 

This is to certify that the
dissertation entitled

AN EMPIRICAL TEST OF THE APPLICABILITY
OF GENERAL CRIMINOLOGICAL THEORIES
TO INTERNET CRIME

presented by
SAMEER HINDUJA

has been accepted towards fulﬁilment
of the requirements for the

Ph.D. degree in Criminal Justice

 

Major Professor’s Signature

Date

MSU is an Afﬂnnative Action/Equal Opportunity Institution

 

 

 

LIEﬁRAdRY'
Michigan State
University

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

CT 31 20:29

 

.0 TAjUg 3112014

 

 

.2.

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClFiCIDateDuopss-sz

 

An Empirical Test of the Applicability of General Criminological Theories
to Internet Crime

By

Sameer Hinduja

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
School of Criminal Justice

2004

ABSTRACT

An Empirical Test of the Applicability of General Criminological Theories
to Internet Crime

By

Sameer Hinduja

Accompanying his seminal examination of white collar crime in the early 20‘h
century, Edwin Sutherland (1947; 1973) maintained that crime theorists should not solely
focus on explaining deviance among the lower class, but should expand their
paradigmatic scope and seek to explain a wide range of crimes committed by a wide
range of offenders. Three general theories - Agnew’s (1992) General Strain Theory,
Gottfredson and Hirschi’s (1990) Self-Control Theory, and Akers’ (1985) Social
Learning Theory - have been posited since Sutherland’s assertion, and should by
deﬁnition have the predictive capacity to explain nontraditional types of wrongdoing
such as Internet crime. Indeed, the respective authors of each theory have speciﬁcally
stated that their perspective is panoptic and all-inclusive. As of yet, however, no research
has sought to test the “generality” of general strain, self-control, and social learning
theory and by studying a decidedly nontraditional form of deviance occurring in a
decidedly nontraditional environment. Intellectual property theﬁ on the Internet is one
such illegality. Concomitant with the rapid global advance into an information-based
society and economy, the critical role that intellectual property plays in the stability,
vitality, and growth of private sector companies, public sector organizations, and even
individual lives demands that it is secured and precluded from misappropriation,

exploitation, and manipulation for criminal purposes. Through its analysis of the piracy

of digital music, the current research seeks to contribute to the criminological literature
base by examining and testing the purported universal applicability of these three general
criminological theories. It also seeks to obtain a deeper understanding of the
phenomenon at hand, and shed light on commonalities and differences inherent in

Intemet-based criminality as compared to traditional forms of crime.

Coryn'ght by
SAMEER HINDUJA
2004

To Him.

vi

ACKNOWLEDGEMENTS

I have learned that intellectual and social support is foundational in the success of
any graduate student. It is needed almost every day. And these days blur into nights,
which blur into weeks and weekends, which blur into months and years. And here I am
reﬂecting upon this season of my life, and the upon many quality individuals who were
instrumental in my attainment of this goal.

To begin, God has been, and remains, my everything - and this would not be
possible without Him. You have made my paths straight, and You have blessed the work
of my mind and my hands. All of this points to Your reality, faithfulness, and provision.
Mom and Dad, your unconditional love, support, and prayers have known no bounds, and
have kept you close despite the geographic distance. Thank you for cultivating discipline
in me and a constant desire to learn, and for teaching me about the unmatched importance
of family. Babita, you are my biggest fan, and your love has been a constant source of
stability and reassurance in my life.

Justin Patchin (Patch), I am honored to have you as my friend. We've gone
through this together, and have kept each other inspired and motivated. I am indebted to
you in so many ways, and couldn't have hoped for a better companion with whom to run
this race. Thank you for your personal and practical support. I hope we have left an
impression - not because of what we did, but because of how we are. Amanda Burgess-
Proctor (Amandabear), we've shared so many experiences over these years, and my life
has been enriched by your fantastic storytelling, your thoughtfulness, and your warmth. I

have learned so much from you in terms of both style and substance. Beth Huebner

vii

(Betharoo), you have made this journey easier because of your willingness to offer sage
advice on graduate school matters, and your insightful comments on scholarly works. I
have enjoyed (and now miss greatly) our chats about the highs and lows of our career
choice. More than anyone else, you have served as the model of academic success that I
have sought to emulate while in graduate school.

Certain professors have also contributed greatly to my development as a scholar.
Mahesh Nalla, I have appreciated your undivided attention when we met to discuss my
work, and for always broadening my perspective of possible correlates of criminological
phenomena. You have been an exceptional mentor, chairperson, and ﬁ'iend. Christina
DeJong, thank you for the time and energy you devote to investing in your students, and
for your methodological and statistical expertise. Tim Bynum, I am so grateful to have
had the opportunity to work for you over these years, and to partake in unparalleled
research opportunities. Thank you also for your mentorship, wise counsel, and for going
to bat for me. Sandeep Kulkarni, I have beneﬁted ﬁom your advice related to the
integration of criminal justice and computer science. Karen Busch, your insight into
Actor-Network Theory has introduced me to a novel approach in studying any
phenomenon. John McCluskey, being a Yankees fan might be the only obvious fault you
have. Thank you for your perceptive comments and feedback on many projects. Mark
Lanier, thank you for convincing me to go to MSU - it was the best decision I ever made.
Thank you also for keeping in touch and guiding me along the way. Each of you have
supported my research interests and have challenged me to expand traditional

disciplinary boundaries.

viii

Two former members of Team Bynum deserve acknowledgement. Joe Schafer,
thank you for being a ﬁne colleague and a hilarious friend. Sean Varano, thank you for
providing great advice over the years as I navigated through the degree program. Thanks
to both of you for also giving me grief about so many things, particularly when I
deserved it. A handful of current members of Team Bynum have also added value to my
life at MSU. Ryan Martz, Cheryl Reid, Alison Blair, Catharine White, Jason Ingram,
Eric Grommon, and Karen Ream - you have provided the perfect compliment to the
stresses and pressures of graduate school by being a part of many fond memories on a
day-to-day basis. Though I will miss you, I take comfort in knowing that the legacy of
Team Bynum will live on. Other graduate students have contributed to both small and
large successes in my professional life. These include Kristy Holtfreter, Yan Zhang,
Cedrick Heraux, Pete Blair, Brandon Kooi, Jeff Cancino, Meghan Stroshine, Amanda
Robinson, Matt Mamer, and Michael Alper. I am thankful for the friendship that each of
you have provided, and it has been a pleasure to work with you.

Pastor Mark Evans, you have seen me grow both professionally and personally
during this season, and you have believed in me and what I am trying to do with my life.
Thank you for seeing my heart and inspiring me towards greatness. Christian Gendreau,
I have appreciated your unique and intelligent perspective on things, and your invaluable
ability to make me laugh out loud all the time. I am grateﬁrl that we have known each
Other for so long. Tammy Dahl, you have sent so many encouragements this way, and
they have always come at the perfect time. Finally, I would like to thank the

M("Pherson’s the Strombeck’s, the Lynch’s, and the Bryan’s. You know how much you

mean to me.

ix

TABLE OF CONTENTS

CHAPTER 1: INTRODUCTION .................................................................................................................... 1
PREMISE ...................................................................................................................................................... I
PRACTICAL BACKDROP OF RESEARCH ........................................................................................................ 3
THEORETICAL BACKDROP OF RESEARCH .................................................................................................... 6
NECESSITY FOR RESEARCH AND RESPONSE ................................................................................................ 8
VALUE AND CONTRIBUTION OF THE RESEARCH ........................................................................................ 1 1
ORGANIZATION OF SUCCEEDING TEXT ..................................................................................................... 13

CHAPTER 2: REVIEW OF THE LITERATURE ......................................................................................... 15
HISTORICAL BACKGROUND OF DIGITAL AUDIO FILES .............................................................................. 15
NIP3 TECHNICAL BACKGROUND ............................................................................................................... 16
NAPSTER - A REVOLUTIONARY APPLICATION .......................................................................................... 22
PLAYERS IN THE W3 PHENOMENON ........................................................................................................ 25
REVIEW OF MP3-RELATED RESEARCH ..................................................................................................... 28
THEFT, LAW, AND AMBIGUITY — HISTORY AND RELEVANCE .................................................................... 38
THE VALUE OF INFORMATION AND INTELLECTUAL PROPERTY ................................................................. 40
COPYRIGHT LAW ...................................................................................................................................... 45

US. Copyright Law {Title I 7 US. C. Section 101 et seq.. Title I 8 US. C. Section 2319} ................... 48
Fair Use Doctrine ................................................................................................................................ 49
The Audio Home Recording Act (AHRA) of 1992 ............................................................................... 49
The No Electronic Theft Act (NET) ..................................................................................................... 50
The Digital Performance Right in Sound Recordings Act (DPRA) of 1995 ........................................ 50
The Digital Millennium Copyright Act (DMCA) of I 998 ............ 2 ....................................................... 51
CIVIL AND CRIMINAL SANCTIONS RESULTING FROM COPYRIGHT VIOLATION ......................................... 52

CHAPTER 3: THEORETICAL BACKGROUND AND APPLICATION ................................................... 57

GENERAL STRAIN THEORY ....................................................................................................................... 57
Empirical Support for General Strain Theory ..................................................................................... 59
General Strain Theory Applied To The MP3 Phenomenon ................................................................. 66

SELF-CONTROL THEORY ........................................................................................................................... 7 0
Empirical Support for Self-Control Theory ......................................................................................... 74
Self-Control Theory Applied To The MP3 Phenomenon ..................................................................... 80

SOCIAL LEARNING THEORY ...................................................................................................................... 85
Empirical Support for Social Learning Theory ................................................................................... '89
Social Learning Theory Applied T o The MP3 Phenomenon ............................................................... 94

CHAPTER 4: METHODOLOGY ............................................................................................................... 108
POPULATION ........................................................................................................................................... 108
INSTRUMENT ........................................................................................................................................... 109
SAMPLING PROCEDURES ......................................................................................................................... 1 13
PRETEST .................................................................................................................................................. l 16
SPECIFICATION OF VARIABLES OF INTEREST ........................................................................................... 1 17

Strain ................................................................................................................................................. 11 7
Self-Control ....................................................................................................................................... 11 7
Control Variables .............................................................................................................................. I 18
Social Learning ................................................................................................................................. l l 9
Dependent Variables ......................................................................................................................... [26
HYPOTHESES ........................................................................................................................................... l 29

CHAPTER 4: ANALYSES ......................................................................................................................... 130

DESCRIPTIVE STATISTICS ........................................................................................................................ 131
BIVARIATE STATISTICS ........................................................................................................................... 137
Analysis of Variance .......................................................................................................................... I 4 l
MULTIVARIATE STATISTICS .................................................................................................................... I44
Ordinary Least Squares (OLS) Regression ....................................................................................... I 44
Limitations of Ordinary Least Squares (OLS) Regression ................................................................ 146
Logistic Regression ........................................................................................................................... I4 7
Multinomial Logistic Regression ....................................................................................................... l 5 6
CHAPTER 5: DISCUSSION ...................................................................................................................... 169
CHAPTER 6: LIMITATIONS .................................................................................................................... 17 7
LIMITATIONS OF GENERAL STRAIN THEORY ........................................................................................... 177
LIMITATIONS 0F SELF-CONTROL THEORY .............................................................................................. I79
LIMITATIONS OF SOCIAL LEARNING THEORY .......................................................................................... 180
METHODOLOGICAL LIMITATIONS ........................................................................................................... 182
CHAPTER 7: POLICY IMPLICATIONS ................................................................................................... I91
POLICY IMPLICATIONS OF SELF-CONTROL THEORY’S RELEVANCE...........I ............................................. 191
POLICY IMPLICATIONS OF SOCIAL LEARNING THEORY’S RELEVANCE .................................................... 194
EMBRACING A NEw BUSINESS MODEL ................................................................................................... 195
COPY PROTECTION SCHEMES .................................................................................................................. 199
COGNITIVE RESTRUCTURING INITIATIVES ............................................................................................... 209
SHAPING MORALITY To COHERE WITH LEGALITY .................................................................................. 215
CHAPTER 8: THE HETEROGENEOUS ENGINEERING OF INTERNET CRIME ................................ 220
THE SOCIAL AND THE TECHNICAL .......................................................................................................... 220
AGENCY AND STRUCTURE ...................................................................................................................... 223
ACTOR-NETWORK THEORY .................................................................................................................... 225
APPLICATION TO DIGITAL MUSIC PIRACY ............................................................................................... 229
ACTOR-NETWORK THEORY AND CHARACTERISTICS ............................................................................... 240
FURTHER APPLICATION OF ACTOR-NETWORK THEORY ......................................................................... 243
IMPLICATIONS ......................................................................................................................................... 249
SUMMARY ........................................................................................................................................... 253
CONCLUSION ........................................................................................................................................... 255
APPENDICES ............................................................................................................................................ 257
REFERENCES ............................................................................................................................................ 276

xi

LIST OF TABLES

TABLE 1: STRAIN FACTOR SCORE ................................................................................................................ 1 17
TABLE 2: ATTITUDINAL SELF-CONTROL FACTOR SCORE ............................................................................ 1 18
TABLE 3: BEHAVIORAL SELF-CONTROL FACTOR SCORE ............................................................................. 1 18
TABLE 4: DIFFERENTIAL ASSOCIATION FACTOR SCORE .............................................................................. 121
TABLE 5: IMITATION/MODELING FACTOR SCORE ........................................................................................ 121
TABLE 6: DEFINITIONS FACTOR SCORE ........................................................................................................ 122
TABLE 7: DIFFERENTIAL REINFORCEMENT FACTOR SCORE ......................................................................... 122
TABLE 8: MUSIC PIRACY DEPENDENT VARIABLE FACTOR SCORE ............................................................... 127
TABLE 9. DEMOGRAPHIC CHARACTERISTICS AND PARTICIPATION IN PIRATING ......................................... 134
TABLE 10. DISTRIBUTION OF MUSIC PIRACY VARIABLES ........................................................................... 136
TABLE 1 I. BIVARIATE CORRELATION MATRIX OF VARIABLES ................................................................... 140
TABLE 12: ANOVA RESULTS BETWEEN DEMOGRAPHIC VARIABLES AND .................................................... 142
OVERALL MUSIC PIRACY (ADDITIVE MEASURE) ............................................................................................ 142
TABLE 13. ANOVA: PROFICIENCY AND VARIETY OF INTERNET USE AND OVERALL MUSIC PIRACY ......... 144
TABLE 14. OLS REGRESSION: IV FACTOR SCORES ON OVERALL PIRACY FACTOR SCORE ......................... 145

TABLE 15. LOGISTIC REGRESSION FOR MP3 FILES DOWNLOADED IN LAST WEEK, LAST MONTH, SINCE
BEGINNING OF 2003, AND ON AVERAGE PER MONTH (N=2032) .................................................................... 151

TABLE 16. LOGISTIC REGRESSION FOR MP3 FILES DOWNLOADED IN AVERAGE WEEK AND MONTH, BOTH ONE
YEAR AGO AND Two YEARS AGO (N=2032) ................................................................................................. 152

TABLE I7. LOGISTIC REGRESSION FOR MP3s DOWNLOADED IN 2002, 2001, 2000, TOTAL COMPLETE MP3

ALBUMS, AND MP3s DOWNLOADED DURING ONE’S LIFE (N=2032) ............................................................. 153
TABLE 18. MULTINOMIAL LOGISTIC: How MANY MPBS DOWNLOADED IN THE LAST WEEK? ...................... 159
TABLE 19. MULTINOMIAL LOGISTIC: HOW MANY MP3s DOWNLOADED IN THE LAST MONTH? ................... 159

TABLE 20. MULTINOMIAL LOGISTIC: HOW MANY MP3s DOWNLOADED SINCE THE BEGINNING OF 2003?... 160
TABLE 21. MULTINOMIAL LOGISTIC: HOW MANY MPBS Do YOU DOWNLOAD ON AVERAGE PER MONTH? .. 160

TABLE 22. MULTINOM IAL LOGISTIC: HOW MANY MP3s DOWNLOADED IN AN AVERAGE WEEK EXACTLY ONE
YEAR AGO? .................................................................................................................................................. I61

xii

TABLE 23. MULTINOMIAL LOGISTIC: HOW MANY MP3s DOWNLOADED IN AN AVERAGE MONTH EXACTLY
ONE YEAR AGO? ........................................................................................................................................... 161

TABLE 24. MULTINOMIAL LOGISTIC: How MANY MP3s DOWNLOADED IN AN AVERAGE WEEK EXACTLY TWO
YEARS AG0? ................................................................................................................................................. 162

TABLE 25. MULTINOMIAL LOGISTIC: HOW MANY MP3s DOWNLOADED IN AN AVERAGE MONTH EXACTLY

TWO YEARS AGo? ......................................................................................................................................... 162
TABLE 26. MULTINOMIAL LOGISTIC: HOW MANY MP3s DID YOU PERSONALLY DOWNLOAD IN 2002? ........ 163
TABLE 27. MULTINOMIAL LOGISTIC: HOW MANY MP3s DID YOU PERSONALLY DOWNLOAD IN 2001? ........ 163
TABLE 28. MULTINOMIAL LOGISTIC: How MANY MP3s DID YOU PERSONALLY DOWNLOAD IN 2000? ........ 164

TABLE 29. MULTINOMIAL LOGISTIC: HOW MANY TOTAL COMPLETE MUSIC ALBUMS IN MP3 FORMAT HAVE
YOU OBTAINED ONLINE? .............................................................................................................................. 164

TABLE 30. MULTINOMIAL LOGISTIC: HOW MANY TOTAL MP3s HAVE YOU DOWNLOADED OVER THE COURSE
OF YOUR LIFE THUS FAR? ............................................................................................................................. 165

xiii

CHAPTER 1: INTRODUCTION
Premise

For centuries, scholars and thinkers have sought to explain why individuals
engage in criminal behavior. This has resulted in a number of theoretical paradigms that
concentrate on singular (e. g.,Hirschi, 1969) or multiple factors (e.g., Cohen & Felson,
1979) that inﬂuenced our understanding of the onset, incidence, and perpetuation of
deviance, delinquency, or crime. Moreover, a few “general” criminological theories
have been professed and reﬁned in the last two decades, each ostensibly seeking to
predict and explain variation in all types of deviant behavior (e.g., academic dishonesty,
substance abuse, domestic violence, and embezzlement). These include Agnew’s (1992)
General Strain Theory (GST), Gottfredson and Hirschi’s (1990) Self-Control Theory
(SCT), and Akers’ (1985) Social Learning Theory (SLT).

Accompanying his seminal examination of white collar crime in the early 20th
century, Edwin Sutherland (1947; 1973) maintained that crime theorists should not solely
focus on explaining deviance among the lower class, but should expand their
paradigmatic scope and seek to explain a wide range of crimes committed by a wide
range of offenders. The three general theories mentioned earlier have been posited since
Sutherland’s assertion, and should by deﬁnition have the predictive capacity to explain
nontraditional types of wrongdoing such as Internet crime. Indeed, the respective authors
of each theory analyzed in this work have speciﬁcally stated that their perspective is
panoptic and all-inclusive.

The applicability of these theories to traditional forms of wrongdoing has been

empirically explored by a host of social science researchers since their initial

conceptualization. However, very few studies have attempted to determine their
explanatory power on nontraditional types of crime - particularly those fostered through
the use of computers and the Internet. With the burgeoning frequency, prevalence, and
scope of hi gh-tech illegalities, such phenomena merit immediate and substantive
attention by scholars and practitioners, especially those in the domain of criminal justice.
As of yet, however, no research has sought to test the “generality” of GST, SCT,
and SLT by studying a decidedly nontraditional form of deviance that occurs in a
decidedly nontraditional environment. Moreover, very few pieces of criminological
research have been published on a purely Intemet-based illegality', and no study has
attempted to analyze the digital music phenomenon from a social science perspective.
Concomitant with the rapid global advance into an information-based society and
economy, intellectual property will be afforded an increasingly immeasurable value. The
critical role that it plays in the stability, vitality, and growth of private sector companies,
public sector organizations, and even individual lives demands that it is secured and
precluded from misappropriation, exploitation, and manipulation for criminal purposes.
The current research attempts to examine the applicability of these general
theories on one speciﬁc form of Internet crime: online intellectual property theft, as
measured by participation in illegally uploading and downloading digital music (MP3)2
ﬁles. Such empirical assessment should determine the extensibility of criminological
theory to crimes in cyberspace, and should contribute to the discipline’s knowledgebase

through its testing of the universality of three major theories. Ideally, this study will

 

' The author knows only of his work of online software piracy (see Hinduja, 2001, 2003a).
2 MP3 is an abbreviation for MPEG Audio Layer 3, an audio compression technology that shrinks the ﬁle
size of CD-quality audio while maintaining the level of ﬁdelity. It is explicated in detail later in this work.

discern the most salient predictors of such wrongdoing and serve as a foundational
inquiry into novel forms of deviance engendered by computers and the Internet. Fruitful
policy initiatives may consequently be developed to restrict the propagation of this
criminal activity, and to simultaneously work to reduce copyright infringement of this
and other forms of intellectual property in the ﬁiture.

Practical Backdrop of Research

MP3 technology has been heralded as the music lover’s dream. Its popularity has
grown from only being known among small circles of Internet technophiles to competing
with “sex” as the most-queried keyword on search engines across the World Wide Web
in its pinnacle year of 2000 (Knight Systems, 2000). Since that year, the phenomenon
has lost a little momentum, but is still wildly popular. According to the most recent
annual statistics available from Google, the most popular search engine as of today,
“MP3” was still the most popular technology-related search term queried among over 55
billion searches during 2002 (Google, 2002).

One might wonder, what exactly is the basis for such tremendous popularity? To
begin, the technology3 has granted free, unrestricted access to songs of extremely high
ﬁdelity by practically every musical artist, past and present. Also, it has allowed
individuals to amass enormous collections of digital music ﬁles, provide them to others,

make custom audio CDS of favorite tracks, and transfer them onto portable players to

 

3 MP3 audio ﬁles are by far the most popular form of digital music compression technology. Other forms
include Ogg Vorbis (.ogg), Advanced Audio Coding (.aac)., and Windows Media Audio (.wma). These are
not studied in the current work due to the comparatively small number of individuals who are familiar with
them Potential competing music ﬁle formats have been hard-pressed to capture the attention of millions of
MPB participants who became booked and hopeless devoted to encoding, sending, receiving, and playing
music in this format. With these facts in mind, “MP3” and “digital music” are used synonymously in this
paper for the purposes of simplicity. The “technology” is inclusive of the compression algorithm and all of
the software that facilitates the exchange of MP3 ﬁles.

satisfy music needs on-the-go. It has spawned massive virtual communities in Internet
chat rooms, message boards, newsgroups, and other cybervenues - in existence solely for
the purpose of distributing MP3s. Moreover, it has facilitated the growth of hundreds of
“dot.com” businesses, allowing millions of dollars to be earned by capitalizing on the
proﬁtability and marketability of this method of distributing compressed audio. Finally,
the pervasive and ubiquitous nature of MP3s has transformed music for the recording
industry,’the artists, and especially the general consumer.

