items, five of which address obsessions and five of which address compulsions. The scale was adapted to assess obsessive thoughts about shoplifting and compulsions to shoplift. For example, one item in the original scale asks, “How much time do you spend performing compulsive behaviors?” For this study, this question was changed to “How much time do you spend shoplifting?” The Y-BOCS exhibits good internal consistency with reliability estimates of .87 (95% CI = [.86, .89]) in past research (Goodman et al., 1989) and .86 (95% CI = [.83, .89]) in the present study.

Loss history

Cupchik (1997) and Tyminski (2014) theorize that shoplifting is related to a lifelong history of loss. To test the relevance of loss, the researchers administered an adapted version of the Holmes–Rahe Life Stress Inventory (Holmes & Rahe, 1967) including only the items that relate to loss, such as death of spouse, divorce, marital separation from one’s mate, death of a close family member, and child leaving home. The scale was adapted to ask about events across the lifespan, instead of just in the last year, and to document the number of times loss was experienced. These adaptations were made because Cupchik (1997) and Tyminski (2014) connect theft to general experiences of loss across the lifespan, including in childhood (Tyminski, 2014). The scale has strong face validity, criterion validity, validity in predicting mental illness and stress, and test–retest reliability (Spurgeon, Jackson, & Beach, 2001).

Trauma history

Cupchik hypothesizes that some individuals shoplift because of their trauma history (1997). Participants’ trauma history was measured using the Trauma History Questionnaire (THQ; Green, 1996), a 24-item self-report questionnaire that assesses for the occurrence of crime-related events, disasters, and traumatic physical and sexual experiences using a binary yes/no format. Hooper, Stockton, Krupnick, and Green (2011) provided strong support for the THQ’s construct validity, cultural validity, test–retest reliability, and interrater reliability.

Shame and guilt

Overall shame and guilt, as well as shame and guilt when caught shoplifting, were measured using the Personal Feelings Questionnaire–2 (PFQ-2; Harder & Lewis, 1987). The PFQ-2 is a 22-item measure that asks participants to indicate how much they have experienced a variety of feelings on a 5-point scale. For the scale administered when participants indicated they had been caught shoplifting, the instructions were amended to ask participants to consider their feelings when they were caught. In its initial validation, the internal consistency estimate for the shame subscale was .78 (95% CI = [.69, .85]), and the estimate for the guilt subscale was .72 (95% CI = [.61, .81]; Harder & Zalma, 1990). The internal consistency of the scale was supported in this study with a reliability estimate of α = .92 (95% CI = [.90, .94]) for the shame subscale and α = .88 (95% CI = [.85, .90]) for the guilt subscale.

Attitudes about shoplifting

Participants’ ethical attitudes about shoplifting were measured using the short form of the Multidimensional Ethics Scale (Reidenbach & Robin, 1990). This scale is an eight-item measure of ethical judgments on three dimensions: moral equity, or justness and fairness; contractualism, or the extent to which behavior is seen to violate rules; and relativism, or the extent to which a behavior is culturally accepted. In this study, the scenario used as the prompt was drawn from Babin and Babin’s (1996) study of moral cognitions and shoplifting. Extant studies have produced a reliability estimate of .80 (95% CI = [.75, .85]; Reidenbach & Robin, 1990), and the estimate in the present study was .83 (95% CI = [.79, .86]).

Antisociality

The presence of antisociality was measured using the Levenson Self-Report Psychopathy Scale (LSRP; Levenson, Kiehl, & Fitzpatrick, 1995). The DSM-5 distinguishes between shoplifting associated with antisociality and not associated with antisociality (APA, 2013). Brady (2013) also uses antisociality as a differentiating factor between types of individuals who shoplift. The LSRP is a self-report scale that evaluates interpersonal, affective, and social deviance. Evidence for the scale’s internal consistency is strong; past research supports a coefficient α of .83 (95% CI = [.80, .86]; Miller, Gaughan, & Pryor, 2008), and in this study, the estimate was α = .89 (95% CI = [.87, .91]).

Meaning in life and motivation to find purpose in life

The degree of participants’ meaning in life and their strength of motivation to find meaning in life were assessed using the Meaning in Life Questionnaire (MLQ; Steger, Frazier, Oishi, & Kaler, 2006). These facets were used to categorize individuals in McShane and Noonan’s study of shoplifting (2004). The MLQ includes two subscales, Presence of meaning and Search for meaning, each consisting of five items. Both subscales exhibit strong internal consistency; the Presence subscale has a Cronbach’s alpha of .86 (95% CI = [.80, .91]), and the Search subscale has an internal reliability estimate of .92 (95% CI = [.89, .95]; Steger et al., 2006). In the present study, the internal reliability estimates were α = .94 (95% CI = [.93, .95]) and α = .97 (95% CI = [.96, .98]), respectively.

