An exploration of the relationship between MP3 player ownership and digital piracy

Abstract

A great deal of research has explored the impact of technology on human behavior, particularly the emergence of the Internet and computer technology to facilitate digital piracy. Few studies have, however, considered how portable digital music, or MP3, players facilitate or reduce involvement in piracy. This study explores the relationship between digital ownership and participation in digital piracy among a population of college students. The findings suggest that MP3 player ownership is significantly connected to piracy, along with deviant peer connections. In addition, there are significant differences in the beliefs and attitudes of owners and non‐owners toward online deviance and piracy.

Keywords:

• digital piracy
• MP3 players
• technology
• cybercrime
• techniques of neutralization

A great deal of research has explored the impact of technology on human behavior (Bryant, 1984; Forsyth, 1986; Holt, 2007; Melbin, 1978; Ogburn, 1932; Quinn & Forsyth, 2005). Individuals adapt their norms and behaviors in response to scientific and technological innovations. Eventually, new forms of behavior may supplant old practices resulting in behavioral shifts referred to as ‘technicways’ (Odum, 1937; Parker, 1943; Vance, 1972). Understanding technicways has significant value for criminologists, as offenders change their patterns of behavior due to evolving technologies (Quinn & Forsyth, 2005). For example, pagers, cellular telephones, and the Internet are increasingly used by prostitutes to attract and solicit customers (Holt & Blevins, 2007; Lucas, 2005). Embossing, scanning, and printing technologies have also been employed to improve the quality and volume of counterfeit credit cards (Mativat & Tremblay, 1997) and in developing counterfeit currency (Morris, Copes, & Perry‐Mullis, 2009).

Recently, the development of the Internet and computer technology has facilitated the growth of a variety of crimes, particularly digital piracy, or the illegal copying of digital media such as computer software, digital sound recordings, and digital video recordings without the explicit permission of the copyright holder (Gopal, Sanders, Bhattacharjee, Agrawal, & Wagner, 2004). Such files can be easily downloaded from one of many Internet file sharing services or websites and commonly do not stem from a single user (e.g., torrent sites). ¹ The financial losses estimated to result from digital piracy are staggering and participation levels in this illegal activity are commonplace, particularly among college students (Hinduja, 2001, 2003; Ingram & Hinduja, 2008; Morris & Higgins, 2009; Rob & Waldfogel, 2006; Zentner, 2006).

One of the most prominent pieces of technology that may facilitate piracy is the portable digital music player, or MP3 player. Products such as iPods are extremely popular among consumers and can store large volumes of digital media. Thus, the widespread use of these products may be an enabling or debilitating factor in illegal file sharing among individuals. Another issue is that regularly occurring advancements and price reductions for this technology make macro‐level assessments of the extent of digital piracy rather short on external validity. This study addresses this issue through an initial exploration of the relationship between digital media player ownership and participation in illegal file sharing among college students in the USA. The findings provide insights on the role of technology in facilitating criminal behavior, and attitudes toward deviance.

The financial impact of digital piracy

The illegal duplication of copyrighted material is not a new phenomenon. For example, Liebowitz (1985) explored the impact of Xerox copying on library journal purchases and suggested that photocopying actually increased revenue for copyright holders. In fact, the financial benefit to the industry as a whole in the long run is known as the exposure or sampling effect in economic theory. ² Likewise, as noted by Liebowitz (2005, 2006), the birth of home video recording spawned an entirely new industry, pre‐recorded video sales, that ultimately produced profits far in excess of box office ticket sales. The present‐day sound recording and motion picture industries have not experienced the creation of equivalently successful markets as a result of affordable and efficient digitization of copyrighted material. ³ Since 1999, the sound recording industry has experienced drastic declines in record sales that correspond to the time that file sharing witnessed widespread popularity through services such as Napster (Liebowitz, 2006). The reason behind such losses is largely attributed to increased file sharing among individuals who would otherwise purchase the media legitimately (Liebowitz, 2006). In fact, by 2003, the recording industry had filed law suits against several hundred heavy downloaders (Recording Industry Association of America [RIAA], 2003).