These are just some of the fruits born from this revolutionary technological
advance. Many would argue it has done a world of good for Internet users and music
fans. Others, however, disagree strongly and point towards the inherent illegality of
distributing and reproducing copyrighted works without authorization. Despite the
outcry from the federal government, the major record labels, and other entities dealing in
intellectual property, MPBS continue to be uploaded and dOwnloaded with impunity.
According to the Recording Industry Association of America (RIAA), music piracy in the
form of bootlegged and counterfeit recordings on physical media costs the record
industry $5 billion a year, with $1 million lost each day in the United States (RIAA,
2000a). Illegal exchanges of digital ﬁles occurring over the Intemet (such as MP35) are
exponentially more difﬁcult to track, and one can only estimate through anecdotal
evidence the amount of revenue being denied to musicians and record companies through
this practice. Admittedly, it is difﬁcult to conceptualize and measure potential “sales”
lost by an artist if the MP3 versions of the songs were not available online. Nevertheless,
it is unarguable that at least some individuals are losing out on due compensation because

their music is being circulated over the Internet without their consent.

 

Accurate statistics associated with the phenomenon are relatively rare, and are
often perceived as biased depending on their protagonist or antagonist source. Some
numbers, though, from the International Federation of the Phonographic Industry’s 2002
Music Piracy Report paint a compelling picture of the scope and prevalence of
participation in online intellectual property theft. Piracy is argued to be the greatest
threat to the respective industries of music, movies, and software (IFPI, 2002), and with
more instances of plagiarism and misappropriation of textual content, the press and
mainstream media may ﬁnd itself increasingly harmed by piracy of their own produced
material in the near future. Copyright industries in 2001 accounted for 5% of the GDP,
which equates to $535 billion dollars (IFPI, 2002). The availability of so much
copyrighted material worth such a signiﬁcant amount provides the opportunity and the
rewards for potential and actual perpetrators to participate in and beneﬁt from piracy.

IFPI (2002) also estimates that at any given time in May 2002, there were
approximately 3 million participants in music piracy providing around 500 million music
ﬁles for unauthorized downloading by anyone with the inclination to do so. Much of this
occurs through point-and-click software that easily facilitates MP3 transfers among users,
such as KaZaa, the most popular ﬁle-exchange program in current existence following
the demise of Napster It is estimated that 10% of Internet users have downloaded KaZaa
(Black, 2003); the application’s web site boasts that it has been downloaded over 197
million times, with around 3 million program downloads each week (Sharman Networks,
2003). IF PI (2002) also estimates that 200,000 FTP and web sites exist online, which
host or link to at least 100 million music ﬁles available without proper permission from,

or payment to, the creators and producers of the music works.

Unquestionably, since computer crimes are not easily identiﬁable (let alone easily
measurable), many are currently unaccommodating to empirical research. Crimes such
as hacking, child pornography, and Internet fraud can possibly be examined through the
use of case studies and through content analysis of message texts and documents
transmitted between participants. A quantitative piece would likely have a very small
number of cases, simply because of the difﬁculty associated with identifying, soliciting,
and tracking participants for inclusion in a sample. Conversely, copyright inﬁ'ingement -
the unauthorized duplication or distribution of software, music, movies, and other forms
of intellectual property - is a crime committed by millions of individuals on a regular
basis. As such, this particular act provides a large population from which to obtain study
elements (unlike other forms of high-tech deviance), and therefore appears to be the only
computer crime capable of being rigorously analyzed at the present time.

Theoretical Backdrop of Research

When considering MPB participation and intellectual property theft in general,
some criminological and sociological questions inevitably come to mind. What
motivates or impels individuals to partake in this illicit activity? Do certain dispositions
and inclinations differentiate participants from nonparticipants? What micro- and macro-
level factors play predictive roles? In the 20th century, three general criminological
theories carved for themselves a substantial niche in the knowledgebase of explanations
for crime and deviance. Individually and collectively, they appear useful in answering
speciﬁc inquiries which follow neatly from the aforementioned questions.

First, general strain theory — asserted by Robert Agnew (1992) may also Shed

light on the impact of maladaptive affective responses on copyright inﬁingement. The

primary question to be asked and answered with this theory is: DO feelings of strain or
dissonance, when engendered among those consumers who are not able to afford or
Obtain certain intellectual property but who still desire to appropriate it, induce
participation in criminality? Second, Gottfredson and Hirschi’s (1990) self-control
theory — also known as the “general theory of crime” — appears to also be a valid
framework in which to view intellectual property theft. Speciﬁcally, if an individual has
low self-control, does that make him or her more likely to participate in the behavior
when presented with the right opportunity? Third, social learning theory - as proposed by
Ronald Akers (1985) - sets the stage for a deep analysis into social elements that facilitate
music piracy participation. It may be able to answer the following: How and from who
are the techniques and justiﬁcations employed to participate in the deviance learned? Is
the behavior modeled after the actions of others? How is the behavior reinforced and
perpetuated, or punished and ceased?

In addition, these theories might also have interaction effects with each other.
Low self-control might increase the likelihood of a person succumbing to the pressures of
antinormative peers and pressures, or make a person more inclined to respond in an
antisocial manner to perceived or real stress and cognitive conﬂict. Moreover, continual
exposure to strain might augment the tendency of a person to display characteristics of
low self-control, and to be swayed by maladaptive social learning. Testing for
statistically signiﬁcant interactions is beyond the scope of the current work, but merits

explanation in future research endeavors.

Necessity for Research and Response

A host of reasons underscore the importance of studying this phenomenon and
developing informed ways of addressing its growth. Theft of digital property over an
Internet connection is easier and quicker than doing so from a retail establishment, while
the chance of detection, apprehension, and prosecution is exponentially smaller.
Nevertheless, both are activities prohibited by the law and induce similar harm. Still,
music piracy is often condoned in some circles as a victirnless activity that does not befall
tangible or noteworthy harm to any person or organization. Such an argument can be
refuted by a host of facts. For instance, the accumulated economic loss incurred is
signiﬁcant to the artist, recording company, and industry, and is said to approximate $4.3
billion worldwide (IFPI, 2002). Through piracy, these associated parties are preempted
from receiving compensation for the creation, production, marketing, and distribution of
their intellectual product. The desire to innovate and develop creative works may be
stiﬂed if the rewards are less than anticipated, and if persons are able to appropriate the
product without paying for the good and service (Harris, 1969; Smith & Parr, 1989). The
media- and entertainment-based economy is presumably deprived of investments and
proﬁt from their product, and must devote resources to research and development to
create copy-protection solutions, surveillance and tracking mechanisms, and punitive
policies to discourage or thwart attempted theft. It may also reduce jobs in the industry
as cutbacks are made to counterbalance the incurred ﬁnancial losses.

Additionally, if disregard for intellectual property such as music continues
unfettered, it is not a far stretch to conclude that eventually nothing posted on the Internet

will be safe from misappropriation. A Pandora’s box will be ripped open, and articles,

thoughts, ideas, graphics, art, sound ﬁles, animations, movies, software", and more will
no longer be the property of their rightful owner(s), but will be free-rein for anyone to
copy and use without regard. Indeed, this is already taking place, albeit to a lesser degree
than with the breakneck proliferation and propagation of MP3s. What seems to be
forgotten is this: intellectual property is still property, and is owned by its creator or the
company that has purchased its rights. If one does not have the legal authority to
reproduce and distribute a copyrighted work, but does so anyway, a crime has been
committed. Criminal activity, then, is subject to prosecution, ﬁnes, and incarceration;
Speciﬁc legislation detailing punitive sanctions is discussed below. Copyright
infringement through MP3s must be dealt with and not overlooked or minimized simply
because of its nontraditional nature (as compared to crimes which attract more attention
and criminal justice resources) or because of the unique “Virtual” environment in which
the activity occurs.

Moreover, the college—age population who disproportionately participates in the
music piracy phenomenon merits attention. Hinduja (2003b) argues that a “slippery-
slope” effect might be manifested, as digital theft may precipitate more signiﬁcant forms
of computer- and Intemet-related deviance. That is, music piracy may possibly serve as a
gateway to more severe forms of high-tech crime. Such a correlation has yet to be
determined, of course, but its relational viability appears quite real. Finally, the integrity
of the educational establishment at which such behavior takes place is undermined, and
the ethical and normative standards of individuals who participate are compromised and

seemingly weakened, rather than fortiﬁed in this scenario.

 

4 For examples, please see: Berst, 1997; CyberAtlas, 2001; Dyrness, 2002; Evangelista, 2003; Gentile,

Several research questions are posed:

0 Do the purported “general” theories of crime have the predictive capacity and
ﬂexibility to explain Intemet-based criminality?

o Are certain elements of each theory more salient than others?

0 Do these theories need to be modiﬁed or augmented to address issues related to
the “virtual” environment of cyberspace over which such illegalities transpire?

o Are these theories accurately termed “general,” or is a partial or complete
theoretical reconceptualization required to most thoroughly explain both
traditional and nontraditional deviance (such as crime executed on the Internet)?
To reiterate, the current work attempts to systematically explain one type of

Internet crime - copyright infringement in the form of digital music piracy - by
conceptualizing three general criminological theories as predictors, operationalizing
concepts inherent in these perspectives, and measuring their proposed relationship with
involvement in the wrongdoing. Fleshing out the cognitive, behavioral, psychological,
and sociological factors that play a role in effectuating copyright infringement will prove
valuable as a cogent theoretical foray into the mind of a computer deviant. This
undertaking will hopefully result in the identiﬁcation of some causes of Internet crime,
thereby producing new knowledge and insight that will enhance our current
understanding of the phenomenon. AS such, the application of theory, the empirical
examination of contributive factors, and the subsequent construction andimplementation
of policy solutions to prevent and suppress the illegal activity will ideally result from this

research effort.

 

2003; Haney, 2000; Harris, 2003; Jacobs & Allbritton, 2001; Johnston, 2000.

10

Value and Contribution of the Research

The current study focuses on a controversial and questionable activity involving
computers and the Internet. Participation in the MP3 phenomenon may not seem as
criminogenic as pirating software, hacking into networked systems, or writing viruses,
especially when coupled with its panoptic reach, exponential growth, and unrivaled
popularity. However, except under limited circumstances it equates to the contravention
of extant copyright law, and therefore is illegal. This illegality has resulted in the
crippling of some MP3-based businesses (e. g., Napster, Scour) and the incurring of
severe ﬁnancial penalties on others (e. g., MP3.com), as well as social, civil, and criminal
penalties for otherss. Intellectual property theft via MP3s is a computer crime, and its
inherently unique qualities render it interesting as the subject of a research project, and
arguably as an appropriate proxy for the phenomenon of Internet crime in general.

Formal research concerning computer crime has increased over the ﬁve years, but
still leaves much to be desired both in its scope and theoretical application, as well as
with the development of effective policy initiatives. The subject of intellectual property
theft in general has been discussed and debated by legal scholars, as the law is constantly
shaped and changed by new judicial decisions in this area (see e.g., Lessig, 1997, 1999a,
1999b; Luckenbill & Miller, 1998). A few philosophers (e.g.,Tyler, 1996) have also
written about the subject to ﬂesh out ideas which may possibly stem the tide of copyright
infringement. Software piracy has been Speciﬁcally studied in the business ethics and

management information systems ﬁelds, but these inquiries have been primarily

 

5 See e.g.., (A & M Records Inc. et al. v. Napster Inc., 2001; CNN .com, 2000b, 2000c; Davis, 2003; Healy,
2003; Jones, 2000; Lipton, 1998; Mendels, 1999; Patrizio, 1999; Philipkoski, 1999a, 1999b; RIAA, 2000b;
Spring, 2000).

ll

 

descriptive in nature and only a few have developed or tested a theoretical model to shed
light on the impetus of the behavior (see e.g., Gopal, Sanders, Bhattacharj ee, Agrawal, &
Wagner, forthcoming; Wagner, 1998). Indeed, most of the current policies in place have
stemmed from anecdotal accounts rather than empirically grounded studies (Rogers,
2001)

The salience and diffused nature of computer crime, however, demands more
formal and intensive analyses to best devise competent proactive and reactive strategies
to address the matter. In the past decade, some scholars (Sherizen, 1997; Skinner &
Fream, 1997) have sought to determine the applicability of sociological and
criminological theories on software piracy, but no previous studies in the social sciences
have examined the MP3 phenomenon. With the continual advances in information
technology and the increasing presence and distribution of intellectual property online, a
rigorous theoretical approach to interpreting and analyzing copyright violators and
violations holds much value. Legislators, academicians, and practitioners can also
beneﬁt from this research, both by garnering a deeper knowledge of the nuances of both
crimes and criminals and by obtaining direction in how they might attend to the issues at
hand (Denning, 1998). Ideally, the current work will result in a greater comprehension of
the contributive factors of music piracy, and in the generation of policy that will curtail
the problem. This inquiry is warranted so that the novel occasions for deviance resulting
from technological progress do not overshadow the promises and proﬁts of the continued
progress into an information-based society.

Most importantly, though, this study attempts to assess the generality of three

criminological theories which have been professed to explain all types of crime. It is

12

important to determine if GST, SCT, and SLT are extendable to Internet crime. The
deﬁnitional validity of these theories hinge on identifying their relevance in predicting
nontraditional forms of wrongdoing rarely explored in empirical criminological research.
Such a rigorous theoretical examination is the linchpin of the current project.
Organization of Succeeding Text

In this paper, an examination of the etiology of MPBS and copyright violation is
conducted, against a backdrop of correct interpretation of the activity as a civil and often
criminal offense in the vast majority of cases6. An introduction to Internet audio will ﬁrst
be given to provide fundamental knowledge requisite for a thorough understanding of the
issue at hand. Then, a technical breakdown of the MP3 phenomenon will ensue,
including an analysis of the speciﬁcations of the ﬁle format, means of production, and
methods of delivery to Internet users. An examination of literature and empirical
research on MP3 technology follows, primarily derived from Intemet-based news sites
and online marketing and research ﬁrms. Next, a review of extant literature on GST,
SCT, and SLT will be provided to demonstrate the pertinence of each framework to
traditional forms of crime. Their analogous relevance to the nontraditional crime of
music piracy on the Internet will then be posited to depict how the applicability of the
theories can be extended.

A quantitative analysis through the use of a survey instrument is subsequently

conducted on data collected from a sample of university students to more accurately

13

assess the applicability of each criminological theory on participation in the MP3
phenomenon, and to provide statistical ﬁndings which can be used to shape policy and
other productive implementations to combat online intellectual property theft on the
Internet. These suggested measures will then be discussed in detail, with the intention of
framing ideas into feasible practices that can accommodate the beneﬁts of the new digital

economy, the music industry, and the perpetually growing wired world.

 

‘5 While much social and legal controversy surrounds this issue, copyright infringement is currently deﬁned
as an unlawful offense, and the current research is conducted with that perspective in mind. Civil remedies
are available for copyright infringement irrespective of the intention or knowledge of the perpetrator or the
amount or degree of harm done to the victim. Criminal remedies are available for intentional acts that
result in private ﬁnancial gain or commercial advantage. Financial gain does not only refer to proﬁting by
the perpetrator but also refers to possible ﬁnancial loss to the victim (RIAA, 2000c).

l4

CHAPTER 2: REVIEW OF THE LITERATURE
Historical Background of Digital Audio Files

The frrst multimedia personal computers were introduced to the world in 1985
with the Commodore Amiga. These systems not only provided a graphical user interface
and a multitasking operating system to individuals, but also integrated relatively
adVanced sound and graphic capabilities (Patterson, 1998). AS Auriga, IBM, IBM-PC
Clones (e. g., Compaq), Macintosh, and other personal systems became increasingly
advanced in their processing power and ﬁmctionality, enhanced multimedia capacities
were standardized as fundamental features of computers. CD-ROM drives were
marketed with new personal computers beginning in the mid 19908, and complex
software applications and games were developed and sold on CDS to utilize the new
technology.

Concurrently, audio CDs competed with and ultimately surpassed the popularity
of cassette tapes for market share in retail recordings. Music on cassettes were recorded
in analog format, but music on CDS was digital in format and could be digitally extracted
into waveform format using a computer and software available to the average consumer.
Nonetheless, the sheer size of the resultant native audio ﬁle precluded ease of distribution
and exchange. Waveform ﬁles of CD quality consumed approximately 10 megabytes per
each minute Of audio and circulation of these high-quality music ﬁles just did not happen,
as most individuals had extremely Slow Internet connections. With the explosive growth
of the Internet, different audio ﬁle formats allowing for music to be more easily posted

and distributed became popular.

15

The initial formats included .AU (Sun Microsystems’s proprietary audio format),
MIDI (Musical Instrument Digital Interface - a format which stores descriptive
information on how musical notes should be played in a particular arrangement, rather
than the music itself), and WAV (the Microsoft default audio format, and the most
popular because of the pervasiveness of Microsoft operating systems on personal
computers). Due primarily to issues related to ﬁle size, those clips that were posted
online were typically created with low-quality bitrates and smaller sample rates, better
suited for speech and perhaps quotes from television or movies.

A monumental event occurred in 1995 when RealNetworks, Inc. released
RealAudio 1.0, a streaming audio format to which web surfers could listen within a few
seconds of clicking on a hyperlink, rather than waiting until an entire piece of music
downloaded onto the user’s computer. While this is representative of the traditional
broadcast model of information dissemination, the quality of this streaming multimedia
left much to be desired and was arguably unpractical for anything other than news bites
and speech clips. Despite the advances in multimedia, listenable popular music of
reasonable quality was not available at this time on the Internet, and formats that
“streamed” music for playback through the end user’s computer speakers could not be
saved to enjoy at a later time. MP3 technology, however, was able to meet these desires
of music fans, and ultimately exceeded everyone’s expectations through the way in which
it facilitated delivery of a valued commodity.

MP3 Technical Background
MP3 (an abbreviation for MPEG-1 Layer Audio 3) is an audio compression

format that enables audio ﬁles to be compacted into relatively small ﬁle sizes, while

16

maintaining near perfect ﬁdelity when played back. It is a direct descendant of MPEG-1
(low-bandwidth Video compression, typically used over the Internet) and MPEG-2 (high-
bandwidth audio and video compression that is the standard for DVD technology)
(Midgley, 2000). A general idea of the heuristics of the MP3 compression process is
useful to note. Compression occurs through the use of perpetual coding techniques,
where auditory information from large digital multimedia ﬁles that exceeds the
perceptual range of human hearing is removed, resulting in smaller ﬁle sizes (Crawford,
2000). Its functionality relies on mathematical algorithms developed using knowledge on
how the human ear hears sounds. These algorithms are then able to analytically
determine which components of the audio data can be heard by the human ear and those
that are inaudible or masked. By discarding those data which do little to contribute to
perceivable sound quality, the size of the ﬁle is greatly reduced. This process is very
similar to how the Internet graphics format J PEG works, essentially eliminating visual
data in images that human eyes cannot easily detect (Heid, 1997).

As a general metric, it is said that one megabyte (MB) is typically equivalent to
one minute of music in compressed MP3 format. MP3 compression can produce audio
ﬁles of several different quality levels measured by the amount of data per second -
required to reproduce that second of sound. A larger amount of data results in higher
audio quality, but at the expense of a consequently larger ﬁle size. To convert an analog
sound recording or “waveform” into a digital format, a process called “sampling” must
take place. The more often one samples a waveform per second, the more accurately
sound can be reproduced. Audio data on CDS and, as a consequence, practically all MP3

ﬁles are sampled at a rate of 44.1khz — where 44,100 16-bit-precision samples are taken

17

each second to accurately reproduce the sound, and separate samples are taken for the
right and left speakers in a stereo system.

Exactly 1,411,200 bits (or 176,000 bytes) of data are needed for each second of
music on a CD. As a consequence, a three-minute song would occupy 253,440,000 bits
(31,680,000 bytes). Compressing it in MP3 format, however, results in a multiplicative
decrease of its ﬁle size. As such, 1,411,200 bits per second can be reduced to 128,000
bits per second (or 160,000 or 192,000 -— common bitrates for MP3 ﬁles) by using the
compression algorithm. This translates into a ﬁle size 11.025, 8.82, or 7.35 times
smaller, respectively, than the original. Furthermore, CD audio requires a bandwidth of
1.5 megabits per second in order to play perfectly. The compression technology also
signiﬁcantly reduces the bandwidth requirement for CD quality playback (Karagiannis,
1999). Through this process, audio ﬁdelity is largely preserved while at the same time
simplifying and accelerating the transfer and storage of muSic tracks.

Fidelity refers to the degree to which an electronic system reproduces sound
without distortion. For many pieces of music, the MP3 sound quality at 128kbps,
reducing the music ﬁle Size to approximately 9.1% of the original, comes negligibly close
to music from a CD (Calpo, 2000). In recent years, the use of 192kbps rates have
increased in popularity, but this more accurate hi gh-end reproduction comes with an
approximately 33% increase in ﬁle size (approximately 13.6% of the original ﬁle) when
compared to the same ﬁle encoded at 128kbps (Calpo, 2000). With the increased
availability of gratuitously large hard drives, ﬁle size no longer seems relevant as

individuals tend to seek the highest quality digital audio they can obtain.

18

In 1987, Fraunhofer Gesellschaft — a prominent technology institute in Germany -
began work on the audio ﬁle compression algorithm. After perfecting and patenting it in
1989, it was submitted to the International Standards Organization (ISO) to be integrated
as an audio subset into the speciﬁcation for the video compression technology termed
MPEG-l7 (Nijmeh, 2001). MP3 began to gain prominence as an audio-only compression
scheme among the Intemet diehards at the beginning of 1997 when software to play
MPBS were written and released as freeware online. Concurrently, the hardware
infrastructure had developed to the point where it could support the software - Intel
Corporation ﬁnally released a central processing unit (CPU) fast enough to decode MP3
ﬁles in real-time, enabling playback as soon as the user clicked on the ﬁle, rather than
requiring wait time while the song made an uncompressed copy of itself on the hard drive
and then began to play (Weekly, 2000). It was not until 1999, however, that MP3 began
to take off among the general and less technically inclined online users.

To be sure, MPB not only “took off,” but ushered in a revolution of sorts among a
consumer population desiring their music. MPB players, such as Nullsoft’s WinAmp and
MusicMatch Jukebox, encoders such as Xing’s AudioCatalyst and Telos Systems’
Audioactive Production Studio, and organizers such as ShufﬂePlay and Helium, were
rapidly developed and deployed to worldwide users enamored by the promise of this new
technology. The hardware and software infrastructure in place was and continues to be

vastly sufﬁcient to perpetuate the exponential growth of the phenomenon despite the

 

7 MPEG is an acronym for Moving Pictures Expert Group, a subcommittee of the International Standards
Organization (ISO). That organization sets worldwide standards for business, technology, and society;
over 13,700 standards have been published since 1947 (International Standards Organization, 2003).

19

 

 

legal troubles that MP3 has caused certain businesses and individuals (which are
discussed below).

With regard to hardware, computer hard drives of multi-gigabyte disk space had
been appearing in the retail market at progressively reasonable prices, which afforded
many online music aﬁcionados the opportunity to amass large holdings of MP3 ﬁles on
their systems. Gains in hard drive size also beneﬁted the distributors of the music, as
web and ﬁle servers that individuals accessed on the Internet could now accommodate
larger collections without devastating the pocketbooks of the online businesses.
Throughout the 19903, hard drive capacity increased by 60% each year, and the average
size of hard drives sold in 2000 was at least 10 gigabytes (de Fontenay, 1999; Quantum
Corporation, 2000). If nothing but MP3 ﬁles of an average size of 4 megabytes each
were stored on a 10 gigabyte hard drive, an individual would have an easily accessible
MP3 jukebox containing around 2,500 songs. Even just a single gigabyte of space can
hold approximately 250 high-quality MP3 ﬁles.

