An Exploratory Study of Adolescent Shoplifting Behavior

ABSTRACT - This paper presents the results of an exploratory study of adolescent shoplifting behavior. A conceptual framework is first presented as a blue print for studying shoplifting behavior. Within this framework hypotheses are developed based upon theories of delinquent behavior and some scattered empirical data. The study results suggest that peers may play an important role in the development of shoplifting behaviors. Some directions for future research are suggested.


George P. Moschis, Dena Saliagas Cox, and James J. Kellaris (1987) ,"An Exploratory Study of Adolescent Shoplifting Behavior", in NA - Advances in Consumer Research Volume 14, eds. Melanie Wallendorf and Paul Anderson, Provo, UT : Association for Consumer Research, Pages: 526-530.

Advances in Consumer Research Volume 14, 1987      Pages 526-530


George P. Moschis, Georgia State University

Dena Saliagas Cox, Georgia State University

James J. Kellaris. Georgia State University


This paper presents the results of an exploratory study of adolescent shoplifting behavior. A conceptual framework is first presented as a blue print for studying shoplifting behavior. Within this framework hypotheses are developed based upon theories of delinquent behavior and some scattered empirical data. The study results suggest that peers may play an important role in the development of shoplifting behaviors. Some directions for future research are suggested.


Marketers and consumer behavior researchers traditionally have investigated consumer behavior in order to understand how to better market products/services. As a result, the focus of their research efforts has been upon understanding normative consumer behaviors (e.g., comparison shopping, information seeking) as well as desirable behavioral patterns, at least from the marketer's point of view, such as brand loyalty and store loyalty. Considerably less attention has been devoted to understanding undesirable consumer behaviors that have negative consequences upon businesses and society as a whole.

Shoplifting is an undesirable activity which has become a growing concern not only among retailers but also among consumer educators, governments and social scientists. It is the nation's largest monetary crime, accounting for up to 7.5 percent of dollar sales (Messenger 1975) and over $16 billion annually (Forbes 1981), and it is on the rise (Velocci 1978).

Retailers concerned with shoplifting generally have two broad strategies for reducing shoplifting losses: shoplifter detection and shoplifting prevention. Although most of the efforts have focused upon shoplifter detection, this strategy remains highly questionable for at least two reasons. First, reliable records show that for every one theft apprehended thirty-four get away (Taylor 1979). Second, often those apprehended are more likely to do more subsequent shoplifting than those not apprehended (Klemke 1978).

Shoplifting prevention, on the other hand, involves increasing the difficulty of shoplifting (e.g., installing electronic warning systems) and decreasing the desire for shoplifting among potential shoplifters. The latter involves identifying characteristics of those likely to shoplift that would help in alerting security personnel, as well as understanding the process leading to the development of shoplifting behavior and its prevention.

Previous research attempting to identify characteristics of shoplifters has offered insights into the profile of the typical shoplifter (e.g., Robin 1963; Cameron 1964; Won and Yamamoto 1968; Brandy and Mitchell 1971; Wright and Kirmani 1977; Beck and McIntyre 1977; Klemke 1982). Unfortunately, many of these studies use only subjects who have been apprehended. Furthermore, their results are not interpreted in the context of any theories that would help us better understand why people shoplift.

The purpose of this paper is to present a conceptual framework which suggests antecedents and social processes contributing to the development of shoplifting behavior, and to test this proposed approach and resultant hypotheses on a small scale basis using teenagers. Teenagers were chosen as an appropriate group to study because young people are reported to account for most shoplifting (Verill 1978; Stores 1971; Security Management 1978). Explanations for this particular aspect of deviant behavior are sought in theories of developmental psychology and sociological models of human behavior which are integrated into a broader socialization perspective. The socialization approach makes the assumption that in order to understand human behavior one must specify its social origins and the processes under which it is acquired and maintained (McLeod and O'Keefe 1972).


Two general models of human behavior are used to explain the acquisition of consumer behavior in general and delinquent behavior in particular among youths: the cognitive developmental model and the social learning model. Studies utilizing the cognitive developmental approach seek to explain the development of youths' delinquent behavior as a function of qualitative changes (stages) in cognitive organization occurring between infancy and adulthood. Such stages are often defined in terms of cognitive structures the child can use in perceiving and dealing with the environment at different stages. The acquisition and subsequent enactment of delinquent norms is assumed to occur as the child moves from one stage to the next. The social learning approach, in contrast, seeks to explain learning of delinquent behaviors, norms and values as a function of sources of influence in the child's environment. Learning of such orientations is assumed to be taking place during the person's interaction with such sources of influence, commonly known as "socialization agents," in various social settings defining the structure of the child's physical and subcultural environment. Such social "structures often have a direct effect upon the learning of various behaviors as well as an indirect effect by influencing the person's level of interaction d the the various socialization agents.

