Understanding factors associated with online piracy behaviour of adolescents

Abstract

This study aimed at evaluating self-control theory in the context of online music piracy by adolescents in Hong Kong. In addition to self-control measures, social variables and strain variables were measured in this study with 634 secondary school students. The results did not strongly support the general theory of crime's claim that the impact of self-control is universal across gender and cultures. Individuals' self-control was correlated with boys' online piracy behaviour, but not that of girls. A stronger influence of social factors (differential association) compared with self-control on online piracy behaviour was also observed. This research contributes to knowledge of the interplay of gender and culture in shaping the applicability of self-control theory in understanding delinquent behaviour.

Keywords:

• online piracy behaviour
• adolescents
• general theory of crime

Introduction

The music industry's profits have been declining. One of the reasons for this may be the advances in Internet and compression technologies that have enabled people to circulate large collections of copyrighted music files over the World Wide Web. Although a growing body of literature concerning software piracy (Glass & Wood, 1996; Gopal & Sanders, 1997; Moores & Dhillon, 2000) has emerged, relatively little work has addressed online music piracy. Music files are much smaller than computer software, and so people find it even easier to distribute the files freely over the web (Bhattacharjee, Gopal, Lert Wachara, & Marsden, 2003). This study aimed at evaluating Gottfredson and Hirschi's (1990) general theory of crime – self-control theory – and, on the basis of the theory, investigating the underlying factors associated with adolescents' propensity to obtain copyrighted music files online illegally.

Self-control theory

Gottfredson and Hirschi's (1990) theory of self-control is one of the most popular theories in the field of crime and deviance (Pratt & Cullen, 2000; Vazsonyi, Pickering, Junger, & Hessing, 2001). The theory proposes that individuals' self-control determines the extent to which people commit to criminal behaviour. They claimed that self-control is a stable trait, and people who possess this trait are more likely to be self-centred, impulsive, impatient, and have a preference towards simple tasks over complicated ones that require persistence. As a consequence, people with low self-control would find it difficult to delay the immediate gratification associated with deviant acts. Research has consistently demonstrated that low self-control is significantly associated with a range of criminal and deviant behaviour in different populations, including: self-reported and official delinquency among school students (Cochran et al., 1994; Mason & Windle, 2002; Nakhaie et al., 2000; Polakowski, 1994; Vazsonyi et al., 2001) and street youth (Baron, 2003); future criminal convictions in adolescents (Polakowski, 1994); academic cheating, class-cutting and intimate violence among university students (Sellers, 1999); imprudent acts (such as, drinking, smoking, and gambling) in general adult populations (Burton, Cullen, Evans, Alarid, & Dunaway, 1998; Wright, Caspi, Moffitt, & Silva, 1999) and adult criminal offenders (Longshore, Turner, & Stein, 1996); as well as victimisation in violence (Stewart, Elifson, & Sterk, 2004). In a recent meta-analysis, Pratt and Cullen (2000) found that low self-control had a mean size effect of 0.257 across 21 empirical studies.

Although there is a growing empirical literature showing support for the general self-control theory, some issues of the theory need further investigations, such as the explanatory power of the theory when other perspectives are also considered, the propositions regarding the invariant influence of gender and culture on deviance after self-control is controlled, and the extent to which self-control theory can be applied in the context of Internet crime. To examine these issues, this study aimed to evaluate the strength of self-control theory in the prediction of online music piracy behaviour among Chinese adolescents in Hong Kong.

Self-control and social learning perspectives

Social learning theory holds that relationships between individuals and their significant others restrain them from deviant behaviour by exerting informal social control (Hirschi, 1969). However, Gottfredson and Hirschi (1990) did not support the view that socialisation or culture is a major causal factor in deviant behaviour. Instead, they argued that people with low-self control are more likely to put themselves in environments that facilitate delinquent behaviour. Previous studies examining these propositions gave mixed support (Burton et al., 1998; Mason & Windle, 2002; Polakowski, 1994; Pratt & Cullen, 2000). For instance, Wright et al. (1999) supported Gottfredson and Hirschi's (1990) view that low self-control predicts broken adolescent and adult social bonds (with family, school, peers, employment, and marriage) and criminal offending later in life. In a recent meta-analysis, Pratt and Cullen (2000) demonstrated that deviant peers and deviant attitudes remained significant predictors of crime independent of low self-control. Evans, Cullen, Burton, Dunaway, and Benson (1997) reported that only a few social bonds significantly influence criminal behaviour after controlling for self-control. Mason and Windle (2002) found that gender may moderate the effects of self-control and informal social control on adolescent delinquency. They observed that low self-control was associated with both major and minor delinquency among boys through informal social control variables, whereas for girls low self-control was significantly associated with their major delinquency, but it only indirectly influenced minor delinquency through the informal social control variables.