Moreover, the falling prices of CD burners - drives capable of recording data onto
blank CDS — permitted individuals to easily dump an average of 12 hours of MP3 music
onto very inexpensive recordable discs (Crawford, 2000). Portable devices which ﬁt
hours of compressed music in the resident memory modules also gained a great deal of
popularity as prices decreasedg. Thus, because of the new digital music paradigm

fostered by MP3 technology, the distribution of audio over the Internet has become less

 

8 Many new personal and car CD players have the functionality to decode and play ﬁles in MP3 format
(Consumer Electronics News, 2003; Weekly, 1998). Interestingly, research indicates that the market for
audio electronics that play compressed digital music ﬁles will grow to almost $44 billion dollars by 2007
(Consumer Electronics News, 2003).

20

of a broadcasted service. Rather, it has taken on the qualities of a copious and valuable
property acquired, stored, and circulated as desired.

While the majority of digital music transfer occurs to satiate the auditory palates
of music lovers, others download and produce MP3s primarily so that they can have a
digital backup of their music collection in case a CD is lost, stolen, or scratched
(Swiatecki, 2000a). Furthermore, MP3s enable individuals to compile collections of their
favorite songs and create a custom playlist on their computer, to play tracks in a
sequential manner, and not have to worry about switching CDS in and out of their player.
While record companies have felt their grip over the control, production, and distribution
of music slowly slipping away, consumers have rejoiced because of the ability to hear the
work of musicians before they spend their hard-earned money on an album with perhaps
only one or two decent songs. In addition, MP3 participants praise the opportunity to be
exposed to a wider variety of music genres and to hear the. creative productions of
thousands of unsigned and yet talented bands and artists (Swiatecki, 2000a). Others
argue that the music industry has held an unfair monopoly over the music market and has
maintained an inﬂated price for CDS, compensated artists comparatively little for their
efforts, and have reaped sizable proﬁts from these exploitative practices (Swiatecki,
2000a). Conversely, many contend that while it is nice to sample full-length high-quality
music for ﬁee, they will not be deterred or swayed from purchasing album-length CDS
that they can easily take with them to work, to use in their vehicle, and which contain
liner notes, lyrics, and cover art (Swiatecki, 20003).

Other digital music compression formats have been recently formulated by some

of the largest IT companies in the world, but often require certain proprietary hardware or

21

software for playback, and generally include digital protections and limitations to
preempt their unauthorized use or dissemination. End users, though, have become
accustomed to, and have in fact demanded, freedom of choice in the ability to control
their music experience. MP3 technology accordingly became the standard because of its
widespread acceptability, comparatively fast download speed, minimal storage
requirements, near-CD quality of sound, ease of use, and ﬂexibility (Kibbee, 1999). It
did not achieve pervasive popularity, however, among the general populace until the
development of a user-ﬁiendly freeware software application called “Napster” hit the
Internet in 1999.
Napster — A Revolutionary Application

Napster was a small peer-to-peer (P2P) ﬁle exchange application that transforms
users’ computers into de facto ﬁle servers, enabling them to upload to and download
from millions of other MP3 enthusiasts around the world: It was the software — “the
killer app” - that revolutionized the worlds of both producers and consumers of music
(Petersen, 2000). As mentioned, no longer were recordings on physical media such as
CDS and tapes the only way in which a desired commodity could be obtained. No longer
required was the payment of a premium price to possess high-quality music. Now, it .
could be simply and freely be downloaded from the Internet - thereby providing the same
aural beneﬁt to the end user, but in the package of a small digital ﬁle that could be stored
and transferred to others at virtually no cost (apart from the expense associated with
owning a computer and an online connection). How, though, might individuals be
introduced to, and become enamored with, the beneﬁts of MP3 ﬁles? What might serve

as the catalyst to make a sparsely known audio compression technology ubiquitously

22

embraced by millions and millions of people? Napster, the freely available software
program that brought the technology to millions around the world, was the actual
instrument of change.

This is how it worked: interested individuals visited the Napster web site,
downloaded the software, and installed it. Upon signing on to the Napster network, the
application scanned the end user’s hard drive for MP3 ﬁles, and catalogued the name of
the artist, the song title, and other related variables that designate ﬁdelity of the track and
speed of the user’s Internet connection. Then, while connected to the network along with
thousands of others, the user’s catalogue of MP3 ﬁles was concatenated into a giant
database with the lists of every other individual who was signed on to that particular
system. Search queries could then be run for particular artists or tracks, and results were
displayed showing the other persons who had available the sought-after MP3 ﬁle. Then,
with the click of a button, a download could be initiated, transferring the ﬁle from the
request grantor to the request initiator.

Prior to the invention of this program, MP3s were distributed primarily through
Internet Relay Chat (IRC) channels and through downloads from web pages (which often
offered unreliable links to MP3 ﬁles housed on other servers, and which were usually
shut down by the RIAA after a brief period due to their case of detection) (Harari, 1999).
File transferring using specialized but arcane programs were often employed by the
Intemet-savvy, but the general public was neither cognizant nor inclined to spend the
time and effort to use such a method to obtain music ﬁles. Napster’s easy-to-use

interface and pre-establishment of a network between all of its users greatly conduced to

23

simple, point-and-click accumulation of desired digital music tracks onto the personal
and ofﬁce computers of individuals.

0 Students at universities and many other music fans with fast Internet connections
began to ﬁll up hard drive space and CDS with MP3s. This was presumably for the
purposes of obtaining free music they would otherwise have to purchase, and also simply
for the sake of possessing them and increasing the size of their music collection.
Unfortunately, this clogged network pipelines and retarded the ability of other individuals
from using the Shared bandwidth resources for legitimate purposes, such as downloading
an open-source operating system for their research work, sending documents and work-
related ﬁles to other users for the purposes of collaboration, or even merely browsing the
World Wide Web (Krochmal, 1998; Stenger, 2000; Swiatecki, 2000b).

As a consequence, academic institutions such as Indiana Universityg, San Diego
State University, and the University of Chicago were forced to disallow access to the
TCP ports used by Napster, which resulted in an outcry of censorship by university
students around the nation (Kover, 2000). Other schools attempted to arrive at a common
ground, either by regulating the amount of bandwidth a user could commandeer, or by
instituting data packet analysis and ﬁltering programs10 to keep tabs on which users are
consuming the largest amounts of data (Stenger, 2000; Swiatecki, 2000b).

Napster allowed for giant repositories of music to be easily accumulated by

individuals, providing accessibility to hits of yesterday, today, and tomorrow with

 

9 Incidentally, Indiana University and a host of other colleges overturned their decision to ban the program
within a few months of their initial decision following the negative publicity that ensued.

1° Proxy servers and other workarounds which can be implemented by the end user to skirt the restrictions
in place that deny access to Napster and other ﬁle-exchange programs. Therefore, these network
management tools seem essential despite their somewhat intrusive nature.

24

incredible ease and functionality, and requiring no extra physical storage space other than
that taken up by one’s computer. It is true that some MP3s available on Napster were
legal ﬁles, supported by the artist or band. The vast majority, however, were illegal ﬁles
of copyrighted works by a sizable number of commercial artists ranging from the latest
pop track by Madonna, to the country sounds of the Dixie Chicks, to the hardcore rap
songs of Snoop Dogg, to the old-school classic rock of 22 Top. The owners of the
copyright (usually held jointly by the artist and the record company) were not
compensated through unauthorized downloads of their music on the Internet, thereby
engendering the problem.

While programs with similar ﬁle exchange capabilities have been developed since
1999 (e.g., Gnutella, Wrapster, Napigator, iMesh, Scour, Kazaa, Morpheus, Bearshare,
eDonkey) and continue to be used by millions of individuals on a daily basis (Black,
2003; Sharman Networks, 2003), Napster was the harbinger of the boom in digital music
downloading (Spring, 2000). The primary beneﬁt of MP3s through these and other
mediums seems to be the receipt of a valued product and service that otherwise would
have to be purchased as a physical recording. Such a concept outraged the producers but
delighted the consumers, and other individuals and entities implicated in the phenomenon
began to take sides.
Players in the MP3 Phenomenon

In the controversy, there are eight primary players whose interaction between and
among each other fuel the dynamic nature of the issue. On one side, there is the general
public - who by and large are supporters, endorsers, and participants in uploading,

downloading, and otherwise distributing MP3s. Independent artists - those not signed to

25

m):
a tool
Ines;
lily)
atom
on 11'
bus

Item

cons
11‘
his

Fun

and 1'
IS pa;

A703

a major music label - often claim allegiance to this side as well because they see MP3 as
a tool of promotion, and a way to provide their music to anyone and everyone
irrespective of whether they have been “discovered” by a recording company (Bowman,
2000). Internet startups and "dot.com ” entrepreneurs - those who have built businesses
around the potential technological, economic, and social beneﬁts offered by MP38 - stand
on this side as well. Some have succeeded (or did succeed before the “dot.com bubble”
burst), which testiﬁes to the fact that the goods and services they provide are valued and
tremendously popular among music aﬁcionados (CNN.com, 2000a).

HackersH are another group who realize a signiﬁcant value and beneﬁt associated
with MP3s This is likely because their computers are often an extension of themselves
for a lengthy period of time each day, and digital music at their ﬁngertips furnishes a
constant amount of listening entertainment while they engage in various computing
activities. This is also a collective who actively champion the voice of the average
Internet user and generally oppose any perceived semblance of capitalistic exploitation.
Furthermore, hackers commonly organize and assist in the circulation and “sharing” of
songs and albums in MP3 format to the masses, and work to “crack” software and digital
protections that might hamper or restrict the end user’s goals to acquire music for free.

Antagonists to the MPB phenomenon are less in number but arguably more potent
and inﬂuential as a whole because of their relative position on the economic ladder. This
is partly due to established relationships with the corporate and government sector.

Another reason is that the social structure endows them with a disproportionate amount

26

of power to deﬁne and enforce their constructed deﬁnitions of legal and acceptable
behavior. That is, what comes to be known as propriety seems shaped not so much by
legal, institutional, or societal factors but is rather an extension of economic self-
aggrandizement by the privileged. The federal government and the courts have largely
defended the interests of the private corporations and businesses and have imposed
restrictions and harsh penalties on entrepreneurial MP3-based enterprises (e.g., MP3.com,
Napster, Inc.) and individuals because of their alleged copyright-infringing practices (A
& M Records Inc. et al. v. Napster Inc., 2001; CNN.com, 2000b, 2000c; Davis, 2003;
Healy, 2003; Jones, 2000; Lipton, 1998; Mendels, 1999; Patrizio, 1999; Philipkoski,
1999a, 1999b; RIAA, 2000g; Spring, 2000).

The most powerful player is the recording industry, a $40 billion dollar behemoth
of authority and clout (RIAA, 2003). Its chief voice is the Recording Industry
Association of America (RIAA), a trade consortium comprised of record companies that
distribute approximately 90% of legitimate sound recordings in the United States and
seek to foster a “business and legal climate that promotes [their] members’ creative and
ﬁnancial vitality” (RIAA, 2000i). Through the introduction and distribution of MP3s
online, the traditional medium (and “cash cow”) of distribution through packaging,
marketing, and selling recordings on compact discs and tapes has been forcibly
weakened. Commercial musical artists signed to the major record labels generally are

hostile towards the unauthorized availability and distribution of their creative works, as

 

" The term “hacker” is chosen for use here for the purposes of Simplicity. In cyberspace social circles,
“hacker” is properly used only for those who explore systems for the purposes of well-intentioned
knowledge discovery, while “cracker” is the term used to signify those who break into systems - and who
break copy-protection mechanisms — for illicit gain.

27

they can no longer control how their product is obtained (Breen, 2000)”. As mentioned
earlier, MPB allows individuals who normally would have to purchase their music from a
retail establishment to acquire and enjoy it for free. Indicative of the consensual attitude
towards the phenomenon and its attendant software facilitators, one popular artist
publicly denounced Napster as “bullS---t hippie capitalism,” and others have expressed
similar sentiments (Bowman, 2000; RIAA, 20000. The ﬁnal player is the attorney, and
there are throngs trumpeting the causes of both conﬂicting positions, each eager to
litigiously prove that the intents and practices of one side are not damaging to the other.
Review of MP3-Related Research
Intellectual property is legally deﬁned as:

1. A category of intangible rights protecting commercially

valuable products of the human intellect. The category

comprises primarily trademark, copyright, and patent

rights, but also includes trade-secret rights, publicity rights,

moral rights, and rights against unfair competition.

2. A commercially valuable product of the human intellect,

in a concrete or abstract form, such as a copyrightable

work, a protectable trademark, a patentable invention, or a

trade secret. (Garner, 1999)

Digital intellectual property can be characterized as a “public good,” in that its

utility is not decreased or removed if given to other individuals. Such a characteristic

encourages the distribution and “sharing” of such property, often with the implicit

 

‘2 Some major artists who are in support of the technology include Limp Bizkit, The Offspring, Chuck D,

28

assumption that others will distribute and “Share” Similar property to collectively meet
the desires of all who participate. In addition, digital intellectual property is an
“information good” as the marginal cost of production is virtually zero. Software and
digital music are two types of digital intellectual property, and Bhattacharj ee, Agrawal, &
Wagner (2003) have stated that they differ in four primary respects:

1) Fidelity. The ﬁdelity of digital music is not as high as the quality of the same
music ﬁ'om an original CD,because digital music is compressed. Fidelity is not an issue
with software.

2) Size. Songs in digital format generally have much smaller ﬁle sizes than
software applications, enabling their transfer at quicker speeds

3) Price. Songs in digital format are generally much less expensive than software
applications

4) Volume. The variety and availability of digital music is vastly greater than that
of software applications

5) Support. No product support or service from the author or manufacturer is
needed for digital music, unlike software applications.

These differences endow digital music with unique qualities which augment its
attractiveness as a valued commodity to be acquired, and which point to reasons why the
music industry has largely refused to embrace the changes this new format has introduced
to their business model.

To determine the extent to which music ﬁle exchanges over the Internet have

proliferated and how they may affect the music industry’s revenue stream, a handful of

 

Deﬁones, and AF 1.

29

empirical Studies have been conducted. However, these have been primarily sponsored
by music business stakeholders in an attempt to validate or refute claims that the
availability of digital audio ﬁles adversely affects CD sales, the recording industry, and
the artists themselves. Any comprehensive examination of the MP3 phenomenon
necessitates a review of these works. Most of these examinations occurred in 1999 and
2000, the years in which Napster served as the catalyst that propelled MP3 technology
into the limelight. It must be reiterated that the subject matter of these studies focused on
determining the acceptability and extent of music piracy among certain populations,
assessing the inﬂuence of piracy on CD purchasing behavior, identifying whether
individuals would be willing to pay for digital downloads, and generally examining ideas
and opinions related to the technology and its consequences for all parties involved.

To begin, in an informal research endeavor, Archambault (1999) posted an online
survey to his webpage and collected data from ﬁfty individuals from the USENET
newsgroups regarding the inﬂuence of MP3s on music sales. It was found that while
16% of respondents stated that they downloaded an MP3 in place of purchasing an
album, 66% stated that they bought an album because of a MP3. Methodological
problems were rife concerning the target population and the generalizability to the rest of
the MP3 community, but the study did identify some interesting results. In his
unpublished doctoral dissertation on entrepreneurial market research, Stenneken (1999)
surveyed 128 individuals from the USENET newsgroups and discovered that 47.7% of
respondents listened to music more than twenty hours a week. Furthermore, he found
that while 93.2% had not paid for a download of a song, many were open to the idea,

with 34% willing to pay between $5-$ 10 for a full-length album, and 26.6% willing to

30

pay less than $5. MP3 participation increased the music purchasing habits for 34.7% of
those surveyed, and did not affect 53.4%. Notably, the music purchasing behavior of
individuals who partook in MP3 distribution had either increased (34.7%) or had not
changed (53.4%).

Data analysis from a sample of 3,300 international Internet users conducted by the
online research ﬁrm Angus Reid Worldwide (2000b) revealed that 36% of the adults and
41% of the teenagers and young adults had downloaded MP3s Of those users between
the ages of 12-17, 44% were participants in the phenomenon, compared to 38% of those
ages 18-24 (2000b). Webnoize (2000) surveyed 4,294 students at ten New England
colleges in 2000 and found that 57% were at least weekly users of Napster. Other
important ﬁndings were that 73% of students used Napster on a regular basis, and that the
majority would be willing to pay $15 a month to do so. Partially illustrating the impact
that digital music has had on the use of CDS, the study also found that 22.8% spent
signiﬁcantly less time with their CDS, and that 63% listened to music downloads more in
the current year than in the previous year. Conversely, though, the USC Annenberg
School of Commurrication’s Norman Lear Center surveyed 275 college students at USC
and found that 63% were purchasing the same amount of CDS as they had prior to the
technology’s introduction (Latonero, 2000). The USC study also found that 69% of those
surveyed downloaded MP3s, 70% learned about MP3s through a close social network,
and that 69% believed copyright holders should be remunerated for downloads of their
music (Latonero, 2000).

In June 2000, the Pew Internet and American Life Project examined data from

over 2,503 individuals over the age of 18 and ascertained that over 1/5th (21%) of the

31

1,345 who use the Internet have downloaded music ﬁles, and 14% have downloaded
music that they do not own in any physical form (i.e., on CD) (Pew Internet & American
Life Project, 2000). Interestingly, the researchers also determined that approximately
37% who downloaded MP3s at the time the research was conducted were college
students. With regard to the breakdown of participation among age groups, those
between the 18-29 constituted 48% of users, and those between 30-49 comprised 42%,
and those over 50 made up 9% (Pew Internet & American Life Project, 2000). Auguring
bad news for independent online startups at the time who hoped to capitalize on the sale
of music singles in MP3 format was the ﬁnding that only 2% of individuals thus far had
at that time paid for such a product (King, 2000b). It bears mentioning here that this
study suffers from external validity issues because of the exclusion of those under 18
from the sample.

In a large-scale study, the Digital Media Association commissioned a survey of
16,903 Internet users/music lovers between the ages of 13-39 on their purchasing
behavior since the arrival of MP3 technology. Two thirds (66%) of respondents asserted
that downloading music has led them purchase CDS of particular artists (DMA, 2000).
Canadian digital music enthusiasts reported similar feedback concerning the effect of
MP3 downloads on CD sales. In the aforementioned Angus Reid study, 72% of the 1,018
respondents said that their continued participation in the MP3 phenomenon would have
no diminishing effect on their purchasing of music CDs (Angus Reid Worldwide, 20003).

A survey conducted by Field Research Corporation and commissioned by the
RIAA, however, provided contrasting data. In order to determine the way that Napster

has affected their purchase of CDS, 2,555 Intemet-using college students were surveyed

32

over a two-week period in May 2000. When asked about its impact on purchasing music,
41.1% of respondents gave feedback indicating that Napster allowed them to stop buying
CDS, to buy less CDS, or to create their own CDS (Jay, 2000). Approximately one-third
(33.2%) of students had downloaded over seventy-ﬁve songs using Napster, and 56% of
this group indicated that the program’s easy retrieval of MP3s had displaced their
purchase of CDS (Jay, 2000). Finally, of those songs downloaded, less than 10% were
from CDS previously owned by the user, and less than 10% resulted in the purchasing of
a CD following acquisition. In their own ﬁinded study, the RIAA discovered that 40.4%
of 2,555 college students surveyed were unreceptive to purchasing music primarily
because Napster facilitates the free exchange of music ﬁles online (King, 2000a). It was
also found that approximately half of Napster users downloaded music which they had
never owned before, supporting the contention that music piracy is rampant. AS might be
expected, the RIAA used these data in their primary argument supporting a legal
injunction against Napster.

In a richer and more directed effort, Tag It, Inc. and MusicDish conducted an
online survey in May-June 1999 with collaboration from MP3.com to determine the
ramiﬁcations of the new technology on the music community. Targeting a- focus group
of 817 self-proclaimed “music enthusiasts” - the majority of who work in the ﬁeld - the
survey posited questions related to the perceived impact of MP3s on: the recording
industry, the artist, and the consumer; the value of music as a product to be paid for; and
their level of support for participation as paying consumers in music distribution over the
Internet. The research indicated that 47% of respondents were willing to pay for digital

music singles available online, and that 58% were willing to pay for full-length

33

recordings (Kibbee, 1999). Further, it was found that only 13% of respondents had
changed their purchasing habits of CDS because of MP3 technology, while 62% had not
changed such habits (Kibbee, 1999). This may be due to the relatively limited
availability of broadband technologies at the time, as the vast majority of Internet users in
1999 were forced to download songs over slow dialup connections which allowed for
speeds no greater than 56kbps (Cravotta, 2000; Petreley, 2000)13 . AS mentioned, MP3s
are usually 3-6 megabytes in size; accordingly, each ﬁle Would take between 15-25
minutes on average to download via the small datapipe afforded by a dialup modem.

Almost half (46%) of the respondents stated that they would pay to download a
music single, but most were willing to spend less than a dollar on each14 (Kibbee, 1999).
Also, the majority were unsure whether they would stop downloading MP3s if they were
charged for each ﬁle, and felt that a new copyright and licensing model for music on the
Internet needed to be developed and instituted so that artists could beneﬁt from digital
dissemination and earn royalties on their works (Kibbee, 1999).

Remarkably, only 43% of respondents stated that the majority of MP3s are used

illegally, indicating widespread misinformation and a lack of education of existing

 

‘3 If and when speedy, dedicated connections become the household norm, and when full length albums can
be downloaded in a few minutes by the majority of Internet users, it is plausible that more individuals will
stop purchasing CDS. To note, recent statistics from Nielsen NetRatings (2003) indicate that broadband
connectivity among Americans is continuing to increase, and has risen 49.2% from 26.1 million to almost
39 million from May 2002 to May 2003. Narrowband usage declined over the same period by 12.3% -
from approximately 79.4 million to 69.6 million (Nielsen Net Ratings, 2003). Overall, 13% of Americans
had dedicated high-speed online access, while 23% dialed in at speeds no greater than 56kbps to access the
Internet in 2003.

'4 In this author’s opinion and as largely evidenced by the review of MP3-related research, the term “MP3”
has seemingly been equated with the notion of “ﬁ'ee music,” and that relationship has been consistently
reinforced through the variations of ﬁle exchange programs such as Napster. This has made it very
difﬁcult (if not impossible) for users to begin to pay for something which ﬂiey have been obtaining at no
cost all along. The truth of this contention remains to be seen.

34

copyright laws15 . The June 16, 2000 issue of Rolling Stone magazine published a Similar
ﬁnding, where 71% of the 600 respondents to an email survey felt that stealing music is
acceptable behavior (Stone, 2000). To further illustrate their dogmatic beliefs on the free
nature of MP3s, 59% remarked that they would not be willing to purchase legitimate
digital music ﬁles online (Stone, 2000). The sample, however, is subject to internal
validity problems (such as a lack of randomization and the presence of self-selection
bias).