Empirical evidence from the economics literature is strongly suggestive of the idea that file sharing is financially harmful to copyright holders (Liebowitz, 2006; Michel, 2006; Rob & Waldfogel, 2006; Zentner, 2006). For example, Siwek (2007) reported that the US sound recording industry loses over $12 billion annually due to piracy. In addition, the US government loses $422 million each year in tax revenue that would have been generated via corporate and personal income taxes (Siwek, 2007). IDATE (2003) has suggested that illegal file sharing accounts for over 4 times the amount of official sales of sound recordings worldwide. The same report suggested that peer‐to‐peer (P2P) file sharing accounts for between 50 and 90% of all broadband Internet traffic in any given day, depending on the time of day. The motion picture industry has reported similar losses. The Motion Picture Association of America (MPAA) reported fiscal losses upwards of $6 billion in 2005 from movie piracy in the USA alone. Over 40% of these reported losses were argued to be a result of university students in the USA (MPAA, 2007).

Criminological explorations of piracy

In light of the significant economic impact of digital piracy, research has attempted to identify the prevalence of file sharing/digital piracy, especially among college students (e.g., Hinduja, 2001, 2003; Ingram & Hinduja, 2008; Morris & Higgins, 2009; Rob & Waldfogel, 2006; Zentner, 2006). These studies suggest that the large numbers of file sharers may be due in large part to improved technology becoming more affordable and widespread, increases in readily available high‐speed Internet access, and the lasting popularity of movies and music as desirable forms of entertainment.

Recent criminological explorations on piracy have attempted to identify the sociological and structural individual level predictors of engaging in the behavior. The bulk of these studies find that males tend to engage in piracy at higher rates (Hinduja, 2001, 2007; Ingram & Hinduja, 2008; Kini, Ramakrishna, & Vijayaraman, 2003; Oz, 1990; Paradice, 1990; Peace, Galletta, & Thong, 2003; Rahim, Seyal, & Rahman, 2001; Ramakrishna, Kini, & Vijayaraman, 2001; Solomon & O’Brien, 1990). Studies that have evaluated variables stemming from social learning theory (Akers, 1985, 1998) have been highly suggestive of the idea that one’s own piracy is highly contingent on the piracy of one’s peers (Higgins, Fell, & Wilson, 2006; Higgins & Makin, 2004a, 2004b; Hinduja & Ingram, 2008; Ingram & Hinduja, 2008; Morris & Higgins, 2009; Skinner & Fream, 1997). In addition, those who have developed definitions that support or neutralize responsibility for piracy behavior are more likely to illegally download media (Higgins et al., 2006; Higgins & Makin, 2004a, 2004b; Hinduja & Ingram, 2008; Ingram & Hinduja, 2008; Morris & Higgins, 2009; Rahim et al., 2001). Attitudinal, or cognitive, low self‐control, as defined by Gottfredson and Hirschi (1990), has also been found to be predictive of digital piracy among college students (Higgins, 2005; Higgins et al., 2006; Higgins & Makin, 2004a; Hinduja & Ingram, 2008 – for an exception see Morris & Higgins, 2009).

Though the explanatory power of behavioral theories in explaining digital piracy may depend on the type of piracy being measured (see Morris & Higgins, 2009), the preponderance of evidence suggests that several popular theories of crime may apply to individual digital piracy. However, to our knowledge, no studies to date have directly explored the relationship between MP3 player ownership and participation in digital piracy. As such, it is unclear how the popularity of such devices has impacted digital piracy behaviors at the individual level. While this possibility may apply more to music and video piracy than to software piracy, the affordability and popularity of such devices may be related to pirating behaviors and it is thus worthy of empirical exploration. In fact, Apple Inc. reported strong gains in iPod sales in 2008 with over 11 million iPods sold in the fourth quarter alone (Apple Inc., 2008). This study adds to the current body of literature by exploring the relationship, if any, between digital media player ownership and participation in digital piracy while controlling for several individual level and structural covariates. In addition we explore any variation in the ethical attitudes and beliefs about piracy and computer deviance between MP3 player owners and non‐owners to understand the impact of this technology.