Self-control and strain perspectives

From the perspective of strain theory, negative life events, negative relations with parents, and negative school experiences can precipitate adolescents' delinquency (Agnew, 2001). Gottfredson and Hirschi (1990) addressed strain perspectives in ways similar to how they treated the social learning perspective. They argued that individuals with low self-control are more likely to provoke negative treatment by other people. Previous research comparing both perspectives is sparse and has provided inconsistent views on the controversy. For instance, Burton et al. (1998) found that low self-control was associated with adult crime, whereas the effect of strain became insignificant after controlling for self-control variables. In Baron's (2003) study, however, the findings are not consistent with those of Burton et al. (1998). Baron (2003) found that both low self-control and strain were significantly associated with criminal behaviour of homeless street youth. In the face of the under-researched and inconsistent findings regarding the relative impact of strain, self-control, and social variables on adolescents' delinquent behaviour, this study attempted to compare these three perspectives simultaneously in the context of online music piracy.

Self-control and gender

One major component of Gottfredson and Hirschi's (1990) theory is the claim that individual differences in self-control can explain variability in crime and analogous behaviour. They make a similar claim regarding gender differences in criminal involvement, contending that the gender gap in crime and delinquency results primarily from gender differences in self-control. Thus, girls are less deviant than boys because they have greater self-control. Burton et al. (1998) found that, after controlling for self-control, the gender difference in criminal behaviour was eliminated. Similarly, LaGrange and Silverman (1999) observed that self-control reduced an association between gender and delinquent behaviour among secondary school students. However, Gibbs et al. (1998) demonstrated a direct effect of gender on non-criminal deviance (including alcohol use, cheating, and cutting classes); when self-control was made constant, gender exerted an indirect effect on this behaviour. Although this research shows some support for self-control theory, it is still not clear whether and how self-control explains the associations between gender and deviance. This study aims at evaluating the role of self-control in accounting for the gender differences in online music piracy, if any. Furthermore, since different factors may influence males and females in different ways, this study also examined how self-control, strain and social variables interact in predicting the delinquent behaviour of male and female adolescents.

Cultural influences and delinquent behaviour

Gottfredson and Hirschi argued that ‘cultural variability is not important in the causation of crime, that we should look for constancy rather than variability in the definition of and causation of crime …’ (1990, p. 175). Self-control was seen by these authors as a mechanism that can predict any cultural differences observed in all kinds of crime and deviance. Most of the research evaluating self-control theory has been conducted in American samples, and very few non-American samples have been investigated. Caspi et al. (1994) and Wright et al. (1999) provided support for the relationship between self-control and delinquency in New Zealand. In Canada, LaGrange and Silverman (1999) found support for the negative association between self-control and general delinquency, property offenses, violence, and drug offenses. The self-control–deviance relationship was also supported in Spain (Romero, Gomez-Fraguela, Luengo, & Sobral, 2003) and Russia (Tittle & Botchkovar, 2005). In a large-scaled cross-cultural study on self-control in Hungary, the Netherlands, Switzerland and the United States, Vazsonyi et al. (2001) showed that low self-control accounted for about 20–25% variance in total adolescent deviance in all four countries, and contended that the effect of self-control on deviance is largely invariant by national context. However, some research has shown the power of self-control is less consistent in eastern countries. Vazsonyi, Clifford Wittekind, Belliston, and Van Loh (2004) demonstrated a self-control–delinquency relationship in Japan, whereas Hwang and Akers (2003) showed that in Korea the effect of low self-control did not hold when the effects of social, learning variables were taken into account. Wang, Qiao, Hong, and Zhang's (2002) study on Chinese adolescents' deviance indicated that self-control theory might not be applicable in China. They found no evidence of direct relationships between self-control and delinquency. Given these inconsistencies, there is a need for further examination of the generalisability of self-control theory in the Chinese culture.