Locus of control

Locus of control was measured using the Rotter (1966) Internality–Externality Locus of Control scale. Brady (2013) categorizes individuals who shoplift based on their locus of control. The Rotter scale is a 29-item self-report questionnaire that measures generalized expectancies for internal versus external control. Extant research has found that the internal reliability coefficient estimate of the scale is as high as .93 (95% CI = [.85, .98]; Beretvas, Suizzo, Durham, & Yarnell, 2008). In this study, the internal reliability coefficient estimate was .76 (95% CI = [.71, .81]).

Social desirability

To ensure that data in this study were not compromised by socially desirable responding, participants completed the MC-SDS (Crowne & Marlowe, 1960). The MC-SDS is a commonly used measure of social desirability with good evidence of validity and reliability (Moss, 2016). Data from the MC-SDS reflect whether participants are responding truthfully or instead trying to manage their self-presentation. Higher scores indicate more socially desirable responding, where “High Scorers” are characterized by “respond[ing] to test items in such a way as to avoid the disapproval of people who may read their responses” (Crowne & Marlowe, 1960). The scale yields strong internal consistency with reliability estimates of α = .79 (95% CI = [.78, .80]; Tanaka-Matsumi & Kameoka, 1986) and α = .78 (95% CI = [.73, .82]) in the present study.

Procedure

A community-based sample of people in the United States was recruited through Amazon’s Mechanical Turk (MTurk). Mechanical Turk is an online platform that allows researchers or “requesters” to compensate workers for simple tasks. Originally used to categorize Amazon products, MTurk has evolved into a reliable and robust method for administering surveys and collecting data in social sciences research. MTurk samples are shown to be highly comparable to standard Internet samples, although MTurk’s samples are significantly more diverse than other samples and more representative of noncollege populations than traditional samples (Gosling, Vazire, Srivastava, & John, 2004). MTurk samples produce data of equal or superior quality to psychometric standards in published research (Buhrmester, Kwang, & Gosling, 2011).

In this study, 695 adult MTurk workers were recruited through the MTurk website for a “short demographics questionnaire” paying US$0.05 for approximately 5 to 10 s of work, yielding an hourly rate of US$18.00 to US$36.00. The questionnaire, administered using the Qualtrics platform, collected data related to the presence of shoplifting, including incidence of shoplifting and frequency over the lifespan. This screening established inclusion criteria, specifically determining that the qualifying participants were above the age of 18 years, lived in the United States, and had shoplifted at least two times in their lives. Three-hundred ninety-four of the screened individuals (56.5%) endorsed having ever shoplifted, “or taken property from a store without paying for it.” This figure is similar to past research on the prevalence of shoplifting (Klemke, 1982; Kraut, 1976). Of these individuals, 91 (23.3%) indicated they had shoplifted only one time. The individuals who endorsed shoplifting two or more times were immediately invited through the survey to complete the full study for a bonus of US$4.00. Each respondent who completed the study was thus compensated a total rate of US$8.00 per hour, given an average length of 30 min.

To ensure the quality of responses beyond that supported by Buhrmester et al. (2011), the project required that each worker had completed more than 100 assignments with an approval rate greater than 95%. Known as reputation screening, this method of ensuring quality responding is supported by Peer, Vosgerau, and Acquisti (2014). In addition, a qualitative, open-ended question examined thoroughness and thoughtfulness of responding. While this qualitative information was not analyzed for the present study, the responses to this question supported the attentiveness of participants to the study.

After providing consent, participants provided information about their shoplifting behavior, including its recency and incidence over the lifespan. Next, they completed the measures described above. Individuals who had been caught shoplifting completed the PFQ-2 on guilt and shame at the time of being caught. Items from the MC-SDS were interspersed throughout the other measures to covertly assess for socially desirable responding.

At the end of the survey, participants completed the PFQ-2 on general guilt and shame, ordered at the end so that thoughts of shoplifting would be salient. Finally, participants completed a demographic questionnaire including sex, age, educational attainment, relationship status, sexual orientation, ethnicity, religion, urbanicity, income, socioeconomic status, and criminal history.