Data and methods

Data for this exploration of piracy were collected from a larger project examining college students’ computer activities, perceptions, and beliefs. Respondents completed a self‐report survey at a southeastern university in 10 courses during the fall of 2006. Five of these courses allowed students from any college to enroll, increasing the representative nature of the sample. The sample (n = 605) was relatively balanced between males (43%) and females (57%) and was predominantly white (78.1%), in keeping with the larger university demographic population (52.5% female and 75% white). This purposive sample was developed because of college students’ involvement in and knowledge of risky behavior regarding piracy and the use of technological devices (see Higgins, 2005; Hinduja, 2003; Skinner & Fream, 1997).

Dependent variable

Respondents were asked how many times within 12 months prior to the completion of the survey they had used, made, or otherwise given pirated media to another person (options being never, 1–2 times, 3–5 times, 6–9 times, and 10 or more times). This measure was collapsed into a dichotomous measure (0 = no piracy; 1 = piracy) and a binary logistic regression was employed. A total of 301 (49.8%) respondents reported engaging in media piracy at some point within the last year, reflecting similar statistics from existing research (see Higgins, 2005; Hinduja, 2003).

Independent variables

Multiple measures of respondents’ exposure to technology, computer use, and attitudes toward technology were included in the instrument to assess potential exposure to and use of technology (Table 1). A self‐assessment item was included to measure the respondents’ knowledge, skill, and computer use (skill level). This three category index was based on general categories of computer proficiency: (1) I can surf the ‘net, use common software, but not fix my own computer (normal); (2) I can use a variety of software and fix some computer problems I have (intermediate); and (3) I can use Linux, most software, and fix most computer problems I have (advanced). Respondents were asked how many computers they owned (computer ownership) (0 = none; 1 = one or two; 2 = three to five; 3 = six to nine; 4 = 10 or more computers). Internet connectivity was a dichotomous variable, with high‐speed connectivity as the reference category to assess the ability to rapidly pirate a variety of media which has been a predictor of pirating behavior (e.g., Hinduja, 2003). Finally, respondents were asked if they owned an iPod or MP3 player, with MP3 player ownership as the reference category (non‐ownership = 1).

Table 1. Sample descriptives (N = 605).

Variables	Mean	Std. dev.	Min	Max
Dependent variable
Digital piracy	0.4975	0.50041	0	1
Technology use
Computer ownership	0.9983	0.30176	0	4
Internet connectivity	0.9521	0.21380	0	1
Skill level	0.6727	0.56849	0	2
MP3 player ownership	0.6017	0.48996	0	1
Hours work/school	1.5521	1.36877	0	4
Hours non‐work/school	1.6083	1.38593	0	4
Offending, definitions, and beliefs
Peer piracy	1.5372	1.28698	0	4
Clear rules online	3.1521	0.85546	1	4
Learned ethics	3.4512	0.72766	1	4
No illegal behavior	2.9124	0.92102	1	4
Negative impact	2.6975	0.99473	1	4
Never snitch	3.3289	0.77582	1	4
Pirate to get friends	1.8033	0.84335	1	4
OK pirate few songs	2.5504	0.98642	1	4
OK no money lost	2.17179	0.94038	1	4
Demographics
Employment	0.8248	0.60274	0	2
Age	0.8529	0.88578	0	4
Race	0.2198	0.41448	0	1
Gender	0.5769	0.49447	0	1

To directly assess the amount of time they spend using a computer daily, respondents were asked to indicate how many hours per week in the last year have they spent on a computer for both work or school (Hours work/school) and also outside of work or school (Hours non‐work/school), with options being less than 5 hours, 5–10 hours, 11–15 hours, 16–20 hours, and 21 or more hours. The frequency distribution of these two measures appear similar, although they are not highly correlated (Spearman = 0.259). Therefore, they are measuring two distinct aspects of how computer usage is integrated into their daily routines.