The current study

This study conducted a comprehensive evaluation of self-control theory in the context of online music piracy by adolescents in Hong Kong. In addition to self-control measures, we also incorporated measures of social variables and strain variables in the study. On the basis of self-control theory and previous findings, we predicted that males would demonstrate more piracy behaviour than females, and that self-control would account for the differences. In addition, owing to the importance of social forces in Chinese culture, we predicted that social variables would be more powerful than self-control variables in predicting adolescents' delinquent behaviour. This study undertook further tests of the possible variations of how different factors influence boys and girls in downloading music illegally from the Internet. It was hoped that through this research endeavour a comprehensive picture of elements associated with online intellectual property theft would be obtained.

Method

Participants

The participants were recruited from secondary school students through random sampling in Hong Kong. Self-reported questionnaires were administered to the students in their classrooms. All students filled out in the questionnaire during one of the regularly scheduled class periods. The sample consisted of 634 secondary students, 290 of whom were female (45.7%), 314 were male (49.5%), and 30 (4.7%) did not provide information about their gender. The participants were aged between 15 and 19, with a mean age of 16. All students had high-speed access to the Internet at home.

Measures

Three sets of independent variables (self-control, differential association, and strain) were used in the present analysis. The dependent variable was the extent to which adolescents participated in online music piracy.

Low self-control

Six items from the scale created by Grasmick, Tittle, Bursik, and Arneklev (1993) were used to assess self-control. A sample item was ‘I often do what brings me pleasure here and now’ and participants responded on a five-point Likert scale, with higher scores indicating higher self-control. As the general theory of crime treats self-control as a unitary concept, we used a composite measure in this study (see also Grasmick et al., 1993; Polakowski, 1994). Moreover, although previous research (Arneklev et al., 1993; Grasmick et al., 1993; Wood, Pfefferbaum, & Arneklev, 1993) has demonstrated through factor analysis that separate factors exist within the trait of self-control, our analysis of the items failed to identify distinct factors using exploratory factor analysis. This scale was found to have good reliability, with alpha = 0.81.

Strain

Six questions from Broidy's (2001) study were used to measure stressful life experiences. These items asked the participants to reflect on the last six months and indicate whether they had stressful life experiences; for example, doing badly academically. They gave their responses according to a five-point Likert scale, with higher scores indicating a greater level of strain. One factor structure was shown from an exploratory factor analysis, and this scale had good reliability with alpha = 0.75.

Differential association

Eight items constructed by the researcher were used to measure differential association. These items included ‘My friends support my MP3 usage’ and ‘I associate with others online who exchange MP3 with me’. Participants responded on a five-point Likert scale, with higher scores indicating higher level of association. One factor structure was shown from an exploratory factor analysis, and this scale achieved good reliability with alpha = 0.84.

Dependent variable

The dependent variable was the extent to which adolescents participated in music piracy. Respondent were asked to indicate how many MP3 files they had personally downloaded that week, how many they downloaded in an average week and in an average month, and how many they had downloaded this year. These variables were factor-analysed using promax solution. One factor structure resulted with alpha = 0.92, indicating that if a respondent answered one of these questions in a certain way, it was likely that she/he answered the other questions in the same way.

Results

Independent t-tests were performed on each independent variable and the dependent variable to see whether there was any gender difference. No significant gender difference was found for all variables: strain, t(602) = − 1.70, p>0.05; self-control, t(601) = − 0.15, p>0.05; differential association, t(592) = 0.74, p>0.05; music piracy, t(600) = 0.49, p>0.05. Separate hierarchical regression analyses were undertaken separately for boys and girls to assess the generality of self-control theory across gender. Table 1 and 2 presents the findings of full models for boys and girls, respectively.

Table 1 Hierarchical regression analyses showing association of self-control, strain, and differential association with music piracy behaviour for girls.

	Model 1		Model 2		Model 3
Variable	β (SE β)	B	β (SE β)	B	β (SE β)	B
Age	0.124 (0.048)	0.156**	0.128 (0.047)	0.162**	0.132 (0.047)	0.167**
Parental income	0.073 (0.035)	0.035*	0.099 (0.036)	0.169**	0.100 (0.036)	0.170**
Self-control			− 0.059 (0.081)	− 0.045	− 0.006 (0.083)	− 0.005
Strain			0.087 (0.032)	0.167**	0.077 (0.032)	0.147*
Differential association					0.219 (0.101)	0.148*
R ^2	0.119		0.149		0.173
ΔR ^2	0.119***		0.30*		0.24*
ΔF	11.01***		4.28*		3.01*
Note: SE, standard error. *p < 0.05. **p < 0.01. ***p < 0.001.