Results and Discussion

Two-hundred two individuals comprised the final sample for this study, with no missing data. These 202 individuals indicated that they had shoplifted between 2 and 1000 times with a mean of 28.60 (SD = 104.23). None of the demographic variables measured were significantly associated with the incidence of shoplifting across the lifespan.

In performing a cluster analysis, the researchers began by conducting a principal components analysis with the scores from the relevant scales. The principal components analysis was critical to addressing two key challenges in clustering: first, that the scores for different measures have different weights, and second, that correlations between variables could lead to a double counting of the constructs (Multivariate Solutions, 2017). Principal component analysis allowed the researchers to emphasize variations in the data and identify strong patterns in the data set (Kelley, 2010). This analysis identified eight principal components in the data set.

Next, hierarchical clustering using Ward’s method allowed the researchers to determine how many clusters best grouped the participants. To determine the appropriate number of clusters in the data, the researchers examined the Agglomeration Schedule for the hierarchical clustering analysis, which indicates the distance coefficient for each stage of clustering, where each stage is a further reduction in the number of assumed clusters. Using the elbow rule to identify the number of clusters (Ketchen & Shook, 1996), the researchers determined that the data were best described by a six-cluster solution.

The researchers then performed a k-means cluster analysis to group the participants into six distinct clusters. This analysis yielded a data set that was divided into six groups, whose values on each scale could be compared. The quality of the clustering was supported by a silhouette coefficient of .55, indicating that participants in each cluster were similar to each other and different from participants in other clusters. Comparing scale scores for each cluster through a collaborative process, key themes emerged that defined each cluster. These results will be discussed within the context of the current literature.

Loss-Reactive Type

The most common cluster revealed in 27.7% of the sample (56 individuals) was labeled the Loss-Reactive Type. This type of individual was characterized by generally law-abiding behavior, stability in mental health, and the preponderance of lifetime loss and trauma. These participants were unlikely to be arrested for another crime and had the highest orientation to traditional ethical values, as compared with the other study participants. They had a high incidence of lifetime losses and traumas. They generally felt a lack of agency and scored highest in external locus of control, compared with the other participants.

Interestingly, these participants presented with the lowest depression, anger, cognitive impairment, addictive tendencies, compulsiveness, impulsivity, and antisociality of all of the types identified in the study. Over the lifespan, individuals in the Loss-Reactive Type had shoplifted the least number of times (Mdn = 4, M = 11.98, SD = 20.79) and the least expensive items (Mdn = US$5.00, M = US$7.92, SD = US$8.94). They were most likely to only use their bare hands while shoplifting, instead of using more sophisticated tools (e.g., a bag with hidden compartments), and did not feel a desire to shoplift. These individuals also had the highest mean annual income of all types in the study (M = US$42,009, SD = US$51,403).

This type aligned strongly with a documented category from the literature. Cupchik (1997) dedicated his career to studying the Atypical Theft Offender, described as being generally law-abiding and having experienced significant losses or traumas. Cupchik’s “Loss-Substitution-by-Stealing” theory describes why these individuals shoplift, noting that shoplifting is a way for individuals to compensate subconsciously for an unfair loss.

Although the present study could not speak to the subconscious motives for shoplifting, the Atypical Theft Offender seems to be evident in the present study’s Loss-Reactive Type. The presence of a Loss-Reactive Type in this study lends the first empirical support for Cupchik’s theory. The Loss-Reactive Type, as with the other types identified in this study, was developed by combining a broad analysis of the full range of shoplifting correlates with a deep examination of existing theory and research in the field. This methodology has not been utilized previously in developing knowledge about shoplifting and individuals who shoplift.

Impulsive Type

The next cluster, which included 20.3% of the sample (41 participants), made up the Impulsive Type. These participants were characterized by high impulsivity; high antisociality 2, a dimension that captures impulsivity; low self-control; and location in more urban or suburban areas. Impulsive Type individuals are presented with ample opportunities to shoplift in an urban or suburban environment and may carry out shoplifting as an impulsive behavior with little self-control. This type was also characterized by having the most ability to pay for the stolen items, as compared with the other types; a higher sense of opportunism when committing the thefts; a low overall incidence of shoplifting across the lifespan; low shame; the lowest sense of guilt of any type; the lowest incidence of life stressors before the shoplifting; the lowest presence of meaning in life; and the most internal locus of control.