Peer involvement in piracy was also measured to consider the impact from friends’ piracy on one’s own piracy. Respondents were specifically asked ‘how many of your friends have knowingly used, made, or gave to another person pirated media’ (options being: 0 = none of them; 1 = very few of them; 2 = about half of them; 3 = more than half of them; 4 = all of them).

Eight items were also included to explore a range of attitudes toward piracy and computer ethics adapted from Skinner and Fream (1997). Students were asked to indicate their agreement with these items, with responses ranging from (1) strongly disagree to (4) strongly agree. Four items were included to assess individual beliefs about ethical and illegal activity online: (1) there are clear rules on what is acceptable ethical behavior online (clear rules online); (2) I have learned what is acceptable and ethical concerning computers and online behavior (learned ethics); (3) I would never do anything illegal with a computer because it is against the law (no illegal behavior); (4) I would not engage in unethical computer activities because it may negatively impact the way I am viewed by others (negative impact). In addition, four neutralizing definitions were assessed through the following items: (1) I would never turn in a friend who pirated software or media (no snitch); (2) I see nothing wrong in giving people copies of pirated materials to foster friendships (pirate to get friends); (3) it is OK for me to pirate music because I only want one or two songs from most CDs (OK pirate few songs); and (4) it is OK for me to pirate media because the creators are really not going to lose any money (OK no money lost).

Finally, we examine whether sex, race, age, and employment affect the risk of engaging in piracy. Gender is a dichotomous measure (0 = male; 1 = female). We include race as a dummy variable where non‐white is equal to 1. ⁴ Age is a 4‐point ordinal scale (0 = 19 and under; 1 = 20–21; 2 = 22–25; 3 = 26 and up) while employment status is measured as a 3‐point ordinal scale (0 = unemployed; 1 = part time/temp; 2 = full time).

Analytical plan

In order to address our research questions, we developed an analytical strategy that took place in three steps. First, we developed binary logistic regression models predicting participation in digital piracy. Second, we stratified the sample based on reported MP3 player ownership and ran equivalent binary logistic models. Here, we relied on a comparative Z‐test to determine whether significant differences in coefficient values existed between the two groups. Finally, we developed a binary model predicting iPod ownership including all of the independent variables as the previous models but including digital piracy prevalence as a predictor. This analytical framework was taken to provide a more robust exploration of this cross‐sectional data, as we could not formally assess whether iPod ownership preceded piracy. The following section presents the findings from each of these analyses.

Findings

Binary model predicting piracy

The results of the binary logistic regression models predicting digital piracy are presented in Table 2. The analysis indicated that peer involvement in piracy is significantly related to individual involvement in digital piracy, which is consistent with previous studies (e.g., Higgins, 2005; Morris & Higgins, 2009; Skinner & Fream, 1997). Though some studies have found a relationship between definitions that support involvement in piracy (Higgins,2005; Ingram & Hinduja, 2008; Skinner & Fream, 1997), the only measure of definitions that is significant in this model is the perception that computer deviance is acceptable despite laws to curb these behaviors.

Table 2. Logistic regression predicting piracy (n = 605).

Variables	B	SE	Odds ratio
Technology use
Computer ownership	0.878	0.396	2.405∗
Internet connectivity	−0.124	0.214	0.883
Skill level	−0.502	0.214	1.652∗
MP3 player ownership	0.501	0.214	1.652∗
Hours work/school	−0.06	0.09	0.941
Hours non‐work/school	−0.024	0.088	0.977
Offending, definitions, and beliefs
Peer piracy	1.082	0.116	2.952∗∗∗
Clear rules online	1.53	0.152	1.165
Learned ethics	−0.084	0.172	0.919
No illegal behavior	−0.549	0.152	0.587∗∗∗
Negative impact	0.084	0.129	1.088
Never snitch	0.168	0.154	1.183
Pirate to get friends	−0.008	0.159	0.992
OK pirate few songs	0.266	0.172	1.305
OK no money lost	0.175	0.182	1.191
Demographics
Employment	0.043	0.191	1.044
Age	0.061	0.136	1.063
Race	0.523	0.267	1.688∗∗∗
Gender	0.051	0.245	0.951
Constant	−3.424
Pseudo R ^2	0.53
∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001.