Previous researchers have placed these models and variables in a context of a socialization perspective that has been used to explain various aspects of the person's behavior, including consumer behavior, over the life cycle (e.g., Churchill and Moschis 1979; Moschis forthcoming). Generally, age is viewed as an index of cognitive development and is perceived along with other variables which define various socio-cultural attributes as antecedent variables. These factors affect the person's interaction with, and processes of learning from the social milieu. Socialization agents define the socialization process which, along with the antecedent variables, can have direct impact upon the acquisition of values, norms, and behaviors -- i.e. outcomes. In sum, cognitive developmental perspectives are used to explain relationship of age with the socialization process and outcomes, while social learning models are used to explain the acquisition of learning properties from socialization agents in various social (structural) settings (Churchill and Moschis 1979, Moschis forthcoming).

Previous researchers have applied socialization perspectives to the study of various aspects of youth's delinquent behavior, including shoplifting. For example, Klemke (1978 and 1982) studied the relationship between age and delinquency among youths and the role of peer influence on shoplifting. Similarly, Powell and Moschis (1986) investigated shoplifting behavior of juveniles from a socialization perspective. The present study extends this sparse body of delinquent socialization research by examining a greater number of relationships between antecedent and socialization process variables and shoplifting motivations, attitudes, and behavior, thereby providing a potentially broader perspective.

Specifically, it is believed that with increasing age the adolescent is likely to become more susceptible to peer pressure. Studies, for example, show that peer influence increases with age during adolescence (e.g., Campbell 1969; Churchill and Moschis 1979). Since shoplifting involves game-like aspects which are shared with others (e.g., fun and excitement) (Gold 1970; Cameron 1964; Klemke 1982), age is expected to be positively related to such sporting motivations for shoplifting, and these sporting motivations are expected to be associated with peer interaction frequency.

H1: Sporting motivations for shoplifting are positively related to age.

H2: There is a positive relationship between frequency of communication with peers about shoplifting and: (a) social motivations for shoplifting, and (b) sporting motivations for shoplifting.

As socialization agents, peers may also affect shoplifting behaviors and attitudes. Klemke (1982) studied 1,189 juveniles and found a strong relationship (r - .82) between the youth's recent shoplifting behaviors and his/her g of a close friend who had also shoplifted. a us, it is possible that peers may serve as a source of awareness as well as attitude and behavior development regarding shoplifting.

H3: Frequency of communication with peers about shoplifting is positively related to:

(a) favorability of attitudes toward shoplifting, and

(b) frequency of shoplifting.

The extent to which peers are likely to be used as a reference group and source of influence appears to be a function of the youth's family structure. Generally, theory and research suggest that youths most likely to engage in delinquent activities are those from large families and unstable homes or from homes where one or both parents are frequently absent, those who are at odds with their parents and whose behaviors are constantly restricted by their families (Moschis forthcoming). These conditions are likely to alienate the youth, leading her/him to turn to peers in an effort to compensate for lack of parental attention, warmth and supportiveness, and even engage in activities disapproved of by parents (e.g., delinquent behaviors). These youths are likely to become assimilated into peer groupings which tend to be less acceptable in the dominant peer groups of the school.

H4: Family size is positively associated with:

(a) social motivations for shoplifting

(b) favorable attitudes toward shoplifting, and (c) frequency of shoplifting.

H5: The degree of absence of parents is positively associated with:

(a) social motivations for shoplifting

(b) favorability of shoplifting attitudes, and

(c) frequency of shoplifting.

Communication structure, as a socialization process variable, may also influence a child's behavior. "Socio-oriented" communication structures are a style of parent-child communication which emphasizes obedience, deference, and family harmony (cf., McLeod and O'Keefe 1972). Because families placing ample restrictions on their children are characterized by such a communication structure (cf., McLeod and O'Keefe 1972) it is expected that:

H6: Socio-oriented family communication will be positively associated with:

(a) social motivations for shoplifting

(b) favorability of attitudes toward shoplifting, and

(c) frequency of shoplifting.

Finally, research shows that socioeconomic characteristics of youth may explain shoplifting behavior and motivations. Specifically, studies point to a moderate negative relationship between socioeconomic status and shoplifting (Won and Yamomoto 1968; Klemke 1982). The likely higher propensity for lower social-class youths to shoplift may be due to lack of economic resources:

H7: The Youth's socioeconomic status is negatively associated with:

(a) economic motivations for shoplifting

(b) favorable attitudes toward shoplifting, and

(c) frequency of shoplifting.