This type bears some similarities with categories defined in the past literature. Moore (1984), for example, identified an “impulsive shoplifter,” who showed little planning and greater impulsivity. Unlike Moore’s “impulsive shoplifter,” the Impulsive Type identified in this research was not characterized by only stealing inexpensive items. Furthermore, the individuals in the present study endorsed low guilt and shame, whereas Moore’s “impulsive shoplifter” experienced intense guilt and shame. The Impulsive Type identified in this research bears some similarities to historical designations of kleptomania: The type is characterized by impulsivity, an ability to pay for stolen goods, and low economic motivation for theft.

Depressed Type

The next cluster, comprising 18.3% of the sample (37 individuals), experienced acute depression and makes up a Depressed Type. Individuals of this type typically scored a 16.03 on the Center for Epidemiological Studies Depression Scale, on which a score of 10 is the cutoff for clinically relevant depression (Andresen et al., 1994; Radloff, 1977). Depressed Type individuals may engage in shoplifting as a way of coping with depression or as a means to feel something instead of anhedonia or apathy. In this study, Depressed Type individuals were marked by the highest overall sense of guilt compared with other types, high shame, the lowest thrill-seeking, and the lowest desire to shoplift. This type endorsed the strongest orientation to traditional ethical values and the highest sense of searching for meaning in life. Individuals in the Depressed Type had shoplifted fewer times than the average study participant and were the least likely to have been arrested for any other crime.

This type is unique in the literature on shoplifting, which has so far not identified a type based solely on the predominance of depression symptoms. This type does not parallel other models, but it emerged clearly in the current study as a group strongly linked to high numbers of depressive symptoms.

Hobbyist Type

Another category of individuals comprised 17.8% of the sample (36 individuals) and is best described as a Hobbyist Type. These participants were marked by enjoyment of shoplifting and high psychological functioning and well-being. The Hobbyist Type had committed the second-most incidences of theft in the sample, with a mean total incidence of shoplifting of 59.39 times (SD = 172.00). These individuals had a high orientation to traditional ethics and yet did not experience distress, guilt, or shame, indicating that they may see themselves as above, outside, or exempt from the law. Hobbyists had the highest presence of meaning in life, as compared with the other study participants, and scored low in depression, cognitive impairment, antisociality, and searching for meaning in life. For Hobbyists, shoplifting may represent an ego-syntonic and enjoyable activity.

Like the Depressed Type, the Hobbyist Type was a unique category that emerged in this study. The Hobbyist Type has likely not been captured in past research because Hobbyists do not seek psychological help and seem to avoid being caught shoplifting, the two primary sources of recruitment in past research. Conducting a study with the general population, as in the present study, may have allowed for the capture of this well-adapted, Hobbyist individual.

Addictive–Compulsive Type

The fifth cluster identified 8.9% of the study participants (18 individuals) and was characterized by an addictive and compulsive pattern of shoplifting with enjoyment of the thrill of the act. Of all the types identified in the study, the Addictive–Compulsive Type is marked by the highest sense of addictiveness, compulsiveness, impulsivity, thrill-seeking, opportunism, desire to shoplift, urge to collect things, and antisociality 1 (measuring general psychopathy). These individuals may enjoy the act of shoplifting and show signs of an addictive–compulsive drive to shoplift. They have the lowest orientation to traditional ethical values and the lowest incidences of loss and trauma over the lifespan. On average, they stole the most expensive items of all types (Mdn = US$13.50), and if caught shoplifting, they experienced the highest shame and guilt. Individuals in this type were more likely to be male and were the youngest age at the time of study completion (Mdn = 27.50 years).

This type bears resemblance to categories identified in previous literature. Shulman’s (2004) Addictive–Compulsive group and Brady’s (2013) Compulsive (Impulse-Driven) group include individuals motivated by repressed anger, addictiveness, and compulsiveness who experience guilt and shame when caught. The Addictive–Compulsive Type identified in the present study is also similar to Shulman’s (2004) Thrill Seekers and Brady’s (2013) Thrill Seeker (Psychologically Motivated). Brady (2013) also suggested a Provisional/Delinquent Shoplifter, characterized by antisociality, hedonism (or a desire to shoplift), peer pressure, and adolescence. The present study found that all of these elements can be found in one generally homogeneous group of addictive–compulsive individuals who are motivated by thrill and enjoyment.