Finally, there is signiﬁcant evidence demonstrating that colleges are hotbeds for
digital song-swapping, fostered primarily because of the high speed, dedicated Internet
connections installed in residence halls (Davis, 2003; Healy, 2003; Latonero, 2000).
Because of this fact, VNU Entertainment Marketing Solutions analyzed approximately
9,000 independent music stores in the immediate vicinity of over 3,000 colleges and
universities in an attempt to discover how MP3 distribution has affected sales of CDS. It
was discovered that while overall national album sales increased during the period from
1997 to 2000, sales in college towns by independent stores decreased approximately 4%
(Learmonth, 2000; Reciprocal Inc., 2000a, 2000b). Contrary to intuition, in those towns
where Napster was banned from usage due to the inordinate strain it caused on the local
university’s network, music sales declined 7% over that same period. However, critics
such as the CEO of Napster attributed the decline in sales at independent record stores to
the increase in sales at giant retailers such as Wal-Mart (Learmonth, 2000). Others

argued that college students are buying less CDs from independent stores because they

 

'5 Only the copyright holder has the authority to distribute and reproduce his or her music. Copyright law
is discussed in detail later in this work.

35

obtain new music through shopping for used CDS and by duplicating the CDS owned by
ﬁiends with CD burners16 (Learmonth, 2000).

As is evident, none of these studies analyzed behavioral inﬂuences or motivations
that conduce to the criminal activity, and no theoretical perspective was used to create the
hypotheses or to shape inferred conclusions. Moreover, the methodology and research
design of most was questionable, and the underlying motive for commencing the study
was typically manifested in the suggested policy solutions. It is predictable, then, that no
dominant theme consistently emerged concerning the relationship between MP3s and the
economic health of attendant players in the industry. Nonetheless, a rough sketch of the
prevalence and pervasiveness of participation in the MP3 phenomenon was obtained
through such research.

Very recently, a few exploratory academic studies utilizing small sample sizes
have been published concerning the MP3 phenomenon. For instance, in a sample of over
200 college students from 2000-2001, (Bhattacharjee et al., 2003) found that price of
music and available bandwidth are positively related to participation in digital music
transfers, and that the issues of digital music ﬁdelity and level of income were not
signiﬁcantly related to downloading “known,” “favorite,” or “popular” music tracks. In
another study, (Agrawal, Gopal, Sanders, & Wagner, 2003) developed a structural
equation model of behavioral determinants related to music piracy, and included: ethical
inclinations; conceptions of justice and belief in laws; “club size,” where individuals
partner with others to increase the availability of desired music amongst the group;

income; gender; age, and amount of money saved. The model provided a good ﬁt to the

 

‘6 This activity is also illegal, and is a separate issue beyond the scope of the current research.

36

data retrieved from surveying 133 undergraduate students primarily majoring in business
and in their third year of school.

As expected, ethical individuals and those with strong conceptions of justice were
less likely to commit intellectual property theft for the purposes of providing music to
others. Older individuals participated less in the activity, as would be expected. The
amount of money saved was a strong predictor of club size, as a greater perception of
reward (i.e., not paying for comparatively expensive music) increased the likelihood that
the individual would exchange music with others. Income was not found to signiﬁcantly
predict club size. The researchers also identiﬁed that a subsample which received an
article delineating the prosecution and penalties of a college student who distributed MP3
ﬁles did not have a signiﬁcant effect on club size. This research is the ﬁrst to
conceptualize music piracy from a behavioral perspective, and therefore merits accolade
despite its incompleteness.

Banerjee et a1. (1998) has stated that piracy is a result of decisions that individuals
consciously make. Other scholars (Agrawal et al., 2003; Gopal & Sanders, 1997, 1998;
Irn & Van Epps, 1991; Kievit, 1991; Thong & Yap, 1998; Wong, 1995) have asserted
that the decision to pirate is inﬂuenced by individual ethical conduct. While such
statements stimulate inquiry into the cognitive impetus for behavior, this author feels that
a host of additional factors (e.g., cognitive, behavioral, psychological, and sociological)
also play a contributory role. That is, a complete understanding of the etiology of the
piracy phenomenon - and any phenomena in general - requires a more explicit

speciﬁcation of all the elements that explain some proportion of the variance in the

37

behavior and that cumulatively engender its execution. Any other attempt is
reductionistic and simplistic at best, and crude and rudimentary at worst.
Theft, Law, and Ambiguity — History and Relevance

Theft, law, and ambiguity all intersect to provide some insight into the etiology of
the dishonest acquisition of property. The common law deﬁnition of larceny serves as
the historical starting point for theft and dishonest acquisition of property. Traditionally,
a trespassory taking was necessary in order for larceny to occur, and larceny was only
applicable when considering certain forms of property. Over time, this law has evolved
to encompass more forms of the act and more forms of the property that can be
dishonestly acquired. The maintenance of the social order has also been paramount to the
development of theft laws, as legal mandates tend to deﬁne and uphold the boundaries of
behavioral propriety to preserve the social, political, and economic system in place.
Jerome Hall, in his senrinal work Theft, Law, and Socieg (1935), points to an appropriate
example which arose in England in the late 14005 when the country was transitioning
item a feudal and agricultural economy into one based on trade and mercantilism.

In this setting, merchants who sold goods to customers would hire individuals, or
carriers, to transport and deliver the goods on a horse-drawn cart. Some carriers,
however, decided to keep the goods for themselves. At the time there was no law which
deﬁned such an action as illegal because social norms dictated that goods belonged to the
individual who had possession of them (Hall, 1935). That is, while the goods were with
the carriers, they belonged to them, and no theft had occurred if the carriers chose to
appropriate what was in their possession. Any harm that befell merchants was their own

fault, because they had chosen to hire someone untrustworthy to deliver their goods.

38

Notwithstanding, the English merchant class vociferously demanded that this activity be
deemed illegal, though their number was much smaller than the population of poor
English men who were the ones fulﬁlling the role of carrier and delivering the goods.

In 1493, judges who came from well-to-do backgrounds and who did not
represent the interests of the majority population of the lower class ruled in favor of the
merchants and created a new crime and deﬁnition of theft that outlawed the retention of
deliverable goods by carriers (Hall, 1935). This decision safeguarded the economic
interests of the privileged merchant class to the detriment of the poor, and violated social
norms in the process - all for the purposes of curtailing behaviors which potentially
threatened the status quo. Over the centuries, the criminalization of dishonest acquisition
has expanded by leaps and bounds, and the contributive force of economic and political
interests cannot be overlooked or dismissed.

Ambiguity in the actual content and application of the law during the Mercantilian
Revolution appears to have contributed to the crime. For example, common sense in the
21St century would indicate that keeping something owned by another but entrusted to
one’s care is unethical and illegal. In the 15th century, though, it was a socially
acceptable behavior and no alarm or question was raised when it occurred. If carriers .
were well-versed in deﬁnitions of property and made conspicuously aware of the
wrongﬁIl nature of misappropriating propriety that belongs to someone else, perhaps no
problem would have arisen. Some ﬁve hundred years later, ambiguity in the Information
Revolution areas appears to once again be relevant in contributing to illegal activity - this

time concerning intellectual property theft over the Internet.

39

Arguably few carriers were aware that their actions might be criminal since the
inception of their occupational duty, but were made aware of the unacceptability of their
behavior through the visible processes and outcomes of the legal system. Similarly, it
can be posited that a signiﬁcant number of MP3 enthusiasts do not completely understand
the illicit extent of their point-and-click actions online. Even if news and media outlets
have introduced that notion to them, it is highly unlikely that they thoroughly
comprehend why their behavior is inappropriate and the reasoning behind policy intended
to restrict and penalize such activity”. Nonetheless, uploading and downloading music
ﬁles would not be a crime if the content being exchanged had no commercial or personal
worth. Any creative work, however, is fundamentally imbued with value and the
usurping of that value without proper authorization is the issue at hand.

The Value of Information and Intellectual Property
“Although intellectual property has become a salient topic
in economic and political circles, it has generally escaped
the attention of criminologists. Such negligence is
unwarranted. It is time to grant that intellectual property is
as valuable as customary forms of property, that its
infringement is as signiﬁcant as burglary and robbery, and
that its violation and protection merit careﬁIl investigation.”

(Luckenbill & Miller, 1998).

 

'7 Though not relevant to the primary hypotheses of this paper, a few questions have been included in the
survey instrument to determine whether respondents believe obtaining or distributing MP3s is a crime,
whether it should be a crime, and how such conceptions condition their participation in the phenomenon.
The instrument is described in detail in the methodology section of this work.

40

Seminal in part to the current research was an interesting article in Listic_e
Quarterly by David F. Luckenbill and Susan L. Miller (1998), who assessed the relevance
of two competing theories related to intellectual property. The ﬁrst was termed the
“intellectual property protection” argument, and speculated an increase in laws protecting
the creative works of authors and punishing violators of copyright. The second was
termed the “intellectual property access” argument, and projected a decrease in protective
laws and prohibitive actions concomitant with a rise in the amount of legislation that
increased access to creative works.

The authors described how the development of exploitative information
technologies foreboded and facilitated the misuse and misappropriation of works without
proper remuneration and rights that should be duly afforded to the originators and owners
(Dordick, 1986). Home satellite dishes, audio-to-digital converters, decoding boxes,
videocassette recorders, and audiocassette recorders were articulated by the researchers
as technologies that have greatly enabled individuals to acquire, copy, alter, and
distribute intellectual property to which they have no legal right (Luckenbill & Miller,
1998). It was predicted that as these and similar technologies became more prevalent,
intellectual property crime (and the difﬁculty in addressing it) would rise (Ofﬁce of
Technology Assessment, 1986). Luckenbill and Miller’s paper was published in 1998, a
year in which the MP3 phenomenon was still gernrinating and yet to fully sprout. It can
be argued that digital music has had much greater ramiﬁcations than the aforementioned
devices for the distribution of creative works due to its global scope and availability on

the Internet.

41

Historically, government agencies paid attention to traditional types of property
crimes but ignored those of the intellectual variety (Luckenbill & Miller, 1998). By the
turn of the 20th century, though, supporters of rights restricting the copying and
dissemination of creative works had successﬁrlly compelled the state to propose and
enact legislation in line with their position after developing strong political and economic
ties (Betti g, 1996). As time went on and more effort was expended by owners of
intellectual property to seek the assistance of the government to protect their interests,
lawmakers mobilized to quell copyright concerns that the entertainment industry voiced
following technological advances such as the player piano (in 1908), broadcast radio
(1931), photocopies (1968), the VCR (1976), the DAT recorder (1990), and portable
MP3 players (1998) (Schoen, 2002).

Luckenbill and Miller (1998) assert that a primary reason for the historical apathy
of the state towards protecting intellectual property may have hinged on the fact that most
individuals lacked the capacity to violate these laws. For example, many recording and
duplicating devices were prohibitively expensive for many years, but the cost of these
pieces of hardware eventually became quite affordable for the average person.
Furthermore, in recent years the advent of the Information Age has facilitated the
opportunity and means to covertly distribute copyrighted data without a signiﬁcant threat
of detection, apprehension, and punishment, and at a comparatively small cost”. Thus,
the combination of lower participation costs and the greater number of deviant

possibilities to exploit has jointly provoked legislators to action.

 

'8 All that is needed is a computer, an lntemet connection, and some freely available software.

42

Researchers also have largely ignored the importance of studying intellectual
property, disproportionately focusing attention on conventional conceptions of property
(Gilbert & Lyman, 1989; Reiman, 1995). Legal scholars, however, have studied it to a
great extent, and continue to do so. Their interest resides primarily, though not
exclusively, in the semantics of laws as crafted and delineated in the books and in their
consequent application”. Analyses of the prevalence, role, and efﬁcacy of intellectual
property laws have been extremely limited - a deﬁciency which Luckenbill and Miller
attempt to address through their work. They examined federal legislative action
involving copyright on both civil and criminal levels from 1949 through 1992, and
uncovered a host of interesting ﬁndings. First, signiﬁcant growth in the amount of
legislation over this time period took place, and 91.5% of the 423 bills that were
introduced favored the owners of copyright rather than the consumers (Luckenbill &
Miller, 1998). Simply put, as the private sector became increasingly vociferous in
petitioning the state to support their interests, a greater number of legislation was
proposed and passed.

The scholars also analyzed civil copyright cases compiled by Administrative
Ofﬁce of the United States Courts from 1955 to 1993, and criminal copyright cases
compiled by the Executive Ofﬁce of the United States Attorneys from 1997 to 1993.
Civil complaints increased from approximately 300 per year in the 19505, to about 2,000
per year in the 19908, as a greater number of civil Violations were deemed worthy of
pursuing formally, while the amount of criminal complaints generally decreased ﬁ'om 93

in 1977 to 12 in 1989 (Luckenbill & Miller, 1998). With regard to the actions of the

 

'9 See e.g., Lessig (1997; 1999a)

43

courts, they acted on a larger proportion of civil cases but a smaller proportion of
criminal cases over the years in which the data were available. Finally, of those that were
handled by the courts, the researchers found that many ended in guilty convictions and
relatively few in dismissals (Luckenbill & Miller, 1998).

Also examined were statistics compiled by the Motion Picture Association of
America to obtain an understanding of investigations into, and legal resolutions of, ﬁlm
piracy. An increase in the amount of criminal action (32.6% in 1986 to 47.5% in 1994)
and a signiﬁcant decrease in civil action (67.4% in 1986 to 30.6% in 1994) was identiﬁed
(Luckenbill & Miller, 1998). Over that same time period, the rates of sentencing and
conviction of intellectual property offenses remained relatively stable, while the severity
of sentences declined.

Overall, the ﬁndings suggested that the legislators supported the private sector
and the courts were more diligent in attending to intellectual property cases, but law
enforcement remained apathetic in fervently seeking to identify and apprehend copyright
criminals. Two reasons are suggested by Luckenbill and Miller (1998). The ﬁrst is that
the copyright laws were implemented to promote a symbolic purpose and to proclaim a
certain value system, rather than as speciﬁc delimiters of behaviors that would be
pturitively addressed. The second revolves around the issue of limited resources and
expertise to identify and combat intellectual property violations, coupled with the
politically- and socially-mandated focus on traditional personal and private property
crimes (to the exclusion of nontraditional forms of illegality). It may be that civil cases
are the only viable option because of the restrictions that are placed on the activities of

law enforcement entities (Luckenbill & Miller, 1998).

44

Copyright Law

Those who illegally publish the material of musicians on the Internet in MP3
format tend to rationalize the questionable nature of their activities. This can take the
form of “disclaimers” on a web site offering unauthorized free music, such as those that
qualify the presence of MP3s on their page as “promotional” or “educational.” Others
caveats used as justiﬁcation include, but are not limited to, the following: “Any ﬁles
downloaded must be deleted from your computer within 24 hours,” “The author is to be
held blameless in all respects for the data available on this site,” “The author of this site is
not responsible for the illegal transfer and possession of copyrighted material by its
visitors,” or “This site is non-proﬁt, and is only providing an evaluation service to visitors
before purchase of the respective music CDS” (RIAA, 2000h). Rationalizations are also
manifested in participants’ outcries that MP3s are not absolutely CD quality, that “clips”
of songs are legally acceptable, that the music being downloaded is for personal use only,
that there is no proﬁt being made, or that a site only containing hyperlinks to MP3s on
someone else’s ﬁle server is legal because it does not actually store the copyrighted
material.

These justiﬁcations are all invalid in light of the fundamental principle of -
copyright - only the owner has the lawful ability to distribute or reproduce that creative
work, and anyone or anything that directly or indirectly contributes to unauthorized
dissemination or duplication of another’s intellectual property is committing a crime.
Disregard for the copyrights of intellectual property, manifested through the purposeful
dissemination of unauthorized digital music ﬁles, is a federal offense (RIAA, 2000e).

The illicit activity falls under the auspices of “Internet crime,” which can be deﬁned as

45

any illegal act fostered or facilitated by the Internet and a computer, whether the
computer is an object of a crime, an instrument used to commit a crime, or a repository of
evidence related to a crime (Royal Canadian Mounted Police, 2000).
As quoted from the United States Copyright Ofﬁce, the owner of a copyright has

the exclusive right to do (or authorize another to do) the following:

To reproduce the work in copies or phonorecords;

To prepare derivative works based upon the work;

To distribute copies or phonorecords of the work to the

public by sale or other transfer of ownership, or by rental,

lease, or lending;

To perform the work publicly, in the case of literary,

musical, dramatic, and choreographic works, pantomimes,

and motion pictures and other audiovisual works;

To display the copyrighted work publicly, in the case of

literary, musical, dramatic, and choreographic works,

pantomimes, and pictorial, graphic, or sculptural works,

including the individual images of a motion picture or

other audiovisual work; and

In the case of sound recordings, to perform the work ‘

publicly by means of a digital audio transmission.

(Copyright OfﬁCe of the United States, 2000a)
The term “copyright” is deﬁned as the legal right granted to an author, composer,
playwright, publisher, or distributor, to exclusive publication, production, sale, or
distribution of a literary, musical, dramatic, or artistic work (de Fontenay, 1999).
Copyrights cover both published and unpublished works, and are secured immediately
upon the expression of an original work in ﬁxed, tangible form (Copyright Ofﬁce of the

United States, 2000a).

46

Sound recordings have two copyrights, one on the underlying musical work (notes
and lyrics), and one on the actual recording itself (the arrangement and layering of the
performance by the artist, the backup singers and musicians, the producers, the sound
engineers, as written to a physical medium (e. g., cassette tape or CD)) (Harari, 1999).
Each copyright grants the owner explicit and sole permission to modify, distribute,
reproduce, perform, or display the work. With the uploading and downloading of digital
music over the Internet, however, these copyrights are violated. For instance, uploading
an MP3 to a web or ﬁle server that can be accessed by others through their web browser
or through a ﬁle transfer program is a form of distribution. If the copyrighted work is not
owned or authored by the uploader, that person is breaking the law. When an individual
requests MP3s from a web or ﬁle server, or uses a ﬁle exchange program to download
MP3s onto his or her hard drive, an exact copy of that sound recording is made on the
recipient’s computer system. This violates the reproduction tenet of the copyright law, as
non-owners must have explicit permission to duplicate protected works, whether for
proﬁt or merely for personal listening pleasure, and regardless if it is for a transitory or
permanent period of time.

The RIAA stresses two legal concepts that come into play with Internet music
piracy - copyright infringement and Vicarious liability. When a person knowingly
facilitates violation of copyright, infringement has taken place (RIAA, 2000c). This can
occur online as web sites link to other servers that host MPBS, even if the illegal ﬁles are
not housed directly on the initial web site. Another instance might be when a person sets
up an automated system to advertise MP3 ﬁles for download in a chat channel. While

that person is not speciﬁcally initiating the transfer of ﬁles, s/he is making available

47

MP3s for distribution without the consent of the copyright holders. Vicarious liability
occurs when a person is able to control the activities of a copyright violator and fails, and
also receives some pecuniary beneﬁt from his or her role in facilitating infringement
(RIAA, 2000e). While contrasting opinions proliferate, the law has ruled that ﬁle
exchange programs such as Napster are vicariously liable for copyright inﬁingement
(Borland, 2000).

There are a few pieces of important legislation that necessitate mention to provide
a richer understanding of the attendant legal issues. Each is explicated below, as is its
relevance to the current controversy surrounding digital music distributed over the
Internet.
US. Copyright Law jTitle l7 U.S.C. Section 101 et 860.. Title 18 U.S.C. Section 2319}

There are ﬁve speciﬁc rights exclusively granted to the copyright holder under
Section 106 of the Copyright Act. These include reproduction, adaptation, distribution,
public performance, and public display. Individuals who exercise these rights without a
license from the copyright holder are committing inﬁingement, and thereby subject to
penalties unless shielded by the Fair Use Doctrine (discussed below). An individual can
be held civilly liable if s/he infringes on a copyright unknowingly, or without forethought
or speciﬁc intention. Criminal liability can occur if an individual duplicates copyrighted
intellectual property for the purposes of obtaining proﬁt or “gain” from it. This is not
limited to ﬁnancial returns, and can include the possibility of denied revenue to the artist.
When the illegally reproduced works are used for commercial advantage, resultant
penalties include incarceration for up to ﬁve years and ﬁnes up to $250,000 (Copyright

Ofﬁce of the United States, 2000b). Additional civil liabilities may include payment for

48

damages incurred by the copyright holder, or statutory damages of up to $150,000 per
inﬁinged work.
Fair Use Doctrine

The doctrine of “fair use” from the 1976 Copyright Act, Section 107, allows a
user to duplicate a copyrighted work for educational or research purposes such as
criticism, news reporting, teaching, or scholarship, as long as the work is not used for
proﬁt and its potential value is not negatively affected (Copyright Ofﬁce of the United
States, 2000b; RIAA, 2000e). The value of a song, however, can be impacted even if
only a small clip of it is expropriated, regardless of how high the ﬁdelity is, and
irrespective of the fact that no monetary gain is derived. “Proﬁt” can constitute any form
of received beneﬁt outside the exceptions in this clause. Moreover, distribution of the
work over the Internet for the purposes of exchanging commercially-produced music
without remunerating the artists does not fall under the exemptions of the “fair use”
doctrine.
The Audio Home Recording Act (AHRA) of 1992

Arising from the development of physical digital playback products such as
MiniDisc (MD) and Digital Audio Tape (DAT) players, the AHRA required
manufacturers to pay a royalty from the sale of each device and device media sold to
musicians, songwriters, and record companies as compensation for lost revenue, and to
implement mechanisms to prevent serial copying or multi-generation duplication. The
Diamond RIO PMP300 player was claimed by the RIAA to be in violation of this law.
However, the US. 9m District Court of Appeals ruled against the RIAA’S charge by

holding that the R10 was not an audio recording device, but rather a playback device

49

incapable of intrinsically facilitating music piracy (CNN.com, 2000c; MP3.com, 1998).
Consumers were given permission to make private, non-commercial copies of
copyrighted music with these devices and exempted from litigation for infringement.
Incidentally, a computer is not covered under the auspices of the AHRA, as it is not
solely designed for digital audio playback and recording, and has multiple noninﬁinging
purposes as well.

The No Electronic Theft Act (NET)

The NET Act, signed into law by President Clinton in 1997, speciﬁes that
copyright violations are now criminally prosecutable and punishable with up to $250,000
in ﬁnes and three years in prison, even when there is no proﬁt motive to the activity
(RIAA, 2000g). Those who derive ﬁnancial gain from the behavior can be imprisoned
for up to ﬁve years and/or be liable for up to $250,000. Additionally, offenders may also
be found civilly liable for damages of up to $150,000 per copyright inﬁingement.
“Financial gain,” according to the law, is not necessarily restricted to pecuniary income,
and also includes the receipt (or expectation of receipt) of a valued item, which can
include MP3 ﬁles and other digital intellectual property (Congress, 1997).

The Digital Performance Rigl_rt in Sound Recordings Act (DPRA) of 1995
The DPRA afforded copyright owners of sound recordings (distinguished from

 

“musical works”) control over the public performance of their work, such as granting or
denying permission for digital dissemination and broadcasting (RIAA, 2000c). It also
allowed for artist compensation when their works were transmitted digitally, excluding

the mediums of radio and television. Previously, copyright owners of sound recordings

50

were not allowed to authorize public performances of their work; this law enabled them
to do so.
The Digital Millennium Copyright Act (DMCA) of 1998

The DMCA essentially criminalizes any act of circumventing copy protection. AS
such, an individual may legally make MP3s from a music CD, unless that CD is copy-
protected. An increasing number of CDS are created with technological restrictions to
prevent digital audio extraction and subsequent conversion to MP3 ﬁles. According to
the DMCA, then, any action that attempts to bypass the protection in place - even for
ostensibly legitimate purposes - is unlawful and subject to sanctions. The action may be
to make a backup of a CD for personal use — either to one’s hard drive or to a CDR — and
is actually legal. However, if circumvention of a protective control is necessary to
accomplish that goal, the action becomes illegal, rendering the DMCA a law that
“extends rights to consumers even as it effectively prevents them from exercising those
rights” (Harmon, 2001).