Table 2 Hierarchical regression analyses showing association of self-control, strain, and differential association with music piracy behaviour for boys.

	Model 1		Model 2		Model 3
Variable	β (SE β)	B	β (SE β)	B	β (SE β)	B
Age	0.095 (0.034)	0.150**	0.093 (0.033)	0.147**	0.076 (0.033)	0.119*
Parental income	0.098 (0.031)	0.170**	0.138 (0.032)	0.238***	0.150 (0.032)	0.258***
Self-control			− 0.209 (0.075)	− 0.154**	− 0.182 (0.073)	− 0.134*
Strain			0.042 (0.027)	0.094	0.020 (0.026)	0.045
Differential association					0.379 (0.086)	− 0.258***
R ^2	0.227		0.264		0.317
ΔR ^2	0.227***		0.037***		0.053***
ΔF	27.5***		7.09***		10.65***
Note: SE, standard error. *p < 0.05. **p < 0.01. ***p < 0.001.

As shown in Table 1, after controlling for demographic variables, we demonstrated that strain, but not self-control, was significantly associated with piracy behaviour for girls (β = 0.087, p < 0.01). For boys, we found that self-control, but not strain, was significantly associated with their piracy behaviour after the demographics were controlled (β = − 0.209, p < 0.01). We can also observe from Tables 1 and 2 that differential association explained an additional variance of the piracy behaviour significantly for both boys and girls (β = 0.379, p < 0.001 for boys; β = 0.219, p < 0.05 for girls).

Discussion

This research explored and compared the strengths of self-control, strain, and social factors in influencing online piracy behaviour among secondary school adolescents in Hong Kong. Contrary to most previous research, our analysis indicated that boys had a similar level of online piracy behaviour to that of girls, and boys and girls demonstrated comparable levels of self-control. Moreover, our study has extended previous work by showing that low self-control was significantly correlated with more piracy behaviour for boys, but not for girls. Whereas a higher level of strain was significantly associated with more piracy behaviour for girls, this association was not significant for boys. In addition to examining the relationships of strain and self-control variables to online piracy behaviour, our study also demonstrated that differential association was significantly related to online piracy behaviour for both boys and girls. The findings of this research may challenge the applicability of self-control theory in understanding factors associated with online piracy in a collectivistic culture.

Gender difference in online piracy behaviour

Previous studies indicated that males had higher rates of delinquent behaviour than females, with the gender gap in crime being highest for serious violent and property crimes and lowest for minor property crime, drug use, and escapist behaviour like running away (Canter, 1981; Chesney-Lind & Shelden, 1992). However, the current study did not demonstrate gender difference in online piracy behaviour. This result could be interpreted in terms of individuals' opportunities for online piracy and the non-significant difference in self-control between boys and girls. Several prominent and influential theorists have attributed gender differences in crime and delinquency to differences in opportunity (Adler, 1975; Simon, 1975). Females were traditionally relegated to uniquely feminine roles that kept them at home, or more closely supervised at school and at work. They were thus less likely to engage in ‘drinking, stealing, gang activity, and fighting’ (Adler, 1975, p. 95) because they had fewer opportunities to do so. However, because everyone can now gain access to the Internet and circulate materials almost freely over the Internet, it is not surprising to see girls behaving similarly to boys. In fact, early studies based on female arrests for serious crimes suggested that as females gained greater freedom and wider social participation, their involvement in crime would also increase and converge with that of males (Adler, 1975; Simon, 1975).

In addition to equal opportunities for boys and girls to use the Internet, we found no gender difference in self-control. This result might be attributed to the influence of Chinese culture on self-control and delinquent behaviour. Self-control appears to be more emphasised in collectivistic Chinese societies than in individualistic western societies (Wang et al., 2002). Young people growing up in Chinese societies attach value to collectivism. Submissiveness, obedience, and discipline are much valued and are taught to young people from their childhood onwards (Leung & Fan, 1996). Individual autonomy is often sacrificed for collective interests, especially those of the family (Chan & Lee, 1995). Thus, young people who grow up in a Chinese culture may acquire a higher level of self-control than their western counterparts. Empirical research that compares Chinese and American adolescents has indicated cultural differences, in that Chinese youth are more reserved and habituated to self-restraint (Song, 1995). One might argue that if levels of self-control are so high in both boys and girls in Chinese culture, we might not be able to predict deviance in the same manner as we do in the West. Indeed, no gender difference in self-control was found in this study, which is not in line with our initial predictions.