Economically Disadvantaged Type

The last cluster in the analyses comprised of 6.9% of the sample (14 individuals) and was marked by the lowest annual income and socioeconomic status. These individuals represent an Economically Disadvantaged Type. These individuals were all highly motivated by economic advantages associated with shoplifting and typically stole more expensive items (Mdn = US$10.00, M = US$22.93, SD = US$51.31) that they had the least ability to pay for, compared with other types, which suggests economic need. They engaged in the most sophisticated methods to aid in their shoplifting, such as a bag lined with foil to evade radio frequency scanners. These participants had shoplifted more times than any other type, a median of 36 times over the lifespan. Importantly, this type was characterized by having the highest degree of anger, life stress, incidence of life stressors before shoplifting, and trauma history. If caught shoplifting, the Economically Disadvantaged Type experienced a low sense of shame and guilt. This may be related to the type’s lack of adherence to traditional ethical values and/or to their sense of economic need, which could buffer against those negative self-perceptions. The age and race or ethnicity of these individuals did not differ significantly from that of the rest of the sample.

This type resembles other types identified in the literature on shoplifting behavior. Cupchik (1997) described what he calls the Typical Theft Offender, an economically motivated individual who consciously steals and does not experience remorse. Shulman (2004) and Brady (2013) identified an Impoverished (Economically Disadvantaged) shoplifter who is motivated by economic need and is hostile against the system (has high anger), both of which were found in this study’s Economically Disadvantaged Type. Another type identified by Shulman (2004) and Brady (2013), the Professionals, was marked by economic motivation, stole expensive items using sophisticated methods, and did not experience remorse if caught. This looks similar to the present study’s Economically Disadvantaged Type, except that the individuals in this research did not sell or trade their stolen items any more than other types. Research indicates that a subset of individuals who shoplift do so out of economic necessity and may be motivated by hostility toward the system that has created a context of disadvantage.

Implications and Limitations

Overall, the typology delineated in the current study can be used to differentiate between individuals who shoplift and to develop specialized interventions to treat these individuals. These findings represent a significant contribution to the current body of research, which has not so far utilized an extensive literature review and robust statistical analysis to determine a typology. Most of the extant typologies fail to exercise quantitative or statistical methods to determine categories, which are critical to individualizing treatment for those who shoplift. Although McShane and Noonan (1993) did use a cluster analysis in their study, they considered a small, limited number of factors. Collecting and analyzing data on the large number of factors related to shoplifting, drawing from multidisciplinary research on shoplifting, and capturing information across more than 16 scales allowed for a deeper, more structured, and more rigorous analysis of the behavior of shoplifting. The cluster analysis in the present study addresses the limitations of past research to reveal an empirically based typology of individuals who shoplift.

Collectively, these findings suggest that individuals who shoplift can be characterized as one of six types, which can inform their treatment and recovery from shoplifting behavior. One recommendation based on our findings is the promise of clinicians conducting clinical interviews to categorize clients in types according to the descriptions above. Knowing each client’s location in this typology will allow clinicians to tailor interventions to the unique concerns, motivations, and behaviors of each type.

Future research must explore how to help different individuals reach their behavioral goals. This research can build on the present study to further explore the utility and effects of shoplifting on the psychology of individuals who shoplift. Future research might also consider shoplifting from each type and investigate how to intervene accordingly. Other studies could explore which treatment approaches best reduce shoplifting behavior in different types. This type of research would result in tailored approaches for each of these outlined above.

For the Loss-Reactive Type, for example, reducing shoplifting behavior may involve interventions that help clients process and move on from losses and traumas. Depressed Type individuals, alternatively, may best respond to direct interventions that address their depression and its underlying causes. Treating individuals who fit into the Addictive–Compulsive Type may involve approaches that echo treatments for substance use, gambling, and other behavioral addictions. Continued research will play an important role in developing and assessing these potential interventions.

In considering the implications of this study, it is important to discuss relevant limitations. First, while the MTurk participants in this study represented a sample with ethnic, gender, and sexual orientation diversity, MTurk workers are not ideally representative of the U.S. population. To assess for the effect of sampling on the study results, the researchers collected demographic information that included various factors and measured the relationship between them and the variables of interest in this study. The data indicated that there were no correlations between MTurk-related demographics and the study factors of interest. Although MTurk samples may be limited compared with the general population, this limitation did not appear to have an impact on the results of the study.

A further limitation concerns the use of self-report measures. Whereas respondents in past research have been noted to be willing to self-report criminal behavior, they still tend to underreport this behavior (Thornberry & Krohn, 2000). Therefore, while participants were likely to be honest about whether they had shoplifted before, they may have underestimated the frequency or incidence of their behavior. To mitigate the effects of possible underreporting, the researchers administered the Marlowe–Crowne Social Desirability Scale and discarded data for individuals who were High Scorers on the scale.