Secondly, the DMCA amended the DPRA to cover transmission over the Internet,
as well as through cable and satellite services (Copyright Ofﬁce of the United States,
1998). Finally, the liability of Internet Service Providers (ISPs) was clariﬁed with the
creation of the DMCA. It was determined that ISPs are not responsible for keeping tabs
on what their customers transmit online or post to web pages and ﬁle servers for others.
However, if an ISP is aware, or is made aware, of copyright inﬁinging practices, that

business has a legal obligation to act accordingly and remove the material, or risk facing

liability (RIAA, 2000c).

51

Copyright laws are in place to give incentive to artists to innovate and produce
creative works. It can be argued that if the Virtual community of today is not able to
respect the intellectual property of musicians and record companies, and continues to
download and distribute digital audio ﬁles without authorization, the Internet users of
tomorrow will be even less likely to attend to ethical behavior. A downward spiral will
consequently ensue, possibly resulting in a complete and utter disregard for creative
“fruits” of labor. Ultimately, this effect would have the potency to severely debilitate the
originative and inventive dogmas of our culture. And, as more and more aspects of our
lives are affected (and more of our needs are met) through the capacities afforded by the
Internet, this will drastically affect the structure and function of our networked economy
and society.

Civil and Criminal Sanctions Resulting From Copyright Violation

Law enforcement acknowledges that while the chances are very slim that a person
will get arrested for intellectual property theft, it is still a possibility (Spring, 2000).
Other detrimental outcomes are more likely, however. For instance, upon joining a ﬁle
exchange network like Napster, a computer system is rendered a little more vulnerable to
viruses and hackers. Additionally, by opening up a ﬁle directory on a computer system
so that others can download your ﬁles, that computer has become a veritable conduit of
music piracy and subject to apprehension and prosecution (Spring, 2000).

To illustrate, the US. Attomey’s ofﬁce obtained the country’s ﬁrst conviction
under the NET Act in August 1997 against Jeffrey Levy, a 22-year old University of
Oregon student who distributed thousands of songs, software, and movies after setting up

his computer as an accessible ﬁle server (Patrizio, 1999; Roth, 1999). In 1998, a suit was

52

ﬁled against a 20-year old junior at Arizona State University for posting approximately
50 copyrighted works by popular artists on his web site (Lipton, 1998). Another suit was
ﬁled in Washington against an individual who made over 1,100 copyrighted songs
available on a server (Lipton, 1998). In October 1999, Carnegie Mellon University used
specialized software to detect potentially copyright-inﬁinging materials on the dorm
room network of a random selection of 250 students. They subsequently disconnected
Internet service from the 71 individuals who were found to be hosting illegal music
archives, and stated that Internet access could be regained following attendance at a
discussion forum on the topic of copyrights (Mendels, 1999; Philipkoski, 1999a). Penn
State, in the Spring of 2003, revoked broadband Internet access from 220 students in
residential halls after it was discovered that they were committing intellectual property
theft (Davis, 2003).

Finally, in a lawsuit that made headlines largely because of the incredible amount
of possible ﬁnancial penalties, four students at Princeton, Michigan Technological
University, and Rensselaer Polytechnic Institute were charged in 2003 with creating and
adnrinistrating local area ﬁle-sharing networks that made available over one million
songs to other students (Healy, 2003). With possible damages up to $150,000 per song,
ﬁnes might have totaled $150 billion dollars; however, a settlement was reached and the
students were each required to pay between $12,000 and $16,000 in restitution for their
criminal activities. This was the ﬁrst time that monetary penalties have been exacted
from individuals for music piracy in the United States, and perceivably sought to ‘
demonstrate that the illegal activity is seriously regarded and will not be tolerated. In

January 2004, the RIAA initiated legal action on 532 individuals who offered large

53

amounts of copyrighted music to others through P2P ﬁle exchange software (CNN.com,
2004). This followed the ﬁling of 261 suits against similar MP3 participants in
September of 200320.

Online businesses, too, are coming under increased scrutiny for their potentially
inﬁinging practices. For instance, Napster defended its services by reasoning that its
software allowed users the same privileges that the Diamond Rio portable MP3 player
accorded; that is, the ability to duplicate copyrighted works for private, non-commercial
purposes (CNN.com, 20000). The RIAA argued that the service was a haven for online
piracy and violated the AHRA. Napster countered that it was only providing a
technology which serves as a conduit for users to exchange music, and that it was not
responsible for, nor could it possibly control, the unethical and illegal behavior of its user
population of millions (CNN .com, 2000c).

MP3.com was sued by the RIAA in January 2000 and later found guilty of
copyright infringement resulting from their practice of allowing users to create digital
“lockers” (Breen, 2000; CNN.com, 2000b; RIAA, 2000b; Swiatecki, 2000b). Music
CDS could be placed in one’s CDROM drive at the home or ofﬁce, and information
would be sent to the web site detailing the artist and album of the particular disc. Then,
upon noticing that the end user did in fact possess the physical CD (indicating actual or
purported ownership), previously created MP3s of each track were placed in the
individual’s personal locker. This consequently allowed the user to access his or her CD
collection, in the form of streaming digital MPB tracks, from any computer at any

location simply by connecting to the MP3.com website. Thus, individuals could access

 

2° It remains to be seen whether this initiative will deter music pirates, alienate them from purchasing

54

their music collections remotely without having to constantly carry around the physical
discs themselves. Apart from placing a CD into one’s CDROM drive and connecting to
the MP3.com website — a service named “Beam-It” allowed members to have a digital
copy of an album placed immediately into their locker after purchasing that album from
an online CD retailing partner. This was known as MP3.com’s Instant Listening Service
(RIAA, 2000b). .

To do this, and allow for millions of individuals to have digital copies of their CD
collections online, MP3.com “ripped” (the term used for extracting digital audio from a
CD into a waveform audio ﬁle) and “encoded” (a term referring to the compression of a
waveform ﬁle into an MP3) approximately 45,000 albums, an activity for which they did
not have permission from the artist or record company to do (Swiatecki, 2000b). The
RIAA made the case that the service offered cannot determine ownership of a CD, and
only relies on possession of a CD to give users the ability to listen to that music in digital
format on demand from MP3.com. Further, the industry claimed that MP3.com does not
have the legal right to broadcast, stream, or otherwise disseminate creative content and
property that belongs to other musicians as well as the respective recording companies
without a license or approval (Jones, 2000; Rosen, 2000). MP3.com countered that the
music industry should not be able to control how a purchaser of a CD listens to that
music. Additionally, MP3.com stated that it was beneﬁting artists to a much greater
degree than the RIAA is by providing a direct link to their target audience, and well as
the tools to promote their music in a new, proﬁtable, and proliﬁc distribution model

(Robertson, 2000).

 

music, or both.

55

A settlement occurred a short time later, as the US. District Court ruled in May
2000 that MP3.com was guilty of copyright inﬁingement due to its unauthorized
appropriation of the property of the recording industry, its music companies, and its
artists. Further, the “Beam-It” and “Instant Listening” services were found to facilitate a
blatant disregard for the rights of the music industry to control and license the creative
content it owns to others (Menta, 2000; MP3 Newswire, 2000). This ruling will serve as
precedent and has the potential to shape the future of online music business ventures by
reinforcing the requirement that all delivery of music to the end user requires licensing

and permission from the copyright holders.

56

CHAPTER 3: THEORETICAL BACKGROUND AND APPLICATION

The current research attempts to clarify the cognitive, behavioral, psychological,
and sociological inﬂuences of online intellectual property theft in the form of digital
music piracy. Towards that end, three general criminological theories appear applicable
in inducing the phenomenon. These include General Strain Theory, Self-Control Theory,
and Social Learning Theory, and prior to empirically analyzing their relevance as
explanatory frameworks, it is important to provide a detailed description of their precepts.
In the subsequent sections, each theory is introduced, and a number of studies which have
respectively tested the salience of their approach are reviewed and discussed.
Additionally, a detailed breakdown of how certain elements of the theories might impel
an individual to engage in music piracy is posited for the purposes of explaining to the
reader why the commission of the crime might occur.
General Strain Theory

Strain, in essence, is the maladaptive response experienced by some individuals
who seek to attain culturally or socially promulgated goals, but are thwarted by a variety
of hindrances. This consequently leads them towards goal achievement via unethical or
illegal means, or towards harmful responses at the perceived sources of their strain
(Agnew, 1985). Since its initial promulgation by Robert Merton (1938; 1968), the
concept of strain has been reﬁned by a host of prominent sociological and criminological
scholars including Cohen (1955), Cloward and Ohlin (1963), Agnew (1985; 1989; 1992);
and Messner and Rosenfeld (1994). Agnew’s conceptualization has received the most
attention and empirical examination in recent years, and has been proffered as a general

theory capable of explaining all types of deviance and criminality. It attempts to

57

understand the detrimental effect that immediate social and environmental pressures can
have on a person, as manifested through the affective vehicles of anger and frustration.

In his General Strain Theory (GST), Agnew (1992:50) vocalized three primary
types of strain that may affect an individual: the threatened or actual failure to achieve
positively valued goals; the threatened or actual removal of positively valued stimuli; and
the threatened or actual presence of negatively valued stimuli. The ﬁrst type is
exempliﬁed by the notions of classic strain theory through its focus on the disjunction
between aspirations and expectations/achievements — or, the ideal and the real. Also
implicated are idealized conceptions of fair, equitable outcomes with those that actually
occur. That is, certain emotive responses often result from an individual’s failure to live
up to certain expectations or from experiences of perceivably unjust outcomes, which
arguably lead to deviant methods of coping or compensating. The second type of strain
regards the removal of certain positives in a person’s life, such as healthy ﬁiendships,
relationships, or environments (e.g., involving home, school, or work). For instance,
stressful life events have the tendency to incite feelings of pain, anger, and frustration
(and arguably subsequent criminality) as the strained individual attempts to prevent or
come to terms with circumstance.

The third type of strain is the presence of irritating, frustrating, angering, painful,
or otherwise noxious factors in a person’s life. These may stem from social,
environmental, or relational inﬂuences, and delinquency might ensue as the individual
attempts to manage, curtail, or eradicate its effect. To reiterate, Agnew expands the
concept of strain to include not only the thwarting of goal attainment, but also the

removal of conducive entities and the introduction and persistence of detrimental entities

58

in one’s life. Also asserted by Agnew (1992) is the magnitude, recency, duration, and
clustering (occurring closely together in time) of strainful events, and the positive
relationship between those factors and the adverse impact of strain.

To note, certain elements moderate the link between strain and delinquent
outcomes, such as the availability of coping resources and positive environmental and
social support, as well as differences in personality, temperament, and aspirations Agnew
(Agnew, 1992:71). For example, those individuals with an internal locus of control, with
normative levels of self-conﬁdence and self-efficacy, and who associate primarily with
law-abiding peers will be less likely to engage in deviance to cope with strain (Agnew,
1992). Accordingly, Hoffman and Miller (1998) have argued that strain is not an
“isolated cause of delinquency, but a facilitative mechanism that interacts with coping
strategies to increase the probability of delinquent behavior.” These conditioning factors,
then, must be acknowledged and accommodated when attempting to test the viability of
the relationship between strain and criminal outcomes.

Empirical Support for General Strgin TheoLv

Empirical research since the theory’s “general” reformulation in 1992 was
initiated by Agnew himself, and followed by an assortment of other rigorous studies. In
the ﬁrst theoretical test, Agnew and White (1992) studied data collected from the Rutgers
Health and Human Development Project, a longitudinal study which interviewed New
Jersey youths about their experiences with delinquency and drug use. They ﬁrst created
multiple scales of strain, including those representing negative life events, life hassles,
negative relations with adults, parental ﬁghting, neighborhood problems, unpopularity

with the opposite sex, occupational strain, and clothing strain. When contemporaneously

59

measuring strain’s inﬂuence on delinquency and drug use on 1,076 kids, the researchers
identiﬁed a signiﬁcant link between stress around the home and the two outcomes. They
also discovered that the relationship between strain and these antinorrnative behaviors
was conditioned by delinquent peer associations, social control elements, and self-
efﬁcacy, which was deﬁned as “perceived personal control over the environment”
(Agnew & White, 1992:488). In terms of the amount of variation explained, R2 values of
.402 for delinquency and .489 for drug use were derived, inclusive of all of the
aforementioned predictors. Next, the relationship between a summary measure of strain
at time 1 and delinquency and drug use at time 2 (three years later) was explored among
798 youths. Strain was found to signiﬁcantly predict delinquency at time 2, and had a
larger effect than any of the differential association, social control, or self-efﬁcacy
elements (Agnew & White, 1992). It was not, however, related to drug use in the
longitudinal analysis.

Paternoster and Mazerolle (1994) performed a more extensive test of the theory
through longitudinal analysis of adolescents from the ﬁrst two waves of the National
Youth Survey. They employed a shorter lag period (one year) than the three-year gap
between re-interviews in Agnew and White’s (1992) study. In addition, they replicated
most sources of strain constructed in that previous work, and extended the analysis by
examining the interaction effects of strain with delinquent peer groups, self control, self-
efﬁcacy, conventional social support, and moral inhibitions (Paternoster & Mazerolle,
1994:240).

Despite the fact that none of the interactions were found to be signiﬁcantly related

to delinquency, some general support was discovered for the theory. Speciﬁcally,

60

negative relationships with adults and friends, negative experiences with school, various
stressful life events, and living in a noxious neighborhood environment were signiﬁcant
predictors of delinquency (Paternoster & Mazerolle, 1994). Also identiﬁed was the
facility of strain to weaken conventional social bonds and strengthen unconventional
bonds; temporal ordering of this relationship, however, was not clear because
measurement of the concepts occurred at the same time. In another longitudinal work,
(Hoffman & Miller, 1998) found that strain (operationalized as negative life events) was
signiﬁcantly related to changes in delinquency when controlling for other exogenous
factors. Contrary to one of Agnew’s hypotheses, however, self-esteem and self-efﬁcacy
appeared unrelated to the relationship between strain and delinquency.

In a study that offered strong support for the theory’s tenets, four scales - negative
relations with adults, school/peer hassles, neighborhood problems, and negative life
events - were employed by Mazerolle and Maahs (2000) as strain measures, along with
one composite additive scale of all the variables which comprised the aforementioned
facets. Utilizing data from the National Youth Survey, the researchers found a linear and
systematic relationship between delinquency and strain. Further, they found that
conditioning variables such as negative peer inﬂuence, low moral inhibitions, and a
behavioral inclination towards delinquency increased the likelihood of wrongdoing on
both a cross-sectional and longitudinal basis.

Broidy (2001) examined the relationship among strain, crime, and two mediating
variables - negative affective states such as anger and the availability of legitimate coping
avenues - among 896 undergraduate students. A signiﬁcant positive link was identiﬁed

between strain and anger, unfair outcomes and anger, and strain and negative emotions.

61

Interestingly, blocked goals reduced the likelihood that individuals responded to anger
with strain, while negative emotions were positively related to legitimate coping
mechanisms. Finally, negative emotions such as anger increased the likelihood of crime
commission. Anger has been identiﬁed as an intervening variable between strain and
criminal outcomes in a host of studies (e.g., Agnew, 1985; Agnew & Brezina, 1997;
Agnew, Cullen, & Burton, 1996; Brezina, 1998; Mazerolle & Piquero, 1998; Piquero &
Sealock, 2000), which have highlighted its importance as the medium through which
violence is manifested following exposure to strain.

Strain, anger, and delinquent behavior in the forms of Violence, drug use, and
school-related deviance were further explored by Mazerolle et al. (2000). The
researchers found that strain exhibited a direct effect on violence when controlling for the
precepts of differential association and social bond theory, and some demographic
measures. It was also found that strain mediated the relationship between anger and
violence. Strain, when it occurred to those who were angry, tended to produce violent
outcomes, while anger did not necessarily lead to violence unless strain was present
(Mazerolle et al., 2000). Such ﬁndings contrast expectations associated with GST but
emphasize the fact that strain can exert a criminogenic inﬂuence through anger.
Additionally, it was determined that strain and anger do not directly affect drug use or
school-related deviance at the bivariate level, and that anger did not serve a mediating
role between strain and such nonviolent outcomes (Mazerolle et al., 2000).

More support for the notion that variations in affective states resulting from strain
predict different types of criminality was found among a population of 150 youthﬁrl

offenders (Piquero & Sealock, 2000). In particular, anger was identiﬁed as a signiﬁcant

62

predictor of personal but not property crimes, while depression was unrelated to either
type. Supporting the contentions of previous research (Broidy & Agnew, 1997;
Mazerolle & Piquero, 1998), the scholars also found that the strength of relationship
between strain and consequent emotional responses varies based on the types of crime
under analysis, and that a link between anger and interpersonal violence and between
depression and self-destructive behaviors merits additional exploration. A similar
conclusion was reached by Aseltine, Gore, and Gordon (2000) when they examined the
effect of life stressors on delinquency through the inclusion of school, work, family, and
ﬁnancial strain variables among high school youths in Boston. Their analysis revealed
general support for GST, but identiﬁed a link only between anger and more serious and
violent forms of criminality. Speciﬁcally, feelings of anger and hostility stemming from
stressful life events appeared to predict aggressive responses more so than nonaggressive
responses or drug use (Aseltine et al., 2000).

The effect of strain on the sexes has also been analyzed by a few researchers in
this area (e.g., Hofﬁnann & Su, 1997; Mazerolle, 1998). Hoffmann and Su (1997) found
similarities across gender for the applicability of GST to delinquency and drug use.
Another study published the same year demonstrated that men were more likely than
women to partake in delinquency following exposure to strainful stimuli (Agnew &
Brezina, 1997). In a conceptual piece, Broidy and Agnew (1997) asserted that gender
inﬂuences the types of strain and the negative affective states experienced, as well as the
resultant methods that individuals employed to cope. Speciﬁcally, they hypothesized that
males are often subjected to ﬁnancial strain - which frequently results in property crime,

and to interpersonal strain - which frequently results in violent crime (Broidy & Agnew,

63

1997:297). Strain experienced by women typically includes disproportionate subjection
to social control and a restriction of opportunities to partake in criminal behavior - which
largely seem to result in self-destructive forms of maladaptive responses like eating
disorders and drug use (Broidy & Agnew, 1997:297). These self-harming behaviors
appear to stem in part ﬁom certain emotions that tend to accompany strain among
women, such as depression, shame, and guilt.

Mixed support for GST was generated by Mazerolle (1998), however, who did
not ﬁnd any difference between the effects of strain predictors on delinquency across
gender categories. Nonetheless, he did ﬁnd that gender differentiated the effect of
negative life experiences on violent crimes, with men more likely to extemalize anger
and women more inclined to internalize such an emotion. These studies jointly highlight
the fact that qualitatively different responses appear to result depending on the emotional
outcome immediately resulting from the strainful experienCe, and point to the importance
of revised conceptualizations of strain when attempting to understand differences in
delinquency among males and females.

Finally, two constructs were found to be signiﬁcantly associated with the
possibility that individuals reacted to strain with delinquency: negative emotionalityand
constraint (Agnew, Brezina, Wright, & Cullen, 2002). The former refers to the proclivity
to interpret events as aversive or malicious, and to respond to them in a hostile or
antisocial manner; the latter concerns self-control, discipline, and delayed gratiﬁcation.
Overall, delinquency was found to be higher for those juveniles who experienced strain in
familial, neighborhood, and school contexts. In accordance with intuition, juveniles high

in negative emotionality and low in constraint had an increased predisposition towards

64

delinquent responses from strain. A wide variety of scales were constructed and utilized
in this analysis, including those measuring family strain, the presence of conﬂict with
parents, whether the parents of the juvenile sometimes lose control and feel they might
hurt their child, feelings of hatred towards school, if the juvenile is picked on by other
kids, and the presence of neighborhood strain. Unquestionably, any or all of these factors
may contribute in some manner towards deviant manifestations.

Certain conclusions based on this literature review can be made. The main tenet
of general strain theory - that a positive relationship exists between strain and
delinquency - has been supported through both cross-sectional and longitudinal research.
Variables in the form of personality, temperament, self-control, self-efﬁcacy, self-esteem,
deviant peer associations, conventional bonds, moral beliefs, and social support systems
have conditioned the effect of the primary predictor on the delinquent outcomes, but not
to a conclusive degree. Additionally, the types of offenses directly or indirectly produced
by strainful circumstances appear to vary, depending on both the content of the strain and
the affective way in which the individual responds (which is often related to gender
differences) (Broidy & Agnew, 1997; Mazerolle, 1998). Further analysis is required to
more accurately tease out the intervening role of emotions, and the way in which the
aforementioned mediators affect the strain/delinquency relationship. It also should be
noted that conceptualizations of strain theory have been integrated with other theoretical
perspectives, such as those in the biological (Walsh, 2000), structural (Agnew, 1999;

Brezina, Piquero, & Mazerolle, 2001), and developmental (Agnew, 1997; 2002) spheres.

65

Genegrl Strgin Theory Applied To The MP3 Phenomenon

As mentioned earlier, anger and frustration are the two primary emotional
outcomes resulting from strain, and the literature has largely explored the path from anger
to crime and delinquency. When considering the subject matter at hand, frustration
seems to be much more relevant as a causal element. Accordingly, its relevance to music
piracy is hereby explained. Individuals who are strained in certain ways may attempt to
cope with the resultant affective state by participating in online intellectual property theft.
On its surface, the relationship between stress-inducing stimuli and this speciﬁc type of
deviant behavior appears to be a stretch. However, Agnew’s ﬁrst type of strain - the
threatened or actual failure to achieve positively valued goals — may be relevant.
Speciﬁcally, strain in the form of ﬁnancial, age, mobility, and parental restrictions on
music, and in the form of a perceived necessity to achieve a certain status level amongst
peers or family members, may cause some to unlawfully obtain and transfer copyrighted
music from online sources. Further elaboration of these points is necessary before 3
proceeding.

Most individuals are not able to purchase the desirable commodity of music CDS
with abandon, simply because of their price. This point is more-pronounced among
children, teenagers, and college students, whose ﬁscal resources are generally minimal to
none due to the fact that they have not yet acquired a well-paying job. The desire to
possess and listen to certain songs and artists, and the inability to purchase them either
because they are not affordable or because they cannot take precedent over bills,
payments, and other imperative destinations for one’s dollar, results in a conﬂict that

must be resolved. To note, though, even those who do make a comparatively large

66

amount of money are still inclined to participate in the wrongdoing, as if the fact that they
could easily afford purchasing the CD is irrelevant”. Seemingly, the appeal of obtaining
something for nothing is too strong to resist for some.

The strained individual can commit larceny by pilfering a coveted music album
from a retail establishment, but runs the risk of detection, apprehension, and punishment.
The advent of MP3 technology, though, provides the conﬂicted person with another
choice that is similarly illegal but socially acceptable, collectively embraced, and difﬁcult
to monitor, curtail, or thwart. As a result, the dissonance stemming from the inclination
to possess and the incapacity or disinclination to pay can be overcome through the
discovery and download of the desired music from P2P ﬁle exchange programs, chat
rooms devoted to the dissemination and exchange of MP3s, bulletin boards and
newsgroups created for the same purpose, web sites, ﬁle servers, and even from others
via instant messaging programs.