Self-administered anonymous questionnaires were distributed by trained doctoral students to a judgement sample of approximately 150 adolescents in middle and high schools. This sample of subjects was representative in terms of the major socio-economic variables, such as parents' occupation, race, etc. for the purpose of this study.

The nature of the shoplifting behavioral self-report items posed a potential threat to obtaining accurate (i.e., truthful) responses. Human subjects have demonstrated a propensity to report "socially desirable" responses in surveys, presumably to reduce or avoid personal discomfort, embarrassment, etc. (Cannell et al. 1977; Sudman and Bradburn 1982). A number of strategies were used in the design and administration of the present study to minimize this problem.

The specific wording of the instructions and questions was a major consideration in the design of the survey instrument. Following the guidelines set forth by Sudman and Bradburn (1982), the questions were carefully worded so as to minimize the potential personal threat to the respondents. For example, printed instructions on the questionnaire included sentences such as "We are interested in your thoughts about a number of things people your age often do," ant, "Below are listed a number of situations that teenagers often experience." Threatening terms such as "Stealing" were avoided in direct reference to the respondent in favor of phrases such as "... happen to have taken from the store without paying...".

Several strategies were also implemented in the administration of the survey to promote accurate responses. First, according to prior arrangement, the subjects' teachers were asked to leave the classrooms during the procedure. The procedure commenced with a brief informal introductory speech explaining the purpose of the study in general terms. The speech emphasized the subjects' anonymity and the voluntary nature of the exercise. Several carefully constructed arguments for responding as honestly as possible were presented. (e.g., "Since You are asked not to put your name on the questionnaire, no one will know who you are.") Next, subjects were given ample opportunity to ask questions about the study and the procedure before they fillet out the questionnaires; also, they were encouraged to ask questions during the session. Subjects were told that they were free to terminate their participation at any time. Completed questionnaires were returned by the individual subjects to an enclosed box. After all questionnaires had been collected, a volunteer was solicited from among the subjects to scramble the order of the returned questionnaires in the box in plain view of the others. Finally, the box was closet and the subjects were thanked for their cooperation.

An informal debriefing was held with a subsample of participants. No significant apprehension about the survey on the part of the subjects was noted. Upon probing, no subjects could think of any way that responses could be "traced back" to an individual.

Dependent Variables

Motivations for Shoplifting. Social, economic, and sporting motivations for shoplifting were measured by asking subjects to respond to statements measured on a 5-point "very much agree" (5) to "very much disagree" (1) scale. The statements were designed to assess 'reasons why teens may steal things". Many of the items used were similar to those used in previous research (Klemke 1982). The items making up these scales are listed in Table 1. Social motivations consisted of five items which loaded on the hypothetical factor. The alpha reliability coefficient of this five-item scale was .89. Economic motivations were measured using seven items designed to measure various "rational aspects of shoplifting. The reliability coefficient of this scale was .72. Sporting motivations were measured using three items which loaded on the hypothetical factor designed to measure the game-like" aspect of shoplifting. the alpha reliability of this scale was .82. The motivations for shoplifting are reliable with alphas of .70 or higher. These motivations also have discriminant validity in that on a factor analysis, each of the items included in the 3 motivations loaded highly (.5 or above) on the hypothetical factors of social, economic and sporting motivations, while the other items not included in that scale loaded at .2 or below. Also, the items on these scales had been used by previous researchers (Klemke 1982).

Attitudes toward shoplifting were measured by summing responses to seven items measured on a S-point "very much agree" (5) "very much disagree" (1) scale designed to measure cognitive and affective orientations toward the act of shoplifting. A typical item of this scale was "Shoplifting is a crime." The alpha reliability coefficient of this scale was .85. Shoplifting frequency was measured by asking respondents to indicate on a 3-point "several times"-"never" scale how often they had taken each of ten specific items from a store without paying during the past year. The items were: candy/sweets, records/tapes, sporting equipment, clothing, health items, school supplies, books/magazines, toys, drugs/alcohol, and cigarettes. These items were suggested by previous research to be among the most frequently purchased products, and therefore the most likely to be shoplifted by teenagers. The reliability coefficient among those who hat shoplifted at least one item during the past year (n-53) was .79.

Independent Variables

Frequency of communication with peers about shoplifting was measured by summing responses to five point "very often" (5) - "never" (1) scale designed to measure cognitive and overt interaction about shoplifting. A sample item was: "my friends and I talk about stealing things." The alpha reliability coefficient for the scale was .83.