In addition, the Cognitive Function Instrument (CFI) may not adequately capture the presence of mental impairment while shoplifting. Although the Cognitive Function Instrument is strongly associated with clinical measures of functional abilities (Amariglio et al., 2015), it cannot show mental impairments during shoplifting. Individuals may not be aware of their impairments, and their responding may be subject to fallible memories; if a long time has passed since shoplifting, individuals may not recall their mental functioning at the time of the last incidence.

Similarly, the CES-D 10 was administered to assess for depression at the time of shoplifting, but its validity may have been affected by memory. If a long period of time passed since the shoplifting behavior, individuals may have underestimated or overestimated the presence of depression symptoms during that time. This may be especially true if their depression symptoms have changed since the shoplifting behavior.

Despite these limitations, the results of this study provide a strong starting point for understanding and helping individuals who shoplift. Developing interventions to help individuals who shoplift remains a challenge for researchers and clinicians. This study provides guidelines to evaluate how individuals who shoplift can be categorized and further studied. Moving forward, further research examining interventions will assist in developing effective treatments for the many who struggle with shoplifting.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was partially funded by the June Marie Gallessich Dissertation Award in Educational Psychology by the University of Texas at Austin.

ORCID iD

Miranda M. Nadeau

References

Amariglio

R. E.

Donohue

M. C.

Marshall

G. A.

Rentz

D. M.

Salmon

D. P.

Ferris

S. H.

, . . . Alzheimer’s Disease Cooperative Study. (2015). Tracking early decline in cognitive function in older individuals at risk for Alzheimer disease dementia: The Alzheimer’s Disease Cooperative Study Cognitive Function Instrument. Journal of the American Medical Association—Neurology, 72, 446-454. doi:10.1001/jamaneurol.2014.3375

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.

Andreassen

C. S.

Griffiths

M. D.

Pallesen

Bilder

R. M.

Torsheim

Aboujaoude

(2015). The Bergen Shopping Addiction Scale: Reliability and validity of a brief screening test. Frontiers in Psychology, 6, 1374. doi:10.3389/fpsyg.2015.01374

Andresen

E. M.

Malmgren

J. A.

Carter

W. B.

Patrick

D. L.

(1994). Screening for depression in well older adults: Evaluation of a short form of the CES-D. American Journal of Preventive Medicine, 10, 77-84.

Babin

B. J.

Babin

L. A.

(1996). Effects of moral cognitions and consumer emotions on shoplifting intentions. Psychology and Marketing, 13, 785-802. doi:10.1002/(SICI)1520-6793(199612)13:8<785::AID-MAR5>3.0.CO;2-B

Beretvas

S. N.

Suizzo

Durham

J. A.

Yarnell

L. M.

(2008). A reliability generalization study of scores on Rotter’s and Nowicki-Strickland’s Locus of Control scales. Educational and Psychological Measurement, 68, 97-119. doi:10.1177/0013164407301529

Blanco

Grant

Petry

N. M.

Simpson

H. B.

Alegria

Liu

Hasin

(2008). Prevalence and correlates of shoplifting in the United States: Results from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). American Journal of Psychiatry, 165, 905-913. doi:10.1176/appi.ajp.2008.07101660

Brady

J. C.

(2013). Why rich women shoplift—When they have it all. San Jose, CA: Western Psych Press.

Buhrmester

Kwang

Gosling

(2011). Amazon’s Mechanical Turk: A new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6, 3-5. doi:10.1177/1745691610393980

10.

Cameron

M. O.

(1964, May 17). The Booster and the Snitch: Department Store Shoplifting. New York: Free Press of Glencoe.

11.

Crowne

D. P.

Marlowe

(1960). A new scale of social desirability independent of psychopathology. Journal of Consulting Psychology, 24, 349-354. doi:10.1037/h0047358

12.

Cupchik

(1997). Why honest people shoplift or commit other acts of theft: Assessment and treatment of ‘Atypical Theft Offenders’. Toronto, Canada: Tagami Communications.

13.

Cupchik

(2002). Why honest people shoplift or commit other acts of theft: Assessment and treatment of atypical theft offenders. Toronto, Ontario, Canada: Tagami Communications.

14.