Experience with technology also increased the odds of reporting media piracy. Specifically, owning multiple computers and reporting greater skill with computer technology and the Internet increased the risk of offending. Internet connectivity and time spent online were not, however, significant in the model. This suggests that an understanding of technology is necessary to increase the likelihood of piracy rather than just the amount of time one spends online. In addition, owning an MP3 player (or iPod) increased the odds of self‐reported piracy. Though the impact is not as large as that of peer involvement in piracy, this may be an important risk factor that has not been explored in previous studies on this form of crime.

The findings in regard to the structural correlates on both self‐reported piracy bear mention as well. In the full model, race was found to significantly impact reporting piracy. Reporting a race other than white was found to positively influence the odds of reporting piracy, which is consistent with some previous research (Hinduja, 2003; Skinner & Fream, 1997). No other demographic controls were found to impact the likelihood of piracy, thus suggesting that piracy is driven by larger sociological forces.

Partitioned model

In light of the significance of MP3 player ownership on the likelihood of engaging in piracy, we partitioned the model to examine any variation in the behaviors and perceptions of owners vs. non‐owners (see Table 3). Following the recommendation of Paternoster, Brame, Mazerolle, and Piquero (1998), we calculated Z‐statistics to determine significant between‐group differences based on MP3 player ownership. The findings suggest there are significant group differences in attitudinal and perceptual indicators of participation in piracy. ⁵

Table 3. Logistic regression models predicting piracy for MP3 player owners and non‐owner subsamples.

	MP3 player owners (n = 364)	Non‐MP3 player owners (n = 241)	Z‐Value
Variables	B	Std. error	Odds ratio	B	Std. error	Odds ratio
Technology use
Number of computers	0.509	0.549	0.859	1.864	0.702	6.448∗∗	−1.520
Internet connectivity	0.101	0.773	1.106	−0.746	0.712	0.474	0.805
Skill level	0.617	0.278	1.854∗	0.440	0.381	1.553	0.375
Hours work/school	−0.110	0.122	0.895	−0.042	0.148	0.959	−0.354
Hours non‐work/school	0.010	0.113	1.010	−0.101	0.152	0.904	0.586
Offending, definitions, and beliefs
Peer piracy	1.057	0.150	2.878∗∗∗	1.266	0.209	3.547∗∗∗	−0.812
Clear rules online	−0.051	0.198	0.950	0.582	0.268	1.790∗	−1.899∗
Learned ethics	−0.125	0.244	0.883	−0.139	0.272	0.870	2.256∗
No illegal behavior	−0.356	0.202	0.701	−0.948	0.271	0.3380∗∗∗	1.751∗
Negative impact	−0.048	0.173	0.954	0.274	0.221	1.315	−1.147
Never snitch	0.195	0.194	1.216	0.178	0.28	1.194	0.049
Pirate to get friends	0.061	0.207	1.063	−0.228	0.278	0.796	0.833
OK pirate few songs	0.500	0.244	1.650∗	−0.007	0.318	0.993	1.264
OK no money lost	−0.096	0.241	0.909	0.636	0.319	1.888∗	1.830∗
Demographics
Employment	0.007	0.247	1.007	0.065	0.341	1.067	−0.137
Age	−0.150	0.184	0.860	0.328	0.228	1.389	−1.631
Race	0.714	0.358	2.042∗	0.367	0.451	1.444	0.067
Gender	0.007	0.316	1.007	−0.093	0.445	0.044	0.183
Constant	−2.301			1.681			1.875
−2 Log likelihood	0.577			0.514
∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001.