Socio-orientation communication structure was measured in line with previous research by asking respondents to indicate the frequency of parent-child communication style stressing obedience, deference, and family harmony (cf. McLeod and O'Keefe 1972; Moschis forthcoming). Responses to six items designed to measure the construct were measured on a five point "very often" (5) - "never" (1) scale. A typical item of this scale was: "(Parents) say you (child) may not buy certain things." The reliability coefficient was .60.

Family size was measured by the number of family members (siblings). Absence of parents was a measure as follows: l-respondent lives with both parents; 2-respondent lives only with one parent; 3-respondent does not live with his/her parents. While recognizing the potential drawbacks of this operationalization, pragmatic considerations led us to favor this simple measure. More complex measures might be used to tap the underlying construct more precisely. However, this would have unduly lengthened the instrument and the potential contribution was not deemed sufficient to merit their inclusion. Socio-economic status was measured using Duncan's (1961) scale of occupations.


Table 2 shows descriptive statistics and correlations among all variables used. Table 3 shows the part: al correlations used to test the hypotheses.

Sporting motivations for shoplifting was positively associated d the age (r - .03) and frequency of communication with peers about shoplifting (r - .01) as posited. However, the level of statistical significance was insufficient to support Hypotheses 1 and 2b. The anticipated relationship between social motivations for shoplifting and peer communication was confirmed (r-. 16; p<.05) (Hypothesis 2a).

Frequency of communication with peers about shoplifting was also found to be positively and significantly related to both favorability of attitudes towards shoplifting (r = .20; p < .01) and frequency of shoplifting (r = .57; p < .001). Thus Hypotheses 3a and 3b were supported by the data.

The expected influence of family size was found to be positive as stated in Hypotheses 4a, 4b, and 4c. Specifically, the association between family size and shoplifting frequency was significant (r=.24, p< .05), which supports Hypothesis 4c. the hypothesized relationship between family size and social motivations was not statistically significant, while the relationship between family size and shoplifting attitudes approached significance (r=  .14, p<.06).

The degree of absence of parents from the home was also positively associated with social motivations for shoplifting (r = .13; p < .06), favorabllity of attitudes- toward shoplifting (r = .14; p < .06), and frequency of shoplifting (r = .14; n.s.). These findings offer only minimal support for Hypotheses 5a, 5b and 5c, respectively.

Socio-orientation in family communication was positively but insignificantly associated with social motivations for shoplifting as posited (r = .05) (Hypothesis 6a). Socio-oriented communication, however, was negatively associated with both favorability of attitudes toward shoplifting (r = -.10) and shoplifting frequency (r = -.20). Not only was the nature of the associations too weak to support Hypothesis 6, but, in the specific cases of 6b and 6c, the association was in the opposite direction.

Finally, social class was positively associated with economic motivations for shoplifting (r = .20; p < .01), shoplifting attitudes (r = .08; n.s.), and shoplifting frequency (r = .07; n.s.). These results are counter to what was hypothesized, providing no support for Hypotheses 7a, 7b, and 7c.




The data presented in this study should be interpreted with extreme caution due to the small sample size used and the location from which they were drawn. The main purpose of the research was to present a conceptual framework and some empirical findings that could be used in order to more systematically investigate delinquent behaviors (e.g., shoplifting) in future research. The exploratory findings of the study suggest some interesting possibilities for future research. One of the areas which deserves further investigation pertains to the role of peer influence on the development of shoplifting behaviors. Unfortunately, in cross sectional studies (such as ours) it is difficult to determine the direction of causality between frequency of communication wi the peers about shoplifting and shoplifting frequency. Longitudinal designs are much more suitable for answering such questions. Another direction for research concerns the role of motivations for shoplifting, and more specifically the antecedents and consequences of such motivations. Our data suggest that much of shoplifting may be done for social reasons, and those shoplifting for such reasons do not necessarily hold positive cognitions toward shoplifting. Thus, it would be useful to develop an understanding of the sequence in which motives, attitudes, and behaviors develop, including their antecedents and consequences on undesirable consumer behaviors. This study represents a preliminary venture into the "dark side" of consumer behavior. The conceptual framework developed herein might be applied to the study of other undesirable consumptive behaviors, such as littering, immoderate or over-consumption of certain products (eg. unhealth snack foots), compulsive shopping, buying, or gambling, and chronic overextension of Personal financial resources.