Forbes

Alkemade

Mitchell

Elhai

J. D.

McHugh

Bates

Lewis

(2014). Utility of the Dimensions of Anger Reactions-5 (DAR-5) scale as a brief anger measure. Depression and Anxiety, 31, 166-173. doi:10.1002/da.22148

15.

Forney

W. S.

Crutsinger

Forney

J. C.

(2006). Self-concept and self-worth as predictors of self-perception of morality: Implications for delinquent risk behavior associated with shoplifting. Family and Consumer Sciences Research Journal, 35, 24-43. doi:10.1177/1077727X06289640

16.

Geyer

M. D.

(2000). A psychological examination of first time apprehended shoplifting offenders: An empirical investigation of Cupchik’s loss-substitution-by-stealing hypothesis (Unpublished doctoral dissertation). Spalding University, Louisville, KY.

17.

Goodman

W. K.

Price

L. H.

Rasmussen

S. A.

Mazure

Fleischmann

R. L.

Hill

C. L.

. . . Charney

D. S.

(1989). The Yale-Brown Obsessive Compulsive Scale. I. Development, use, and reliability. Archives of General Psychiatry, 46, 1006-1011.

18.

Gosling

S. D.

Vazire

Srivastava

John

O. P.

(2004). Should we trust Web-based studies? A comparative analysis of six preconceptions about Internet questionnaires. American Psychologist, 59, 93-104. doi:10.1037/0003-066X.59.2.93

19.

Grant

J. E.

(2006). Understanding and treating kleptomania: New models and new treatments. Israel Journal of Psychiatry & Related Sciences, 43, 81-87.

20.

Grant

J. E.

Kim

S. W.

(2002). An open-label study of naltrexone in the treatment of kleptomania. Journal of Clinical Psychiatry, 63, 349-356.

21.

Grant

J. E.

Potenza

M. N.

(2006). Compulsive aspects of impulse-control disorders. The Psychiatric Clinics of North America, 29, 539-x. doi:10.1016/j.psc.2006.02.002

22.

Grant

J. E.

Potenza

M. N.

Krishnan-Sarin

Cavallo

D. A.

Desai

R. A.

(2011). Stealing among high school students: Prevalence and clinical correlates. Journal of the American Academy of Psychiatry and the Law, 39, 44-52.

23.

Green

B. L.

(1996). Trauma History Questionnaire. In Stamm

B. H.

(Ed.), Measurement of stress, trauma, and adaptation (pp. 366-369). Lutherville, MD: Sidran Press.

24.

Harder

D. H.

Zalma

(1990). Two promising shame and guilt scales: A construct validity comparison. Journal of Personality Assessment, 55, 729-745.

25.

Harder

D. W.

Lewis

S. J.

(1987). The assessment of shame and guilt. In Butcher

J. N.

Spielberger

C. D.

(Eds.), Advances in personality assessment (Vol. 6, pp. 89-114). Hillsdale, NJ: Lawrence Erlbaum.

26.

Holmes

T. H.

Rahe

T. H.

(1967). The Social Readjustment Rating Scale. Journal of Psychosomatic Research, 11, 213-218.

27.

Hooper

Stockton

Krupnick

Green

(2011). Development, use, and psychometric properties of the Trauma History Questionnaire. Journal of Loss and Trauma, 16, 258-283. doi:10.1080/15325024.2011.572035

28.

Kelley

M. E.

(2010). Principal component analysis. In Corsini encyclopedia of psychology. doi:10.1002/9780470479216.corpsy0709

29.

Ketchen

D. J.

Shook

C. L.

(1996). The application of cluster analysis in strategic management research: An analysis and critique. Strategic Management Journal, 17, 441-458. doi:10.1002/(SICI)1097-0266(199606)17:6<441::AID-SMJ819>3.0.CO;2-G

30.

Klemke

L. W.

(1982). Exploring juvenile shoplifting. Sociology and Social Research, 67, 59-75.

31.

Krasnovsky

Lane

R. C.

(1998). Shoplifting: A review of the literature. Aggression and Violent Behavior, 3, 219-235.

32.

Kraut

R. E.

(1976). Deterrent and definitional influences on shoplifting. Social Problems, 23, 358-368. doi:10.1525/sp.1976.23.3.03a00110

33.

Levenson

M. R.

Kiehl

K. A.

Fitzpatrick

C. M.

(1995). Assessing psychopathic attributes in a noninstitutionalized population. Journal of Personality and Social Psychology, 68, 151-158.

34.