Peer involvement in piracy behaviors were significant predictors for piracy across both groups, suggesting deviant peers may have a key role in this form of cybercrime. There were also no significant differences in the use or exposure to technology across these two groups. The belief that there were clear rules about online behavior was, however, a significant predictor of piracy for those who did not own an MP3 player, and was significant between these two groups. In addition, those who did not own an MP3 player reported that they were willing to engage in illegal computer activities regardless of laws to limit these behaviors. This may reflect a perception that there are ethical activities and guidelines online, though piracy does not violate these norms for those who do not own MP3 players.

The view that piracy is acceptable because the victims lose no money was also found to have a significant positive relationship with piracy for non‐MP3 player owners. This same measure had a negative impact for MP3 player owners but was not statistically significant. Thus, this could also mean that, on average, it might take a stronger level of neutralization to initiate piracy when a digital music player is not owned. In addition, this suggests that iPod owners may be beyond justifying digital piracy.

Predicting MP3 player ownership

In an effort to extend our exploration of piracy, we developed a model with iPod ownership as the outcome variable and self‐reported piracy as a predictor (see Table 4). As the data is cross‐sectional and not time ordered, we must cautiously interpret the true impact of these findings within the model. There were no significant impacts from individual piracy and peer piracy on owning an MP3 player. Rather, technology exposure appears to have a strong affect on MP3 player ownership. Specifically, those who own more computers and spend time online for personal reasons appear more likely to own an MP3 player. Those who are not employed are more likely to own MP3 players, in keeping with evidence on computer deviance and free time online (see Hinduja, 2003; Holt & Bossler, 2009).

Table 4. Logistic regression predicting MP3 player ownership.

Variables	B	Std. error	Odds ratio
Technology use
Computer ownership	0.625	0.322	3.771∗
Internet connectivity	0.660	0.417	2.509
Skill level	0.195	0.171	1.302
Hours work/school	0.029	0.071	0.164
Hours non‐work/school	0.223	0.070	10.052∗∗
Offending, definitions, and beliefs
Individual piracy	0.147	0.084	3.057
Peer piracy	0.050	0.095	0.280
Clear rules online	−0.006	0.120	0.003
Learned ethics	−0.045	0.142	0.099
No illegal behavior	−0.004	0.126	0.001
Negative impact	0.035	0.103	0.115
Never snitch	0.023	0.124	0.034
Pirate to get friends	0.032	0.130	0.061
OK pirate few songs	−0.129	0.138	0.875
OK no money lost	0.060	0.146	0.171
Demographics
Employment	−0.391	0.153	6.496∗
Age	−0.195	0.110	3.132
Race	−0.473	0.210	5.5055∗
Gender	−0.044	0.198	0.049
Constant	−3.424	1.225	7.816∗∗
Pseudo R ^2	0.125
∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001.

It must be noted that no neutralizing or moral measures were significant in the model predicting MP3 player ownership. The significance of race may also reflect similar trends in computer deviance among minority groups (see Hinduja, 2003; Skinner & Fream, 1997). Taken as a whole, the findings suggest that individuals may not purchase MP3 players as a consequence of their involvement in piracy. Instead, owners may be more likely to illegally download digital materials after purchase.

Discussion and conclusions

A growing body of research has considered the impact of new technologies on deviant behavior and offending practices, particularly computer technology. These studies provide important insights into the ways that these devices facilitate illegal behavior. None, however, have considered the role of MP3 players on participation in digital piracy. Thus, this study sought to explore any particular relationship between MP3 player ownership, attitudes toward piracy, and involvement in pirating behaviors. Primary to our study goals, we found that individuals whose peers engage in piracy, own multiple computers, and report greater skill with computers were more likely to engage in piracy. MP3 player ownership also had a significant impact on self‐reported digital piracy among respondents, suggesting the adoption of this specific form of technology affects involvement in criminal behavior.