Generality of self-control theory across gender

One purpose of this research was to assess the generality of the theory across gender – that is, would the trait of low self-control affect similarly both boys' and girls' online piracy behaviour? Some researchers have contended that the determinants of crime and deviance – most notably strain, self-control, and differential association – affect both men and women in similar ways (Rosenbaum, 1987; Smith & Paternoster, 1987); and there is a body of evidence suggesting that they have comparable effects across genders (Cerkovich & Giordano, 1979; Giordano, Cerkovich, & Pugh, 1986; Paternoster & Triplett; 1988). Contrary to previous research, our regression analyses did not show support for the generality of self-control's effects across boys and girls.

For boys, we found that low self-control was significantly associated with more online piracy behaviour after controlling for other factors. Although social factors were more associated with online piracy than self-control, the basic tenets of self-control theory still hold relevance for boys in our study. Self-control emerges as one of the important factors distinguishing boys who commit online piracy from those who do not. However, in girls, strain and social factors, but not self-control, were significantly associated with online piracy behaviour. Boys and girls may have different responses to strain. There is some evidence that girls tend to internalise reactions to cope with their stress, whereas boys tend to externalise their reactions by becoming aggressive or getting involved in property offenses (Aneshensel, 1992; Leadbeater, Blatt, & Quinlan, 1995). Instead of engaging in serious misconduct, girls may resort to downloading music from the Internet as one of the coping resources to relieve their stress. In general, the present study demonstrated results that only partially support the culture-free thesis of self-control theory, which may only work for boys in Chinese communities.

Importance of social factors: differential association

Another important finding of this study is the important role of social factors (differential association) in adolescents' online piracy behaviour. We observed a culture-specific pattern of associations, in that social factors play a more significant role than self-control in influencing Chinese adolescents' delinquency. Research has consistently demonstrated that friends' delinquency is a very strong predictor of an individual's delinquency. In the present study, whereas self-control and strain affected online piracy behaviour differentially for boys and girls, differential association could explain the behaviour for both genders in this study. Thus, the behaviour of Chinese adolescents may be more directly influenced by social forces than by individual traits in a collectivistic cultural setting. Although Hong Kong is a modernised city that has been heavily influenced by western culture for many years, Chinese individuals still place a high value on Chinese tradition, and interpersonal ties have a strong influence on people's behaviour (Chan & Lee, 1995; Cheung & Cheung, 2006). In terms of delinquency, familial and school relationships are important sources for providing social support and exerting social control against deviant behaviour among Hong Kong adolescents (Leung & Fan, 1996). The general self-control theory would predict no difference in the relationship between self-control and deviance in different cultural settings. In contrast to previous studies conducted in the United States and Europe, the findings observed in this study challenge this culture-free approach to deviant behaviour. Moreover, providing additional support to previous literature, we found that self-control could not adequately explain the relationships between social factors and delinquency (Baron, 2003; Burton et al., 1998; Evans et al., 1997; Pratt & Cullen, 2000; Wright et al., 1999).

Conclusion

Taken together, findings of the current research on adolescents in Hong Kong do not strongly support the general theory of crime's claim that the impact of self-control is universal across gender and cultures. In Hong Kong, as a Chinese society, while self-control is correlated with boys' online piracy behaviour, it is not related to girls' piracy behaviour. We also observed the stronger influence of social factors (differential association) compared with self-control on online piracy behaviour in our study. This research represents one step closer to understanding the interaction of gender and culture in shaping the applicability of self-control theory in understanding delinquent behaviour.

There are some limitations to our study. First, our hypotheses were evaluated with cross-sectional data, and thus we could not assess the causal relationships among various factors and deviant behaviour. Future investigations should rely on a longitudinal design to track individuals at different points in time to determine the causal nature. Second, a growing body of delinquency research pinpoints the influence of school contexts on delinquent behaviour (for example, Hoffmann, 2006; Hoffmann & Xu, 2002). In the future, researchers may recruit a larger sample size and use hierarchical linear modelling to analyse the school-contextualised effects on delinquent behaviour. Third, more direct comparisons between western and Chinese societies should be conducted to further fine-tune the general theory of crime.