McElroy

S. L.

Pope

H. G.

Hudson

J. I.

Keck

P. E.

Jr. White

K. L.

(1991). Kleptomania: A report of 20 cases. American Journal of Psychiatry, 148, 652-657.

35.

McShane

F. J.

Noonan

B. A.

(1993). Classification of shoplifters by cluster analysis. International Journal of Offender Therapy and Comparative Criminology, 37, 29-40. doi:10.1177/0306624X9303700104

36.

Miller

J. D.

Gaughan

E. T.

Pryor

L. R.

(2008). The Levenson Self-Report Psychopathy Scale: An examination of the personality traits and disorders associated with the LSRP factors. Assessment, 15, 450-463. doi:10.1177/1073191108316888

37.

Moore

R. H.

(1984). Shoplifting in middle America: Patterns and motivational correlates. International Journal of Offender Therapy and Comparative Criminology, 28, 53-64.

38.

Moss

(2016, June 18). Crowne and Marlowe Social Desirability Scale. Retrieved from http://www.sicotests.com/psyarticle.asp?id=229

39.

Multivariate Solutions. (2017). SPSS tutorial-cluster analysis. Retrieved from https://mvsolution.com/spss-tutorial-cluster-analysis/

40.

National Retail Federation. (2016). National Retail Security Survey (Report). Retrieved from https://nrf.com/resources/retail-library/national-retail-security-survey-2016

41.

Odlaug

B. L.

Grant

J. E.

Kim

S. W.

(2012). Suicide attempts in 107 adolescents and adults with kleptomania. Archives of Suicide Research, 16, 348-359.

42.

Patton

J. H.

Stanford

M. S.

Barratt

E. S.

(1995). Factor structure of the Barratt impulsiveness scale. Journal of Clinical Psychology, 51, 768-774.

43.

Peer

Vosgerau

Acquisti

(2014). Reputation as a sufficient condition for data quality on Amazon Mechanical Turk. Behavioral Research Methods, 46, 1023-1031. doi:10.3758/s13428-013-0434-y

44.

Radloff

L. S.

(1977). CES-D scale: A self-report depression scale for research in the general populations. Applied Psychological Measurement, 1, 385-401.

45.

Reidenbach

R. E.

Robin

D. P.

(1990). Toward the development of a multidimensional scale for improving evaluations of business ethics. Journal of Business Ethics, 9, 639-653. doi:10.1007/BF00383391

46.

Rotter

J. B.

(1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs: General and Applied, 80, 1-28.

47.

Sansone

R. A.

Lam

Wiederman

M. W.

(2011). The relationship between shoplifting and borderline personality symptomatology among internal medicine outpatients. Innovations in Clinical Neuroscience, 8, 12-13.

48.

Sarasalo

Bergman

Toth

(1997). Theft behaviour and its consequences among kleptomaniacs and shoplifters: A comparative study. Forensic Science International, 5, 193-205. doi:10.1016/S0379-0738(97)02131-2

49.

Schulueter

G. R.

O’Neal

F. C.

Seiler

G. L.

(1989). Rational vs. nonrational shoplifting types: The implications for loss prevention strategies. International Journal of Offender Therapy and Comparative Criminology, 33, 227-239.

50.

Shulman

(2004). Something for nothing: Shoplifting addiction and recovery. Philadelphia, PA: Infinity Publishing.

51.

Spurgeon

Jackson

C. A.

Beach

J. R.

(2001). The Life Events Inventory: Re-scaling based on an occupational sample. Occupational Medicine, 51, 287-293.

52.

Stanford

M. S.

Mathias

C. W.

Dougherty

D. M.

Lake

S. L.

Anderson

N. E.

Patton

J. H.

(2009). Fifty years of the Barratt Impulsiveness Scale: An update and review. Personality and Individual Differences, 47, 385-395.

53.

Steger

M. F.

Frazier

Oishi

Kaler

(2006). The Meaning in Life Questionnaire: Assessing the presence of and search for meaning in life. Journal of Counseling Psychology, 53, 80-93.

54.

Tanaka-Matsumi

Kameoka

V. A.

(1986). Reliabilities and concurrent validities of popular self-report measures of depression, anxiety, and social desirability. Journal of Consulting and Clinical Psychology, 54, 328-333.

55.

Thornberry

T. P.

Krohn

M. D.

(2000). The self-report method for measuring delinquency and crime. Measurement and Analysis of Crime and Justice, 4, 33-83.

56.

Tyminski

(2014). The psychology of theft and loss: Stolen and fleeced. London, England: Routledge.