Age, mobility, and parental restrictions also may contribute to the strainful
circumstance. Some music albums have explicit content or lyrics and are branded with a
sticker or logo that indicates to sales clerks that purchases must be made by an adult or
with an adult present. The desire among those underage to obtain and listen to this type
of questionable music may induce some amount of strain. Also, children and teenagers
who are not yet able to drive, or who do not have access to a vehicle, may not be able to
venture to stores to purchase certain coveted music, and thus the lack of mobility inherent
in such a scenario may hasten strain. Limitations set by parents on the types of music

that their children may listen to can also lead to strain, particularly if that type of music is

 

2' Please see sonre white-collar crime research exarrrples (e.g., Benson, 1985; Benson & Moore, 1992b;

67

popular and culturally embraced by their children’s peer group. Aspirations for peer
acceptance, social status, and even the Simple desire to possess (or at least be familiar
with) the music of certain genres, artists, or bands - coupled with the inability to do so
because of parental restrictions - may activate strainful feelings”.

The signiﬁcance and acuteness, then, of such uncomfortable stimuli may be
attenuated through the maladaptive response of music piracy. What is essential,
however, is the presence of negative affect stemming from the strain. In these cases, it
would be frustration and aggravation resulting from thwarted ambitions for a desired
product that consequently lead to music piracy.

An analogy might assist in understand this point. Some individuals in ofﬁce
environments congregate around the water cooler to discuss a variety of topics, such as
popular television sitcoms from the previous night. Those who did not watch the shows
being discussed might feel as if they do not ﬁt in because their unfamiliarity with the
subject matter precludes their participation in the dialogue. Indeed, their status level
might be reduced in some capacity in that social group if they continually are present for
the discussions but never actually watch the sitcoms, and strainful feelings might result.
To resolve the dissonance stemming from this predicament and to be able to relate to
coworkers around the water cooler, individuals can choose to familiarize themselves with

the television shows by watching them; this can be generally done at no cost to them.

 

Coleman, 1989; Rosoff, Pontell, & Tillman, 2002)} for evidence towards this end.

22 As mentioned in the literature review, Agnew & White (1992) employed a one-item measure of “clothing
strain,” where individuals indicated whether their parents were able to purchase for them the types of
clothing they desired. It was not, however, signiﬁcantly related to delinquency or drug use in their study.
Nevertheless, this parallels the same type of strain that an inability to purchase socially desirable music
might effectuate.

68

College students aspire for peer acceptance in a Similar way, and may need to
demonstrate familiarity with, and appreciation of, certain music to ﬁt in and relate to their
social group. To counter the strain and negative affective state of frustration that might
result, one solution requires them to purchase CDS and thereby acquaint themselves with
the music that their peers embrace. Unfortunately, the cost of CDS is somewhat
prohibitive, particularly for those in school. Another solution would be to download that
same music quickly and easily from sources on the Internet at no cost to them. Ceteris
paribus, it is patently clear which choice is more appealing in offsetting potential or
actual strain.

Agrrew’s initial presentation of his theory was relatively nascent in its
development, and scholars over the past decade have attempted to augment and reﬁne its
explanatory capacity. The current work takes another step in that direction by applying it
to an Intemet-based crime in the form of music piracy for the purposes of testing its
generality. No attempts have as yet been made to explore computer-based wrongdoing
with this theoretical paradigm. Additionally, most empirical examinations of strain have
utilized middle- or high-schoolers, or nationally representative samples of youths; the
current study employs a population of university students. While detailed hypotheses are
presented below, general strain theory appears relevant to predicting Internet crime by
focusing on the inability to achieve a positively valued goal - the possession of a socially,
culturally, and individually esteemed commodity: commercial music. Individuals may
desire to purchase the creative works of certain artists or bands, but might be unable to do
so because of a lack of ﬁscal resources. Furthermore, certain social pressures may be

present - such as the fact that one’s peer group is participating in the MP3 phenomenon,

69

and it may consequently be important for an individual to partake in music piracy to “ﬁt
in.” The desirability of being perceived as “cool” among one’s ﬁiends and acquaintances
stemming from having a large collection of MP3s, or being well-versed on popular artists
and bands, is another positively valued goal which arguably may incite the behavior.
Self-Control Theory

Self-control theory was articulated in its most developed form by Michael
Gottfredson and Travis Hirschi in their 1990 work The Generﬂheorv of Crime. The
scholars deﬁne crimes as “acts of force or fraud undertaken in the pursuit of self—interest”
(1990:15). In their view, criminal acts generally provide only immediate and short-term
rewards, are easy and simple to enact, are exciting, require little skill or planning, impose
pain on others, and can relieve frustration temporarily (Gottfredson & Hirschi, 1990). By
extension, the argument made is that all types of wrongdoing can be explained by low
self-control and the opportunity structure surrounding the act.

With self-control theory, the line of criminogenic explanation continues to
proceed in the opposite direction of Aquinas’ quixotic belief that man is essentially good
and by committing crime, he actually harms his own humaneness and natural tendency to
abide by the law (Vold, Bernard, & Snipes, 1998). Since the 13th century then, the ﬁrmly
ensconced assumption about human nature is that individuals will take advantage of
others without qualms or misgivings if left to their own devices. Self-control theory
embraces that principle as a foundation for its interpretation. The underlying premise is
as follows: all people are intrinsically motivated to break the rules of society, but

differences exist in people’s innate ability to suppress or restrain urges and drives, and in

70

their needs for excitement, risk taking, and immediate gratiﬁcation (Lanier & Henry,
1 998).

Because of deﬁciencies and weaknesses in their intrinsic personality and
character, then, individuals with low self-control are more likely to engage in crime to
accomplish a goal or to resolve a conﬂict in the most expeditious and effortless manner.
Gottfredson and Hirschi (1990) assert that those persons who demonstrate difﬁculty in
(or apathy toward) accomplishing long-term goals or maintaining long-term relationships,
and those who engage in extreme and decadent activities (such as smoking, drinking, and
promiscuity) are also predisposed towards illegal behavior. Nonetheless, most people do
not break the rules because they have been effectively socialized accordingly by various
institutions. Some, however, have either been inadequately socialized or not socialized at
all, and this lack of constraining values frees them to commit crime (Lanier & Henry,
1998). Gottfredson and Hirschi (1990) state that inadequate parenting during childhood
is a primary reason why some individuals are improperly trained to exhibit self-control.

The theory also incorporates the concepts of stability and versatility (Gottfredson
& Hirschi, 1990:117-9). Stability refers to the fact that because of the relatively
permanent trait of self-control, the role of other inﬂuencing factors later in life is
rendered largely impotent in impelling criminality. By extension, differences in
individual offending should remain generally invariant; those individuals who possess
high self-control will be “substantially less likely at all periods of life to engage in
criminal acts” (Gottfredson & Hirschi, 1990289). To note, it seems that assessing self-
control among college students will presumably measure a characteristic that has, and

will continue to, affect their life decisions in a certain way. According to Gottfredson

71

and Hirschi (1990), if self-control is signiﬁcantly related to the music pirating habits of
respondents, this factor is not dependent on time and its inﬂuence will not vary in the
person’s past or future.

Versatility regards the explanatory power of low self-control predicting all
varieties of criminality, deviance, and even unfortunate occurrences like accidents,
debunking the argument of certain causal factors disproportionately producing different
criminal specialties. Taken to its logical conclusion, both traditional and nontraditional,
blue-collar and white-collar, and real-world and “virtual” crimes are possible by those
who lack a sufﬁcient amount of self-control. These are two of the primary underpinnings
of Gottfredson and Hirschi’s paradigm, and are largely what render it a “general” theory.

While low self-control is a necessary condition to increase the likelihood of
committing a crime, Gottfredson and Hirschi afﬁrm that it is not a sufficient condition.
Opportunity plays a crucial role, and the scholars draw on Routine Activities Theory
(RAT) to speak to its relevance. Proposed by Alfred Cohen and Marcus F elson (1979),
RAT attests that the conditions necessary for a crime to occur include a suitable target,
the absence of a capable guardian for the target, and the presence of a motivated offender.
Indeed, this is a red herring because all theories of criminality require the existence of an
opportunity before the crime occur (Cochran, Wood, Sellers, & Chamlin, 1998;
Grasmick, Tittle, Bursik, & Ameklev, 1993).

Gottfredson and Hirschi’s chief concern involves the role of guardians and
targets; the motivation of offenders is regarded as nonproblematic (Sellers, 1999). In its
very simplest terms, then, it is the degree of availability to a target that produces the

opportunity to engage in deviance. Pratt and Cullen (2000:933) echo similar sentiments

72

by stating that “although people vary in levels of self-control, the world is ﬁlled with
criminal opportunities; after all, crime is easy to commit and requires little planning.”
Due to the tremendous (and almost ubiquitous) availability of digital music ﬁles online
and the presence of a population potentially disposed towards exhibiting low self-control
in a college environment, this theoretical perspective appears highly salient to the subject
matter at hand.

Self-control is considered a nucleus of sorts around which every other known
factor associated with the crime can be conﬁgured, and is perhaps best understood as an
underlying construct which integrates a variety of conceptions about crime (Akers, 1991;
Pratt & Cullen, 2000). Gottfredson and Hirschi do not operationally deﬁne “self-control,”
and it is therefore difﬁcult to measure its causal inﬂuence on crime. As such, and as
identiﬁed by other scholars (e.g., Akers, 1991), the identiﬁcation of the former can be
made by identifying participation in the latter. Research subsequent to the theory’s initial
assertion has therefore sought to indirectly assess low self-control to determine its
underlying predictive inﬂuence on deviant and criminal behavior. Furthermore, as long
as self-control and crime are measured using independent items, any arguments of
tautology are preempted (Pratt & Cullen, 2000). Gottfredson and Hirschi (1990) also
recommended testing the theory through the proxy of participation in “analogous”
behaviors which demonstrate low self-control; this has been done in much of the relevant
research. To note, the current study measures self-control through separate attitudinal

and behavioral operationalizations; this is discussed in detail later in the text.

73

Empirical Support for Self-Control Theory

There have been a multitude of tests examining the linkage between low self-
control and criminal acts, operationalizing its six delimited facets: impulsivity, simple
tasks, risk seeking, physical activities, self-centeredness, and temper (Gottfredson &
Hirschi, 1990:89). Many of these probes have found support for the general theory.
Moreover, the promulgation of the theory has been applied to a variety of crimes, ranging
from imprudent behaviors (e. g., Ameklev, Grasmick, Tittle, & Bursik, 1993) to general
law violations (e.g., Piquero & Tibbetts, 1996). The relationship between self-control
and white-collar crime has even been examined, but little to no support was found (e. g.,
Benson & Moore, 1992a; Steffensmeier, 1989). To note, most research has not
empirically examined how poor parenting is a causal predictor of low self-control.
Instead, studies have disproportionately concentrated on the relationship between self-
control and deviant and illegal behaviors; the current work is no exception.

Grasmick et a1. (1993) were the ﬁrst to test an unidimensional operationalization
of the six components of self-control, and their work resulted in the creation of a Single
factor 24-item scale with four items for each component”. That is, the items comprising
the six dimensions articulated by Gottfredson and Hirschi demonstrated enough variance
in common to be used as a singular scale assessing self-control. In their analyses, three
predictions were tested. First, the interaction between self-control and opportunity

should be positively and signiﬁcantly related to force and fraud; this was supported with

74

standardized coefﬁcients of .156 for force and .235 for fraud. Second, since Gottfredson
and Hirschi assert the necessity of opportunity in inducing those with low self-control to
commit crime, the interaction term should be larger than the singular predictive effect of
self-control. This was corroborated in part, as the interaction term was signiﬁcant and the
main effect of low self-control was not when measured against acts of force. Both were,
however, when measured against acts of fraud. The third prediction was that opportunity
should not be singularly related to either force or fraud beyond its interactive effect with
low self-control. Contrary to that hypothesis, the relationship between opportunity and
the two general types of crime was signiﬁcant, introducing some ambiguity into the
causal chain“.

Gibbs and Giever (1995) analyzed the independent effect of self-control on
criminal equivalent behaviors among a sample of 236 undergraduate students, and
actually characterized such individuals as a group marked by high self-control because
college enrollment requires some amount of academic success. Indeed, they state that
they “would expect to ﬁnd very few wholly unrestrained individuals in a group of
university students” (1995:243). According to the researchers, higher levels of self-
control also lend itself to greater levels of participation in the study and in more valid

measurements based on the responses retrieved. Dependent variables included class

 

23 In their original analysis, Grasmick et al.'s ﬁnal scale had 23 items due to the removal of a Physical
Activity measure (which increased the resultant alpha from .805 to .812) that most aptly and reliably
measured the nature of the characteristic as presented by Gottﬁ'edson and Hirschi (1990). When factor
analyzing the variables, however, a ﬁve-factor solution was identiﬁed; nonetheless, the resultant scree plots
revealed a signiﬁcant drop-off between the ﬁrst and second factors, and so a one-factor model was forced .
( 1993: 1 6).

2’ Incidentally, Grasnrick et al. (1993:25) also suggest incorporating variables that affect individual
motivation, such as those related to strain, to explain more variance and live up to the billing it was given
by its originators. This highlights the possibility in future research to incorporate key individual
components of each general theory to best predict the phenomenon of online intellectual property theft.

75

cutting and levels of drinking alcohol—both technically noncriminal but appropriate for
testing Gottfredson and Hirschi’s theory since they demonstrate low self-control and
share characteristics with actual crime (1995:250). Via both OLS linear and logistic
regression, self-control was signiﬁcantly related to the outcome measures; the R2 ﬁndings
were relatively low (.139 for class cutting and .230 for alcohol consumption) and
suggested the existence of other explanatory elements not included in the models.

A large number of studies have used the Grasmick et a1. (1993) scale both
partially and fully to assess the relevance of self-control on a variety of deviant and
criminal behaviors (Ameklev et al., 1993; Longshore & Turner, 1998; Longshore,
Turner, & Stein, 1996; Piquero & Rosay, 1998; Piquero & Tibbetts, 1996; Wood,
Pfefferbaum, & Ameklev, 1993). Ameklev et a1. (1993) found a link between low self-
control and “imprudent” behaviors such as excessive drinking or a gambling predilection,
and Wood et al. (1993) identiﬁed a strong relationship between self-control and theft,
vandalism, certain forms of violence, and drug use. It is interesting to note that of the six
elements conceptualized by Gottﬁedson and Hirschi and operationalized by Grasmick et
a1. (1993), physicality has been consistently identiﬁed as a very weak predictor (Ameklev
et al., 1993; Cochran et al., 1998; Grasmick et al., 1993; Wood et al., 1993). On the
contrary, the risk-taking and impulsiveness elements of the theory have been consistently
identiﬁed as strong predictors (see e. g., Brownﬁeld & Sorenson, 1993; Keane, 1993;
Wood et al., 1993).

The Grasmick et a1. (1993) scale was also utilized by Piquero and Tibbetts (1996)
to test the applicability of self-control on deviance by 642 university students between the

ages of 17 and 35. They found that four and ﬁve percent in drunk driving and shoplifting

76

respectively could be explained by the direct inﬂuence of self-control, as well as its
indirect inﬂuence through situational factors such as perceptions of pleasure and shame.
When employing 19 out of the original 24 items in the scale, Piquero and Rosay (1998)
found that self-control explained 7% and 13% of the variance in fraud and force measures
when controlling for general demographic characteristics. Indeed, when considering the
utility of the scale, they asserted that it is acceptable for “tapping into the components
alluded to by Gottfredson and Hirschi” (1998:170). Additional works have differentiated
between property and personal crimes to more accurately assess the predictive capacity of
the theory. Longshore, Turner, and Stein (1996) speciﬁcally tested the scale on a sample
of offenders with a history of drug use, and found a weak but theoretically expected
relationship between self-control and both acts of force and ﬁaud. In a study on drug
users’ proclivity towards property and personal crimes, Longshore (1998) determined
that low self-control and high opportunity - as well as the interaction of the two - was
signiﬁcantly related to the outcome variables. The amount of explained variation was
quite modest (4%), but did lend credence to the theory’s main precepts.

Additional strong corrOboration for Gottfredson and Hirschi’s theory has been
found by: Evans et al. (1997), who assessed its predictive capacity on 17 personal and
property crimes and 18 other forms of deviance; Gibbs et al. (1998), who demonstrated
that low self-control among college students is signiﬁcantly related to cheating, drinking,
suspension or expulsion, and Skipping class; and Burton et a1. (1998), who keyed out the
theory’s relevance to a variety of wrongdoing including the ﬁling of false insurance

claims, workplace theft, interpersonal violence, drug use, and automobile accidents.

77

One project related to interpersonal aggression allude to points tangentially but
notably related to the current work. Christine Sellers (1999) retrieved some interesting
results when studying self-control and domestic violence by gathering and analyzing data
from a subset of 985 students at a college in Florida who were then involved in a dating
relationship. Although the statistically signiﬁcant results were relatively weak when
isolating the element of self-control, it could be concluded that low self-control plays at
least some role in predicting the probability of using violence against a dating partner
(Sellers, 1999). Coupled with other factors such as opportunity and the perception of
reward, the explanation gained slightly more strength (Sellers, 1999). From a broad
perspective, the use of physical aggression appeared to offer short-term gratiﬁcation to
the offender, both in terms of the derivation of pleasure - perhaps through enhanced
arousal and a feeling of excitement or thrill, and the reduction of pain - perhaps through
relieving ﬁ'ustration or ending an argument (Sellers, 1999). The researcher also
maintained that the most common form of courtship aggression involved physical actions
that require little effort and no planning which can take place at any time, such as
slapping or shoving (Sellers, 1999). Consonant with ﬁndings ﬁom previously mentioned
studies, this underscores the salience of spontaneity and Speedy gratiﬁcation in
effectuating wrongdoing.

Finally, Pratt and Cullen’s (2000) meta-analyses of 21 studies identiﬁed
impressive support for self-control theory, consistently ﬁnding an effect-size estimate -
the standard correlation coefﬁcient (r) - over .20 for the construct of self-control. These
results remained even when including variables measuring opportunity and the elements

of other criminological paradigms, and despite different operationalizations of the

78

construct by various scholars. It is also worthy of mention that the researchers found that
social learning theory variables, when included in studies of self-control, increased the
explained variation of deviant or criminal behavior — highlighting the validity and
importance of each as predictors of criminality.

Interestingly, Pratt and Cullen’s (2000) meta-analysis identiﬁed 82 attitudinal
measures and 12 behavioral measures of self-control. Attitudinal measures include those
in the Grasmick et al. (1993) scale, while behavioral measures consist of instances of
analogous behaviors. Behavioral measures, in fact, were recommended by Hirschi and
Gottfredson (1993) over attitudinal items, and this assertion has attracted criticism for the
tautology inherent in utilizing measures of deviance as a predictor of deviance. Pratt and
Cullen (2000) found that while behavioral measures had a slightly larger effect size, they
were similar in magnitude as their attitudinal counterparts - demonstrating that employing
one measure over the other will not signiﬁcantly affect the predictive capacity of self-
control, and testifying to the robustness in operationalizing the theoretical construct in
multiple ways. With this in mind, the choice to include both attitudinal and behavioral
measures of self-control was made to provide a more nuanced perspective as to the role
of that dispositional trait on music piracy.

The current study is cross-sectional, and does not purport to offer inconclusive
evidence concerning the causal relationship between the theories and the crime.
Nonetheless, Gottfredson and Hirschi (1990) argue that cross-sectional research of self-
control is not inherently subordinate in quality, utility, or rigor to longitudinal research,

and may in fact be more beneﬁcial in some instances. To note, a meta-analysis by Pratt

79

and Cullen (2000) found that the explanatory strength of low self-control is weaker in
longitudinal research conducted to test the applicability of the theory.

To summarize, Gottfredson and Hirschi’s general theory consists of two primary
components: self-control and criminal opportunity. Individuals, then, are more apt to
engage in wrongdoing if they have low self-control - a stable dispositional trait on an
individual level. Furthermore, they can be characterized as impulsive, insensitive, short-
sighted, and risk-taking (Gottfredson & Hirschi, 1990:90). These persons are also
inclined to partake in “analogous” behaviors such as smoking, alcohol and illegal
substance use and abuse, speeding, truancy, and even an increased proclivity towards
accidents and illness (Gottfredson & Hirschi, 1990; Junger & Tremblay, 1999;
Paternoster & Brame, 1998, 2000). By pointing out fallacies of traditional positivistic
thought and by building and expanding upon classical (or control) theory, the authors
promulgated a new criminological paradigm with signiﬁcant explanatory potential.
Self-Control Theory Applied To The MP3 Phenomenon

Individuals with low self-control, when presented with the opportunity to obtain
high-ﬁdelity commercially-produced songs over the Internet through a few “point-and-
click” maneuvers of their mouse, will not be able to bridle their inclinations and will
download such ﬁles and thereby commit a criminal offense. College students are a
population of individuals who have the opportunity to access the Internet either through a
personally-owned computer or a university-owned machine, install P2P ﬁle sharing
programs, and participate in unregulated data transfers (often over high speed
connections) with either no cost on campus or at an affordable cost at an off-campus

location (such as their home). Indeed, with the continued reduction in computer and

80

connectivity prices, the development and propagation of advanced physical media (such
as ﬁber—optic lines) for data transfers, new technology involving ﬁle compression,
smaller packet sizes, and quicker routing, the opportunity for those interested to have
access to, and use, the Internet will continue to grow. Sufﬁce it to say that the
opportunity is there, and will be in increasing fashion.

The differentiating variable, then, is self-control. Reﬂexive responses to
immediate stimuli (such as the availability of MP3 ﬁles), rather than careful reasoning as
to the acceptability, wisdom, and ramiﬁcations of certain actions, would seemingly occur
amongst those with underdeveloped amounts of self-control. Indeed, typical college
students are already arguably at an age where self-control is not foremost on their mind,
particularly if they have recently left the “nest” and are living outside of the regular
supervision of parental authorities for the ﬁrst time. Persons at this age also tend to
experiment a great deal, perhaps in search of themselves or their own identity. Internal
pressure to participate in a host of questionable activities which previously would have
been impossible or unacceptable are now more plausible, appealing, and even desirable.
Accordingly - and particularly among undergraduate students - low self-control is a
characteristic which might be more frequently found than among children or full-ﬂedged
adults.

Is music piracy characterized by the distinguishing features that Gottfredson and
Hirschi (1990) contend as related to crime, such as the precursor to short-tenn
gratiﬁcations of excitement, monetary gain, and relief from situations that induce
aggravation? Seemingly, participants may get excited when locating and obtaining an

MP3 ﬁle of a song they have wanted to hear. This excitement may be augmented when

81

they realize that no cost is incurred when downloading the music ﬁle to their computer
system for unlimited playback. More excitement may result from dissemination of that
ﬁle to ﬁ'iends or family, as the meaning, importance, or relevance of the song is shared
through its collective experience. Excitement again comes into play when the individual
realizes that vast amounts of high-quality and easily accessible and procurable music by
practically any musician or band from practically any time period is available.