This research has demonstrate i the influence of peers and social situations on adolescent shoplifting behavior as well as the relevant motivations and attitudes that result in a deviant consumption behavior. Hopefully, this framework can serve as a base from which to examine this and other deviant consumption behavior, ant, more importantly, help develop programs that will curtail these deviant Practices.




Beck, Esther Ann and Sherwood C. McIntyre (1977), "MMPI Patterns of Shoplifters Within A College Population," Psychological Reports, 41, pp. 1035-1040.

Brady, J.F. and J.G. Mitchell (1971) "Shoplifting in Melbourne," Australian and New Zealand Journal of Criminology, 4 (September) pp. 154-162.

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

Campbell, Ernest Q. (1969), "Adolescent Socialization in D.G. Goslin (ed.) Handbook of Socialization Theory and Research. Chicago: Rant McNally, pp. 821-859.

Cannell, Charles F., Lois Oksenberg, and Jean M. Converse (1977), "Striving For Response Accuracy: Experiments in New Interviewing Techniques," Journal of Marketing Research, 14 (August), pp. 306-315.

Churchill, Gilbert A. Jr. and Georg P. Moschis (1979), "Television and Interpersonal Influences on Adolescent Consumer Learning," Journal of Consumer Research. (June). p. 2.

Duncan, Otis (1961) "A Socioeconomic Index of Occupations": in Occupations and Social Status ed Albert J. Reiss, Jr., New York: The Free Press.

Forbes (1981), March 16, p. 11.

Gold, M. (1970) Delinquency in an American City, Belmont, California: Brooks/Cole Publishing Company.

Klemke, L.W. (1978), "Does Apprehension for Shoplifting Amplify or Terminate Shoplifting Activity?" Law and Society Review, 12 (Spring), pp. 391-403.

Klemke, Lloyd W. (1982) "Exploring Juvenile Shoplifting; Sociology and Social Research, 67 (1), 59-75.

McLeod, Jack M. and Garrett J. O'Keefe (1972) "The Socialization Perspective and Communication Behavior", in F.G. Kline and P.J. Tichenor (eds.) Current Perspectives in Mass Communication Research, Beverly Rills, California Sage Publications. pp. 121-168.

Messenger, Carol (1975), "Losses Hit $5 Billion Annually," Stores (October), p. 25.

Moschis, George (forthcoming) Consumer Socialization, Boston, M.A.: Lexington Books.

Powell, Judith and George P. Moschis (1986), -A Study of Juvenile Shoplifting Behavior," Working Paper, Georgia State University.

Stores (1971), "NRMA's and INAE's National Campaign to Fight Shoplifting," (November), p.6.

Robbin, G.D. (1963) "Patterns of Department Store Shoplifting" Crime and Delinquency, pp. 163-172.

Security Management (1978), "Shoplifting Keeps Pace with Inflation," (July) p. 27.

Sudman, Seymour and Norman M. Bradburn (1974), Asking Questions, San Francisco: Jossey-Bass.

Sutherland, E.H. (1937) The Professional Thief, Chicago: The University of Chicago Press.

Taylor, L.B. (1979), Shoplifting, New York: Franklin Watts, Inc.

Velocci, Tony (1978) "Can Business Collar White-Collar Crime?" Nation's Business, 66(November) 35-44.

Verill, A.H. (1978) Reducing Shoplifting Loses, (Revised Edition) Washington, D.C.: Small Business Administration.

Wright, Thomas L. and Afshan Kirmani (1971), "Interpersonal Trust, Trustworthiness and Shoplifting in High School: Psychological Reports, 41, pp. 1165-1166.

Won, George and George Yamamoto (1976), Social Structure and Deviant Behavior: A Study of Shoplifting," Sociology and Social Research, 53 (October) pp. 44-55.



George P. Moschis, Georgia State University
Dena Saliagas Cox, Georgia State University. Georgia State University
James J. Kellaris


NA - Advances in Consumer Research Volume 14 | 1987

Share Proceeding

Featured papers

See More


Causes and Consequences of the Expense Prediction Bias

Chuck Howard, University of British Columbia, Canada
David Hardisty, University of British Columbia, Canada
Abigail Sussman, University of Chicago, USA
Melissa Knoll, Consumer Financial Protection Bureau

Read More


Paying to Be Social? How Materialism Shapes Spending on Friends

William Ding, Washington State University, USA
David Sprott, Washington State University, USA
Andrew Perkins, Washington State University, USA

Read More


M3. #FOMO: How the Fear of Missing Out Drives Consumer Purchase Decisions

Michelle van Solt, Florida International University
Jessica Rixom, University of Nevada, Reno
Kimberly Taylor, Florida International University

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.