Given the significance of MP3 players on piracy, we partitioned the analysis to explore whether any factors influencing piracy may be mediated by MP3 player ownership. Significant variation was found between MP3 player owners and non‐owners based on attitudes toward piracy. In particular, non‐MP3 player owners perceived there to be clear ethical rules for behavior in online environments, though they felt that it was acceptable to engage in illegal computer behaviors regardless of existing legal statutes. In addition, non‐MP3 player owners were significantly more likely to support the belief that piracy was acceptable because their actions caused no monetary losses. Thus, piracy may be a more acceptable behavior for non‐MP3 player owners than for those who own such devices.

Another important finding from this study is the key role of associating with peers who engage in digital piracy (e.g., Higgins, 2005; Hinduja, 2001, 2007; Ingram & Hinduja, 2008). The impact of peer involvement remained significant when partitioned on MP3 player ownership, though the impact was attenuated for those who did not own such technology. Taken as a whole, these findings indicate that individuals may not necessarily purchase an iPod as a result of having friends who pirate media. Associating with pirating peers does affect whether one pirates music themselves, regardless of whether they own an iPod. Thus, iPod ownership is not a requirement for piracy; many will burn CDs or listen to music while on the computer rather than using a portable music player. In all, if we assume that our data were time ordered, it could be that piracy prevalence may be more likely to follow the purchase of a portable digital music player rather than stimulate the purchase of a music player. Additional research is necessary to determine whether this finding holds true in general.

Further research is needed to expand our knowledge from these initial findings and generalization from the findings presented here should be made with caution. The cross‐sectional nature of this data did not allow us to identify the time sequence and impact of MP3 player ownership on piracy offending over time. Furthermore, by exploring the prevalence of offending it is not clear what differences exist in the amount of materials that MP3 player owners and non‐owners pirate. Thus, longitudinal data examining the onset and prevalence of piracy behaviors, as well as the initial purchase of an MP3 player are needed to better explicate the relationships initially identified here. Such findings will improve our understanding of the nature of computers and technology in offending behavior in the twenty‐first century.

Notes on contributors

Thomas J. Holt, PhD is an Assistant Professor in the School of Criminal Justice at Michigan State University. His research focuses on cybercrime and the ways that technology and the Internet facilitate deviance. He has published in several journals, including Deviant Behavior and the Journal of Criminal Justice, and is one of the founding members of the Spartan Devils Chapter of the international Honeynet Project.

Robert G. Morris, PhD is an Assistant Professor of Criminology at the University of Texas at Dallas. He studies the etiology of crime with a specific interest in fraud and cybercrime as well as issues surrounding the social response to crime. His recent work has appeared in Criminal Justice Review, Journal of Criminal Justice, Journal of Crime and Justice, Deviant Behavior, Criminal Justice & Popular Culture, and Criminal Justice Policy Review.

Notes

1. We use the terms file sharing and digital piracy interchangeably for the purposes of this research.

2. For a review of other economics‐based theories that may apply to file sharing and music purchases, and their limitations, see Liebowitz (2006). For an in‐depth review of economics‐based findings on the impact of file sharing on record sales, see Liebowitz (2005).

3. The closest exception may be in the individual purchase of songs through outlets such as iTunes, however, some research suggests that this has not had a positive impact on recently declining album sales (Leibowitz, 2006). It may be too early to tell if such enterprises represent a viable market respective of illegal file sharing.

4. Respondents were asked to identify themselves as white, African‐American, Hispanic, Asian, or another racial group. Asians, Hispanics, and other race categories only comprise 5.2, 2.8, and 3.1% of the respective sample. Thus these race groups were combined into one category to simplify the model.

5. We determined whether significant group differences existed using the formula: where a result exceeding ±1.64 (one‐tailed) was significant at p < 0.05 and a result exceeding ±2.33 was significant at p < 0.01 (see Paternoster et al., 1998).