Monetary gain is relevant primarily through the realization that no expenditure of
funds is necessary to receive and enjoy this commodity, or to pass it along to others. In
fact, nothing physical or tangible is required in the acquisition and distribution of these
ﬁles apart from a computer system and an Internet connection. Most - if not all - of the
software necessary for involvement in the MP3 phenomenon is freely available online.
These factors widen the net of potential pirates because no purchases must be made
beforehand to facilitate the activity. With golf, for example, golf clubs, balls, and a bag
need to be purchased prior to participation. The skills and knowledge essential to
competence on the golf course also often necessitate the purchase of lessons. These
requisite expenses serve as a funnel to reduce the amount of golfers in the general
populace. With music piracy, if an assumption is made that individuals have a computer
and an Internet connection as an arguable “necessity” for personal and professional
reasons (particularly among college students), no other equipment is needed. Lessons to
participate are also absolutely superﬂuous because of the simplicity of software
applications that assist interested users.

Finally, downloading (and transferring) MP3$ may provide relief from

aggravating Situations if one considers that individuals may have a desire to enjoy certain

82

songs or albums but lack the funds to legitimately purchase them from a retail outlet.
Perhaps a speciﬁc music ﬁle is (to the initiating downloader) useful in some way for
admiration or esteem among friends, or for use in a school project, or to send to a loved
one. An inability to accomplish any of those may exasperate the individual. Perhaps that
person has also sought to purchase the music ﬁle on CD from legal sources, but has been
unsuccessﬁrl in all attempts to locate it.

These examples provide support for the possibility that copyright infringement
through the acquisition of MP3s may relieve aggravation stemming from situational
factors. To note, this aggravation as mentioned by Gottfredson and Hirschi points to the
role of strain in effectuating criminality or deviance and alludes to some overlap between
the theoretical paradigms. Situational factors increasing strainful feelings or reducing
self-control, though, are not analyzed in the current analyses. Sufﬁce it to say that
reﬂexively responding to immediate stimuli, rather than careful reasoning as to the
acceptability, wisdom, and ramiﬁcations of the response, appears to be a Viable reason
why music piracy occurs.

Gottfredson and Hirschi (1990:168) have argued that self-control and social
circumstances do not interact to induce criminality because those with low self-control .
participate unequally in social institutions and relationships and actually avoid attachment
to others due to the underlying dispositional trait. As such, low self-control - their sole
predictor of participation in crime - is not and cannot be learned in a social setting
(Gottfredson & Hirschi, 1990:95). Interestingly, Evans et a1. (1997) have suggested that

Gottfredson and Hirschi overstated their case when asserting that social learning does not

83

play a contributive role in criminality. These researchers studied 555 individuals25
through self-report surveys and included two social learning measures related to
differential association and deﬁnitions favorable to law Violation. The former was
operationalized by a question concerning the respondent’s number of criminal friends,
and the latter by “statements concerning the degree of tolerance for criminal behavior,
moral validity of breaking the law, and level of agreement with committing criminal acts”
(Evans et al., 1997:487). They reached the conclusion that criminal associations and
criminal values may augment the inﬂuence of low self-control on deviant outcomes,
either through introduction to the activity, pressure to partake in it, modeling of the
behavior, or redeﬁning it in an acceptable light and as a pleasurable endeavor worth the
risks of apprehension and punishment (Evans et al., 1997)“.

Concerning efforts towards theoretical integration, the incorporation of strain and
social learning variables with self-control in an explanatory model has been suggested
(e. g., Grasmick et al., 1993; Mazerolle & Maahs, 2000; Wright, Caspi, Mofﬁtt, & Silva,
2001). For instance, Wright et a1. (2001) sought to determine how certain variables
measuring self-control, social control, and differential association were related to crime.

They discovered that low-self control is positively related to delinquent peer association,

and that self-control and social control were both independently and interactively related

 

25 The analysis was limited to Whites because their response rate was disproportionately higher than other
demographic groups.

2° Other important research involving delinquent peer associations and delinquent value systems support
this ﬁnding (e.g., Heimer, 1997; Matsueda & Anderson, 1998; Warr & Stafford, 1991). Thus, a
relationship between low self-control, social learning, opportunity, and crime has been substantiated. In
another piece, (Agnew et al., 2002), evidence was discovered linking the personality trait of “low
constraint” to crinrinality as a mediating variable between strain and delinquency. This speaks to a
demonstrable association between strain, low self-control, and crime. A juxtaposition of the three general
theories, then, is almost demanded - and an integrative approach subsequent to the current research
endeavor may hold much promise in predicting the most variance in the dependent variable under study.

84

to crime. Mazerolle and Maahs (2000) identiﬁed that variables measuring social learning
theory and self-control theory, when included with GST variables, affect delinquency
independently in both a cross—sectional and longitudinal model. This appears to cry out
for an integrationist approach to collectively account for each theory’s relevant precepts.
Nonetheless, Gottfredson and Hirschi (1990) have gone out of their way to call into
question the predictive capacity of other theories. As such, they have isolated and
aggrandized the role of self-control and have conceptualized and articulated it in such a
way as to render integration with other theoretical components largely impossible.
Social Learning Theory

Reﬁned and developed during the course of subsequent years, social learning
theory was initially proffered as a guiding theory in 1977 by Ronald Akers, and was
based in part upon his research with Robert Burgess (1966) and the earlier works of such
scholars as Edwin Sutherland (1947; 1949a; 1949b), Albert Bandura (1969; 1973; 1977;
1963), Gabriel Tarde ([1890] 1903), and BF. Skinner (1953; 1971). As a general theory
of crime, it seems intuitively applicable to new forms of criminal behavior stemming
from technological advances, and in fact has been suggested and utilized for the study of
nontraditional crimes (Akers, Krohn, Lanza-Kaduce, & Radosevich, 1979; Rogers, 2001;
Skinner & Frearn, 1997). Seemingly, unethical or unlawful behavior involving a
computer and the Internet requires the presence of at least some of the principles of social
learning theory to occur; it is not an action that just anyone can do without learning
certain techniques and mentalities.

Social learning theory is an amalgamation of four singular theoretical tenets into a

cohesive whole. An explanation of each is required before proffering its application to

85

MP3 participation. They include: differential association, imitation and modeling,
deﬁnitions, and differential reinforcement. Differential association occurs as social
interaction with family, friends, and acquaintances provide and strengthen normative
deﬁnitions of acceptable and unacceptable conduct. In this environment, motives, drives,
rationalizations, and methods for behaving in certain ways are learned and internalized.
This ﬁrst facet of social learning is based on Sutherland’s (1947; 1949a; 1949b)
differential association theory, which holds that an individual who associates more with
supporters of criminal patterns of behavior (irrespective of whether they are actual
offenders) than those with anti-criminal patterns of behavior will be more likely to violate
the law. Second, behavior is also learned through imitation and modeling of the actions
of others during the socialization process. Individuals already immersed in the deviant
activity provide a palpable exemplar to emulate, thereby transmitting knowledge,
attitudes, beliefs, and techniques that Si gniﬁcantly inﬂuence a potential criminal’s
participation in the wrongdoing.

The third component of social learning theory refers to deﬁnitions, which are
evaluative criteria designating certain behaviors as good or bad, and thus qualifying them
as appropriate, desired, or justiﬁed (Akers et al., 1979). These also are learned from
social interaction, and are instrumental in determining commencement of, or abstention
from, a certain activity. Deﬁnitions roughly parallel the neutralization techniques
proposed by Sykes and Matza (1957). However, rather than being created and utilized by
a singular individual to free himself or herself from the constraints of law and normative
standards, they stem instead from social interaction and are shared by a group. They are

meanings that deﬁne an action as wrong or right — “discriminative stimuli” (Akers, 198 5)

86

- which serve as cues to participate in or refrain from the behavior. This is analogous to
what people have called the most important principle of Sutherland’s (1947; 1949a;
1949b) theory - that crime will result from an excess of “deﬁnitions” favorable to crime.
Accordingly, the fourth and ﬁnal principle comes into play after deciding to

partake in or refrain from a particular path of action. Behavior is now Shaped by the
consequences that result ﬁ‘om it — also known as operant conditioning (Skinner, 1953) —
and this notion is subsumed under the construct of differential reinforcement. Positive
reinforcement results when a beneﬁcial outcome is produced by the action, thus
strengthening the behavior. Negative reinforcement occurs when behavior is
strengthened or continued through the avoidance of pain. Conversely, positive
punishment ensues when negative stimuli following a behavior serve to weaken it, and
negative punishment takes place when a beneﬁcial outcome is denied after an action, also
weakening the behavior (Akers et al., 1979:638; Skinner, 1957); . The continuance or
cessation of behavior, whether lawful or illicit, stems from such a conditioning process
due to social and nonsocial inﬂuencers. Akers writes:

“Progression into more frequent or sustained use and into

abuse is also determined by the extent to which a given .

pattern is sustained by the combination of the reinforcing

effects of the substance with social reinforcement, exposure

to models, deﬁnitions through association with using [or

participating] peers, and by the degree to which it is not

deterred through bad effects of the substance [or behavior]

87

and/or the negative sanctions from peers, parents, and the
law.” (Akers et al., 1979:638).

Thus, it is important to understand that social forces expose the individual to
prescriptive and communally esteemed conduct, and provide or teach cognitive
restructuring techniques to assuage or render irrelevant any pressing misgivings. They
also provide palpable models to emulate, and train the person or persons in the
identiﬁcation of positive and negative outcomes that may result (Akers, 1996).

Akers has also proposed a sequence by which criminal behavior is learned and
then manifested. Differential association with others who hold deﬁnitions favorable to
the crime ﬁrst occur. This social group then models criminal behavior, which is
consequently imitated as activity-supporting deﬁnitions are shared and adopted. The
continuance or desistance of these imitated behaviors is based on reinforcements — both
social and nonsocial (Akers et al., 1979). To note, since individuals are exposed to
culturally transmitted deﬁnitions, and also are provided with models to imitate and
reinforcement to stimulate either prosocial or antisocial behavior, they can behave in a
conforming or conﬂicting manner to a culture, subculture, or even their own personal
value system (Akers, 1994). In sum, once a social environment is created consisting of
associations with persons inclined to criminality, patterns of imitation and the
internalization of deﬁnitions can then follow, with reinforcing stimuli later playing a
large role in determining perpetuation. Akers further states that the theory links
individual and social processes, as structural conditions inﬂuence a person’s differential
associations, models of behavior, deﬁnitions conducive or aversive to crime commission,

and differential reinforcements (Akers, 1992, 1998).

88

Empirical Support for Social Learning Theory

Akers has emphasized that the social leaming paradigm is empirically testable
through the operationalization of the four particular constructs (Akers et al., 1979). To
corroborate this statement, he reviewed a large body of research in 1994 and asserted
that:

“. . .alrnost all research on social learning theory has found

strong relationships in the theoretically expected

direction. . ..When social learning theory is tested against

other theories using the same data collected from the same

samples, it is usually found to account for more variance in

the dependent variables than the theories with which it is

being compared.” (Akers, 1994)
Many other studies in criminology and sociology have been conducted to assess the
relevance of the theory. It is useful, therefore, to familiarize oneself with the most
important pieces and ﬁndings prior to applying its precepts to a heretofore unanalyzed
phenomenon.

In one examination in 1979 in which Akers was a primary investigator, the
applicability of social learning theory to adolescent drug and alcohol use was tested via a
self-report questionnaire given to 3,065 Midwestern students in grades 7 through 12. It
was found that the combined components of the social learning framework accounted for
68% of the variance in marijuana use and 55% of the variance in alcohol use (Akers et
al., 1979). Cigarette smoking and substance abuse among adolescents has also been

longitudinally analyzed using the principles of social learning, generally showing support

89

in predicting usage over time (Akers & Cochran, 1985; Akers & Lee, 1996; Catalano,
Kosterrnan, & Hawkins, 1996; Krohn, Skillet, Massey, & Akers, 1985). Additionally,
the application of a sensation-seeking component to social learning has been explored
recently on substance use (Wood, Cochran, Pfefferbaum, & Ameklev, 1995). It was
found that positive reinforcement is fostered through intrinsic rewards (such as obtaining
a rush and immediate gratiﬁcation) associated with participating in a marginal activity.
An initial speculation would be that Similar intangible beneﬁts also result ﬁ'om unlawful
computer behavior, such as MP3 participation.

Outcome variables utilized in the testing of this theory have also included
perceptions of the appropriateness of deviant behaviors ranging from cheating to suicide.
Lersch (1999), for instance, surveyed over 500 undergraduate students to empirically
validate the inﬂuencing power of social learning on academic dishonesty. Of the four
tenets of the theory, the most signiﬁcant predictors of deviant behavior were acceptance
of favorable deﬁnitions toward cheating and level of immersion in a peer group which
endorsed the activity. Agnew (1998) discovered that the theory was applicable in the
approval of suicide, as those who had been exposed to ideas or thought processes
conducive to suicide were more likely to perceive it as an acceptable (or at least
justiﬁable way) to deal with one’s problems.

Relevant to the instigating role of an individual’s peer group, Adams (1996)
focused on the role of labeling in effectuating delinquency through social interaction with
others engaged in (or in support of) the deviance. His ﬁndings afﬁrmed the viability of
social learning as a general theory, and also corroborated the signiﬁcance of differential

association on participation in criminal activity. Similarly, Brownﬁeld and Thompson

90

(1991) found that delinquency was dependent on associating with ﬁiends who were
delinquent, a ﬁnding consistent with prior research (Akers et al., 1979; Orcutt, 1987;
Winfree, Backstrom, & Mays, 1994). The theoretical perspective has also been used to
examine participation in gangs. For instance, gang afﬁliation was strongest among those
9th grade youths in a self-report study who were differentially associated with gangs and
had learned or assumed gang-related attitudes (Winfree et al., 1994). In another study
involving incarcerated delinquents, the tenets of differential association and deﬁnitions
were proven signiﬁcant in determining gang membership. That is, gang members
differed from others in the amount of gang-related attitudes they had acquired and in their
proclivity toward gang activities (Winfree et al., 1994).

Sutherland’s primary tenet of differential association — disproportionate exposure
to law-violating peers - has been successfully operationalized in previous studies (e. g.,
Matsueda, 1982). Indeed, there is no paucity of research emphasizing the signiﬁcant
relationship between one’s association with a delinquent subgroup and one’s participation
in delinquent and criminal activities (see, e. g., Adams, 1996; Hawkins et al., 1998;
Orcutt, 1987 ; Warr, 2002). This occurs through training the individual in proﬁciencies to
comrrrit the act, providing extrinsic and developing intrinsic rewards such as gratiﬁcation
and respect among the peer group, and by decreasing the constraining or inhibiting power
of cultural norms, societal dictates, and personal bonds toward normative, lawful conduct
(Kaplan, Johnson, & Bailey, 1987). These aspects of differential association appear
highly relevant to music piracy, and their possible applicability is detailed below.

Social learning theory appears to hold much value for the explanation of computer

crimes, and has been applied to them in a few studies. One research piece published in

91

1997 utilized a self-report questionnaire to assess the inﬂuence of differential association,
imitation and modeling, deﬁnitions, and differential reinforcement on the incidence of
computer criminality among 581 undergraduate students at a major southern university.
Five types of high-tech deviance were measured: whether the respondent “(1) knowingly
used, made, or gave to another person a “pirated” copy of commercially sold computer
software; (2) tried to guess another’s password to get into his or her computer account or
ﬁles; (3) accessed another’s computer account or ﬁles without his or her knowledge or
permission just to look at the information or ﬁles; (4) added, deleted, changed, or printed
any information in another’s computer ﬁles without the owner’s knowledge or
permission; and (5) wrote or used a program that would destroy someone’s computerized
data (e. g., a virus, logic bomb, or trojan horse)” (Skinner & F ream, 1997).

With the inclusion of all relevant variables in the regression model, 37% of
the variance was explained by software piracy while password guessing accounted for an
additional 20%. Furthermore, 16% of the variance was explained by unauthorized access
to browse another’s ﬁles, and 40% was explained by a computer crime index composed
of the sum of respondents’ frequencies in engaging in the ﬁve aforementioned types of
deviance (Skinner & Fream, 1997). Gaining access to change ﬁles and writing a
destructive program were not included in the regression analysis because of extremely
small case numbers. Differential association and deﬁnitions were found to be strongly
and consistently inﬂuential on the outcome variables, imitation varied somewhat in its
strength depending on the computer crime variable utilized, and differential
reinforcement in the form of perceived certainty of apprehension was not related to

wrongdoing. Generally, though, the predictive value of three of four social learning

92

theory variables was demonstrated by the ﬁndings, supporting its use as a guiding
framework.

Rogers (2001) further explored the relationship between social learning theory
and computer crime in his doctoral dissertation. In accordance with his hypothesis,
computer criminals were signiﬁcantly differentiated from non-criminals on the basis of
their associations, but the strength of relationship was quite moderate (112 = .11). Also,
the individual theoretical tenet of differential reinforcement was found to be a signiﬁcant
predictor of computer criminals, but with an even smaller level of variation explained
(112:.04). In this particular analysis, the sample size totaled 132 respondents.

Social learning theory holds much merit because it expands on differential
association and incorporates other chief propositions concerning the acquisition of
criminogenic tendencies. In this context of interpersonal interaction, motives, drives,
rationalizations, and methods for behaving in certain ways are learned, internalized, and
sustained. It is this author’s contention that the theory can be logically extended to online
intellectual property theﬁ. It is hypothesized that the primary method. in which
individuals are introduced to, and become involved in, digital music piracy is through
social learning, whether online in cyberspace via asynchronous (message board,
discussion group, web page, email) or synchronous (chat channels, instant messaging
programs) communication, or in real space Via face-to-face interpersonal interaction. In
the following section, each of the four components of social learning theory is

Speciﬁcally applied to music piracy.

93

Social Learning Theog Applied To The MP3 Phenomenon

Differential Association

MP3 propagation and exchange on the Internet is a wildly social event. A great
number of music aﬁcionados congregate in cyberspace for the distinct purpose of
obtaining and exchanging digital music. Venues which house such collectives include
P2P environments, chat rooms, message boards, and other web communities. The
particular behavior of uploading and downloading digital music is validated and
reinforced by the sheer number of participants and the casual way in which requests for
songs (and advertisements of the availability of other users’ collections) are asserted. Via
a simple illustration, an understanding can be obtained of how MP3 participation is
effectuated by the ﬁrst component of Social Learning Theory — differential association.
If an individual visits a chat channel and someone mentions that he is looking for an MP3
of Madonna’s latest hit single, those not knowledgeable of what an MP3 is are likely to
inquire, and will conceivably be inundated with words of accolade referencing this new
technology that allows for the distribution of near-CD quality music ﬁles over the
Internet, preempting the need to purchase CDS to listen to favorite artists and bands.
Techniques to obtain this type of digital audio ﬁles might then be taught to the “newbie,” -
or person unfamiliar with the technology, as well as motives (free, high ﬁdelity music in
replace of CD purchasing has a tendency to impel many to participate), drives (one heard
with increasing frequency is the necessity to “get back” at the monopolistic recording
industry that has exploited music consumers), and rationalizations (such as how
“everyone is doing it,” or how music must be “free,” or how the chances for getting

caught for distributing copyrighted works is slim to none).

94

When MP3 became a “buzzword” among the technologically-inclined and the
general public, the phenomenon took on a life of its own. The media and popular culture
quickly noticed its sensationalistic quality and ushered it into the public eye. Thousands
of articles have been written about it, hundreds of news sources have covered it, and
individuals of various demographic groups have actively embraced it. Not only has this
attention served to augment the notoriety of MP3 and introduce a greater number of
individuals to take advantage of its characteristics, it also has expanded the scope of
evaluative criteria rendering it favorable. Even if the press makes reference to copyright
law or the recording industry’s hostility towards software that facilitates piracy, the
amount of coverage given to the phenomenon and the statistics proffered to depict its
overwhelming popularity (such as the number of users of P2P ﬁle exchange programs, or
the estimated amount of record sales lost by the music industry) testify to its panoptic
reach. In fact, such attention may even subtly imply that those who are not yet riding, or
who have not yet caught on to, the wave of digital music are seriously missing out on
something special.

The media is speciﬁcally mentioned as a role player in the social learning process
(Akers, 1998). Individuals might interpret from the veritable onslaught of information
they receive about MP3s that the positives it generates substantially outweigh any
perceived or real negatives resulting from participation. It may be that persons partake in
a cost-beneﬁt analysis and come to the conclusion that the questionable behavior is
desirable and even essential for fear of “missing out.” Subsequently, obtaining and
trading MP3s with other individuals may engender positive reinforcement (both tangible

and intangible) which - coupled with the lack of any viable threat of repercussion for

95

downloading or distributing such ﬁles — might more profoundly ensconce the individual
in the practice. This irrrrnersion then can establish the person as an experienced digital
audiophile who is able to pass on techniques, rationalizations, deﬁnitions, models of
behavior to imitate, and reinforcing stimuli to introduce and inculcate others into the
scene. Upon internalization of this learning process by the next generation, the cycle can
continue. Thus, a powerful social system to support the existence and perpetuation of
MPBS can come into existence, facilitating their propagation across the Internet and the
growth of the population who participates.

It should be noted that Akers explicitly differentiates between the differential
association construct and the idea of “peer pressure.” The former is subtle and has a
tendency to shape an individual’s behavior without his or her awareness, while the latter
is couched in overt practices by others to induce the commission of a desired behavior by
a person (Akers, 1998). In cyberspace, differential association might not be as salient
because an individual could simply leave a particular chat channel or environment in
which MP3s are being exchanged if the perceived or actual actions of others do not align
with certain personal standards or mores of conduct. While it is true that a person can
experience ostracism online, the consequences are ephemeral, and can be countered
through the use of a new screen name or user II), which effectively provides a brand new
identity and persona to its claimant. At the same time, it might be argued that differential
association is more pronounced than in real space because a great number of physical,
social, and contextual cues are obviated. Accordingly, this increases the inﬂuence of the

textual communication one witnesses and takes part in online, because there is nothing

96

else to distract, interfere, or add as an ingredient to the contemplation, processing, or
interpretation of an action.

To note, Chantler (1996) has argued that computer-related crime may be more
dependent on differential association than traditional crime because of the fact that those
who want to engage in the former must acquire the technical skillsets necessary to do
so—-skills not learned through common, everyday experiences. In addition, it has been
documented that those who participate in computer-related deviance are more apt to
associate with others of like mind than would normally be the case for conventional
forms of deviance (Chantler, 1996). This association occurs in a variety of venues, both
ofﬂine (e.g., local area network (LAN) computer gaming parties, technology-related
conferences, small grassroots neighborhood computing organizations) and online (e. g.,
chat channels, newsgroups, message boards, mailing lists). For instance, persons
interested in digital music congregate in various settings for the purposes of discussing
and learning more about the utilization of the ﬁle format technology, hardware and
software associated with MP3s, news updates concerned with copyright infringement or
P2P applications, as well as to partake in general conversation on artists, musicians,
genres, albums, and songs (Mindenhall, 2000; Weisbard, 2000). To note, research has
identiﬁed a similar example of differential association among computer hackers, as
online peer groups are formed to share network intrusion knowledge and provide
practical, emotional, and psychological support for hacking activities (Rogers, 2001).

Due to its inception decades ago, social learning theory in its original form does
not explicitly mention the role of the community generated via computer-mediated

communication. It has been established that the pressures exerted by a collective unit in

97

the physical world are just as strong and inﬂuential in a faceless, nameless, virtual milieu
(Etzioni, 1999; Miller & Gergen, 1998). Another interesting point is that the majority of
traditional deviant behavior, particularly as analyzed by Akers, is communally-oriented in
nature - performed and validated in a group setting. Participating in MP3 is inherently
individualistic and private - generally executed by one person at his or her computer in
the'solitude of a home or ofﬁce. It is typically required, however, that persons interact
with others online (to varying degrees, depending on the software interface used) to
obtain digital music. The communal aspect thus undoubtedly comes into play in
cyberspace. Still, associations with other people ofﬂine may also affect participation
merely through casual conversation about the technology or about particular music or
artists. This double dose of societal pressure seemingly exerts a substantial amount of
inﬂuence on individuals, which guides and shapes their level of participation in the
phenomenon.
Imitation

Once individuals are immersed in an atmosphere conducive to the learning of
techniques, motives, and rationalizations, the patterning of their activity based on the
words and actions of others can follow. This is the easiest way for the commission of
the act to occur, and due to the rampant popularity of MP3s, there is no lack of suitable
models to emulate or mimic in behavior. For example, in a chat channel speciﬁcally
related to digital music, a person need only spend a few minutes watching the unfolding
dialogue before certain recurrent themes are detected. One might be the way in which
music ﬁles are requested or offered. Another might be the way in which users gain

access to private ﬁle servers consisting of hundreds or thousands of songs for download.

98

Because there are no other distinguishing characteristics evident in a Virtual
setting apart from the traits evidenced by one’s own way of communicating via a
keyboard, a neophyte can blend in with a population of experienced users Simply by
acting in a similar manner. The anonymous, detached, and wide-open features of the
Internet necessarily remove certain nonverbal cues which generally factor into the way
individuals typically would respond to one another. These include gesture, posture, facial
animation, variations in voice, social role, status, afﬁliation, and a host of demographic
characteristics (Flaherty, Pearce, & Rubin, 1998; Walther, 1992; Walther, Anderson, &
Park, 1994). Their absence makes it difﬁcult to quickly categorize people and reﬂexively
act towards them based on previous experience or preconceptions. Therefore, in an
online setting with only textual messaging as the vehicle of communication, simply
following the lead of others and acting likewise removes any perceived marginality and
results in quick assimilation into the culture. Saint Ambrose’s advice to Saint Augustine,
“When in Rome, do as the Romans do,” seems perfectly applicable in an online milieu.
Imitation of the behavior of other MP3 users that one meets in a virtual setting can then
take place, further facilitating the commission of the act. Speciﬁcally, the actions of
more experienced users are copied by “newbies” either through speciﬁc prescribed
instruction or through emulation of methods in order to exchange or obtain MP3s and to
become popular and respected among others in the MP3 scene.
Deﬁnitions

Deﬁnitions are also used to further the social learning process and to inculcate a
favorable attitude toward participation in the phenomenon. These generally reﬂect the

opinion that MP3 usage is not wrong and is in fact to be heralded for increasing the

99

availability of music to the average consumer and for allowing more artists to be heard.
For the ordinary individual, it is relatively difficult to come up with deﬁnitions
unfavorable to partaking in the activity because the recording industry and musical artists
seem so removed from the simple, largely anonymous process of downloading 3-5
megabyte ﬁles27 from a computer across the Internet. It is perhaps not easily
comprehended how this practice, multiplied by hundreds of thousands of users, could
actually be harmful.

Chat room environments are disproportionately favorable to MPB s, and this
presumably inﬂuences those in attendance, as the statements and actions of each user
signiﬁcantly reinforce the perceived legitimacy of the activity. Deﬁnitions which
characterize the activity as positive, beneﬁcial, and commonly accepted are extremely
present in the textual interaction among individuals. These deﬁnitions not only champion
the beneﬁts of participating in the phenomenon, but also subtly convict or denigrate those
who are not yet well-versed in the exchange of MP3s, and who have not yet realized the
wealth of satisfaction, pleasure, and excitement it provides. Television and the print
media, to some extent, mention the issues associated with copyright and the grey areas
associated with the technology, but dialogue in MP3-related chat channels hardly ever
breaches the subject of intellectual property rights and inﬁingement. In fact, these
channels would not be in existence if their originators and frequenters were not wholly
supportive of the ﬁle format and the free exchange of high-quality music it precipitates.
Therefore, the impression given to those new to the setting is extremely skewed and

particularly myopic. Without any mention of “the other side of the story” in these

 

27 Most, but not all, songs are approximately this ﬁle size.

100

venues, individuals cannot make an enlightened, informed decision as to whether or not
they should personally support participation with MP3s. As such, most “newbies”
succumb to the inundation of positive deﬁnitions and become acclimatized to the pro-
MP3 atmosphere quickly and easily.

Both Sutherland (1947) and Akers (1998) posit that factors such as frequency,
duration, and intensity of the differential associations and deﬁnitions also inﬂuence how
they affect the social learning process. When considered in the context of the MP3
phenomenon, frequency concerns how often a person is blanketed with statements
endorsing music piracy, or how often a person interacts with those who pirate music.
Duration depends on the amount of time spent in an environment supportive of piracy,
whether participating in a P2P network, interacting with a discussion thread on a MP3-
related newsgroup or website, or communicating with MP3 aﬁcionados Via a personal
messaging program or chat room. Intensity refers to how pointed, enthusiastic, cogent,
and passionate both the providers of deﬁnitions and the deﬁnitions themselves are, and
how inﬂuential they are in guiding, shaping, or bringing about certain behaviors. To
note, deﬁnitions can also be general or speciﬁc, with the former geared primarily to
inﬂuencing either conforming or deviating behavior, and the latter more suited towards
affecting speciﬁc actions (Akers, 1998).

A further delineation is made between “positive” and “neutralizing” deﬁnitions
favoring criminality. Positive deﬁnitions are much rarer, and openly champion the
deviant or criminogenic behavior as beneﬁcial. Neutralizing deﬁnitions are much more
common, and — in line with Sykes and Matza’s (195 7) “techniques of neutralization” —

attempt to rationalize or justify the undesirable behavior as acceptable or appropriate

101

even though an awareness of its undesirability or erroneous nature is present. Both of
these seem prevalent in cyberspace, as a blatant disregard of intellectual property is
evident among so many persons. The relative anonymity and distributed nature of the
Internet allows many individuals to incautiously endorse participation in an illegal
practice without care for potential consequence or backlash from such statements (or the
actions they seek to foster). Such positive deﬁnitions are presumed to be more prevalent
among those substantially ensconced in the MP3 scene, and who participate in it to a
disproportionately high degree.

Neutralizing deﬁnitions are likely more rife among new and casual users — those
not yet deeply rooted in the activity and not yet able to completely disregard any qualms
stemming from engaging in the act. This is arguably because their appropriateness is still
a question mark in the mind of new participants, and some cognitive gymnastics must
take place to unharness oneself from the fetters of conscience and normative ethics,
thereby reconciling incompatible feelings through a rationalizing process. Over time, as
the individual becomes more accustomed to the behavior (and beneﬁts in signiﬁcant
ways), no justiﬁcations must be invoked in order to proceed; the action now becomes
almost reﬂexive, undertaken without contemplation or deliberation. Neutralizations also
prevent the imputation of a deviant identity onto the actor, and preclude the development
of resultant guilt stemming from the wrongful behavior. The theoretical relevance of
sykes and Matza’s (1957) techniques of neutralization to music piracy is not explored in

the current research.

102

Differential Reinforcement

After deciding to engage in or refrain from a behavior, social and nonsocial
reinforcers aid in the persistence and escalation of the activity. Social reinforcement
online can occur if an individual needs a particular song in MP3 format that someone else
possesses, and attempts to trade with that person for another requested song ﬁle. Thus,
the acquisition of something desirable by both parties at no cost to either will foster a
pleasant acquaintanceship, and will perpetuate the activity of exchanging MP3s.
Additionally, those knowledgeable about MP3s, the latest music released, ﬁle servers and
web servers where the best and most popular music ﬁles are stored, and those with
extensive and varied collections of MP3 ﬁles are considered “elite” and are respected and
admired in the MP3 community. As afﬁnity-seeking is a natural function of human
behavior, continued immersion in the MP3 scene may aid in providing these social
I beneﬁts. Nonsocial reinforcement might result after discovering the excellent quality
sound recordings available in MP3 format, and realizing how simple it is to amass a
colossal aggregation of all the music one could possibly want, without any monetary cost,
sans the price paid for a computer and an Internet connection (which many individuals
already possess). Other beneﬁts of the technology itself, such as the ease of distribution
to family, friends, and associates, and the ability to make audio CDS or to burn them in
their original ﬁle format onto recordable discs for portability and use at different
locations, will seemingly reinforce involvement in the MP3 scene.

A person may have a proclivity to participate in a certain deviant behavior, due to
learned attitudes, beliefs, and deﬁnitions favorable to commission. However, s/he will

likely refrain from participation if under the impression that punishment is imminent.

103

Conversely, s/he will likely engage in the behavior if a perception is held that rewards or
beneﬁts will result from the act. These preconceptions and perceptions are shaped by
previous actions - either one’s own, or learned through the experience of another - that
resulted in reward or punishment (Akers, 1.996). As applied to the MP3 phenomenon, it
is hypothesized that reinforcement is largely provided through social interaction,
primarily mediated by computers and the Internet. Further, social learning theory
proposes that individuals adhere to dominant prescriptive behavior online, but that
positive reinforcement (e. g., the procurement of free music, social status among friends,
ease and convenience of access) or negative reinforcement (not having to pay
approximately $15 for a CD for only one or two appealing songs might make the
individual more likely to change direction and participate in the activity. Social and
nonsocial rewards, then, are potent enough to overrule subscription to an ethical set of
values, rendering them discardable so that commission of the crime and reaping of the
perceived rewards may occur.

One lesser-known form of positive reinforcement regards affective outcomes
from participation. Downloading an MP3 is an inherently emotional action. It provides
immediate gratiﬁcation when one desires to listen to a particular song. It also presents
the rewards of convenience and self-satisfaction as high-quality music of relatively small
ﬁle sizes can be shared with friends and colleagues without the need for their provision as
a physical recording on tape or CD. By extension, it gives individuals somewhat of a
guilty pleasure by providing a valued commodity for free, something for which they

normally would have to pay. Irrespective of whether the download of an MP3 later

104

effectuates the purchase of an album, that incipient pleasure is present and accordingly
increases the incidence and frequency of the behavior.

An example of how social learning theory is applicable ofﬂine also deserves
comment. A college student might be introduced to MP3 technology by her roommate,
who seems to always have an exceptional variety of music blaring from her computer
speakers — including oldies from the 19608, disco tracks from the 19708, love songs from
the 19808, alternative grunge rock from the 19908, boy-band pop from the 20008, and
even albums and singles that have not even been ofﬁcially released to the public.
Knowing that it is highly unlikely that someone can afford such a vast CD collection and
also pay for college and living expenses at the same time, the curious student might ask
the source of all the great tunes. The roommate can then introduce the student to the
technology, and show her how many other people in the residence halls are exchanging
ﬁles with each other through the local area network. Moreover, the unfamiliar student
can be taught how millions of individuals globally use P2P applications or visit certain
chat channels to obtain practically any song, from any time period, by any artist, all
within a few minutes through effortless use of their keyboard and mouse to search for and
download MP3s.

The socialization process can continue through informal discussions with friends
and classmates on the topic of music. The knowledgeable roommate might initiate the
neophyte into the scene by showing her how to download her ﬁrst MP3, and then the
process of imitation can take place as one individual learns and models the behavior of
the other. This event has undoubtedly unfolded itself numerous times in dorm rooms

across the United States. Deﬁnitions favorable to MP3 use are likely to proliferate in this

105

context, win the allegiance of the student, and become internalized with little (if any)
dissension or dispute. The college student thus becomes enamored with MP3s and
proceeds to accumulate large quantities of ﬁles. She then emails love songs to her
boyﬁ'iend back home, sets up a server on her computer to distribute copies of her MP3s
to others, leaves her computer on for weeks at a time to continue transferring ﬁles, and
develops a web page to promote the distribution of digital music on the Internet.
Reinforcement occurs as she sees all of her peers downloading MP3s and through the
realization that she no longer must buy music CD8 to meet her needs for great tunes.
Moreover, the behavior is further legitimated and ingrained as the sheer number of
participants in the phenomenon (for all intents and purposes) any possibility of detection
and discipline for her individual actions.

It is worth mentioning that Akers (1985) speciﬁcally points to the development of
self-control following proper socialization”. Following this, individuals generally abide
by normative standards through self-policing mechanisms as they are capable of
controlling their own behavior. Proceeding from this line of thinking, more variation in
music piracy may be explained through the juxtaposition and integration of theoretical
concepts from each of these paradigms. While the primary purpose of the current
research is not to integrate theoretical propositions into a cohesive or unifying whole in
an attempt to explain intellectual property theft online. Future analyses on these data will

venture in that direction.

 

28 Interestingly, when comparing social learning theory and self-control theory, the former asserts that
crime results from learning something (e.g., techniques, rationalizations, motives) while the latter contends
that crime results from not learning something (e.g., how to defer short-term pleasures for long-term gains).

106

Social learning theory points to the methods and manners in which an individual
might be introduced to a criminal activity. As illustrated above, its four theoretical
components of differential association, imitation and modeling, deﬁnitions, and
differential reinforcement can conceivably be applied to music piracy in the same manner
that they have been applied to more traditional forms of crime. The ways in which a
social group contributes to the commencement and persistence of a behavior is important
for analyses, particularly because the behavior under study appears to ﬂourish via its

communal nature and the amicable context in which it occurs.

107

CHAPTER 4: METHODOLOGY

The present study seeks to determine whether GST, SCT, and SLT are all valid
explanatory frameworks in which to study and understand criminality that is conceptually
different from traditional types in two ways. First, the phenomenon at hand — music
piracy - is facilitated by a computer, and computer-related criminality has very rarely
been the subject of academic empirical examination or policy development by criminal
justice personnel. Second, the phenomenon occurs online — over the Internet in an
intangible, nonphysical, virtual realm. Crime that occurs solely in cyberspace has also
suffered ﬁom vast inattention by researchers and practically all criminal justice
practitioners. Four hypotheses are presented below. It is hoped that through this research
endeavor, a comprehensive picture of predictive elements associated with online
intellectual property theft will be obtained.
Population

The subject population of the current study is undergraduate students at a large
public university in the Midwest region of the United States. The empirical validity of
many criminological theories have been tested through the use of data collected from
samples of college and university students; indeed, this is a widely prevalent and
acceptable method in this discipline (Mazerolle & Piquero, 1998; Nagin & Paternoster,
1993). Most students have high-speed access to the Internet in their residence hall rooms,
or through cable modem or DSL connectivity in their off-campus homes. Others may
primarily use a dialup modem to connect to the university’s network, and while speeds
are signiﬁcantly slower in this context, online access is still attained. Students are

required to use the Internet for a variety of acadenric reasons, including research,

108

correspondence, and various types of scholarly projects. Moreover, tasks as essential as
registration for courses is only possible online, demonstrating the tremendous necessity
of connectivity for those currently enrolled in the university. Notwithstanding school-
related responsibilities, the Internet has become valuable for meeting social and personal
needs, and thereby plays a large, functional role in the lives of most students.

An increased frequency of online activity by students enhances the likelihood of
being introduced to, and participating in, arguably questionable types of behavior on the
Internet. While this author will leave some of the more seedier and outrageous examples
to the reader’s imagination, online intellectual property theft in the form of music piracy
is one such activity in which students may participate. Additionally, augmenting the
likelihood of the behavior’s occurrence are a host of contributive factors, including a lack
of enforcement of rules governing acceptable use of computer and network resources, a
deﬁciency in delineating ethical and unethical standards of behavior by instructors and
other authority ﬁgures, and a higher level of curiosity, experimentation, and general
deviant inclinations among the college-aged population”.

Instrument

An extensive survey instrument was constructed and reﬁned in order to gather
data to examine the veracity of the aforementioned hypotheses. It has been included in
Appendix A. The questionnaire commenced with a short general introduction of the
study, stated the protections afforded to the subject, provides a summary of how data is

being collected, and gave contact information both of the primary investigator and chair

 

29 Research on the subject of cheating, plagiarism, and software piracy has sufﬁciently illustrated this point
(Agnew & Peters, 1986; Buckley, Wiese, & Harvey, 1998; Eining & Christensen, 1991; Im & Van Epps,
1991; Wong, Kong, & Ngai, 1990).

109

of the relevant institutional review board. Questions representing the three general
criminological theories were then presented to the respondent. First, six questions
intended to measure strainful life experiences were given, and stem from Broidy’s (2001)
empirical test of Agnew’s (1992) GST. These items asked the respondent to reﬂect on
the last six months and indicate whether they received a bad grade in a class, broke up
with an intimate partner, experienced weight gain or 1088, been ﬁred or laid off from a
job, had money problems (i.e., had difﬁculty paying tuition, rent, bills), or been a victim
of a crime. Possible responses were true or false. Next were six items from the 24-item
scale created by Grasmick et al. (1993) operationalizing the six constituent elements of
self-control, in order to assess the relationship of this intrinsic characteristic to music
piracy among the sample. This decision stemmed from its extensive empirical utilization
and agreed-upon appropriateness in the previously reviewed studies. The variables
included: “I often do what brings me pleasure here and now” (to measure impulsivity);
“When things get complicated, I tend to quit or withdraw” (simple tasks); “I ﬁnd no
excitement in doing things I might get in trouble for” (risk seeking); “I try to look out for
others ﬁrst, even if it means making things difficult for myself” (self centered); “I don’t
lose my temper very easily” (temper); and “I feel better when I am on the move rather
than sitting and thinking” (physical activities). AS is evident, one question was selected
for each of the six theoretical components of Gottfredson and Hirschi’s (1990)
conceptualization of self-control. In the current research, these are considered
“attitudinal” measures of self-control.

Five questions measuring various types of antinormative conduct that range in

severity were next presented. Their purpose is to provide a rough sketch of whether

110

“analogous” behaviors are related to participation in nontraditional criminality in the
same way that they are related to conventional forms of crime (Cochran, Wood, Sellers,
Wilkerson, & Chamlin, 1996; Paternoster & Brame, 1998, 2000; Tremblay, Boulerice,
Arseneault, & Niscale, 1995). They included whether the respondent had: skipped more
than 10 class periods in the past year; lied to a professor/instructor either via email,
telephone, or in person at least once in the past year; plagiarized on a school assignment
at least once in the past year; drank alcohol before s/he turned 21; or driven a vehicle
while under the inﬂuence of alcohol at least once in the past year. These variables are
considered “behavioral” measures of self-control.

Following the strain and self-control items are sixteen questions to be used as
controls. These measure: various self-evaluative judgments and notions; and various
social-structural factors that are a part of the respondent’s life. It will be useful to gain
insight into personal and institutional factors which may inﬂuence the behavior of
participants. Furthermore, the research will be able to determine the extent to which
variations in life perspectives and participation in social circles are correlated with the
activity. The subsequent section of the survey presented thirty-seven questions largely
measuring social learning theory; the sheer number of items was deemed essential to -
most accurately grasp the four elements of the theory — differential association, imitation,
deﬁnitions, and differential reinforcement. Twenty-two items seeking to discern the
justiﬁcations that individuals might utilize to free themselves ﬁ'om conventional and
normative constraints and thereby participate in digital music piracy are then presented.

Twenty-two questions intended to elicit the frequency and scope of their actual

participation are subsequently given, and the instrument terminates with an assortment of

111

items seeking demographic information from the respondent. A conscious effort was
made to create and present questions in as neutral a manner as possible, so as not to
offend individuals or prejudice their answers. This is imperative not only to conduct
good research, but also because self-reported criminality is the dependent variable.
Candidness and forthrightness of responses must be encouraged so that internal validity is
not threatened, and so that consistency in interpretation is fostered — as best as possible —
to most accurately evaluate key concepts in the work. To note, the written and oral
introduction to the survey will hopefully assuage any inhibitions the respondent might
have.

Due to the fact that the survey is closed-ended, its structure constrains the
responses and therefore prevents individuals from providing comments, feedback, or
more richer answers to these questions if they so desire. This was necessary, however,
due to resource limitations associated with the project. The vast majority of questions are
provided with a Likert-scale answer set; the remainder are either true or false, or had
answer choices speciﬁc to the inquiry posed. Also, the direction of answer choices was
varied to prevent automatic and lackadaisical responses by participants. Apart from
testing the extensibility of the three major general theories of crime, it is hoped that this
research will cumulatively advance society’s understanding of the causative elements of
online intellectual property theft speciﬁcally, and — to some degree - Intemet-based
criminality in general. Also, such derived knowledge Should inform decision-making
related to policy and programming strategies that can be implemented to respond this

form of wrongdoing in the most advantageous manner. As a consequence, it is hoped

112

that the ﬁscal and social damage caused by copyright inﬁingement across the Internet
will be ultimately curtailed.
Sampling Procedures

While creating a sampling frame of classes to survey, much care was taken to
ensure that the resultant group of individuals were predominately representative of the
entire student population at this university. Three primary stages of data selection took
place. First, a list of the ﬁfteen colleges was obtained, as well as a list of departments
and schools inside each college. Second, two or three majors inside each college were
selected so that speciﬁc classes that might be surveyed could be identiﬁed. Some majors
were somewhat conventional in nature, and offered classes which all undergraduates
would have the opportunity to take - such as introductory courses in Computer Science
and Psychology. Other majors were highly speciﬁc and offered classes that only those in
that department would take - such as Biochemistry, Zoology, and Finance. These courses
ranged from the 100-200 level (generally populated by freshmen and sophomores) to the
300-400 level (largely consisting of juniors and seniors).

Once a few majors in each college were randomly chosen, the third stage of data
selection occurred. A concerted effort was made to randomly select 1 or 2 lower-level
and 1 or 2 upper-level classes through the use of the university’s online course catalog.
A comprehensive list of these potential classes to survey was then created, and emails
were sent to each respective professor or instructor. In these emails, a description of the
project was given, along with a link to a web page where the survey instrument might be
viewed. A request was then made for approximately 20 minutes at the beginning or end

of their class period so that the surveys could be administered. Overall, 169 professors

113

representing 185 classes were emailed, and 15 professors representing 16 classes agreed
to the request.

The aforementioned method is known as purposive sampling for heterogeneity.
This technique seeks to obtain a certain number of people in speciﬁed groups (such as
college majors), and is not inordinately concerned with proportionality but rather to
obtain a sufﬁciently diverse sample on one or more characteristics. Determining the
prevalence of music piracy among university students across the gamut of possible
majors is a goal of this study, rather than restricting it to those in speciﬁc disciplines,
such as computer science or criminal justice. “The general strategy is to identify
important sources of variation in the population and then to select a sample that
represents this variation” (Singleton & Straits, 1999:158). Individual area of study was
one notable variable in which students would differ - and which may accordingly affect
their ideologies toward music piracy. Therefore, a minimum of two classes (one lower-
level and one upper-level) from a minimum of two majors in each of the colleges in the
school where this research was conducted was deemed necessary to facilitate cross-
disciplinary comparison of individuals. The classes and majors selected were chosen
based on this author’s “expert” judgment after consideration of the population at hand
and the goals of the study. For the purposes of this analyses, stratiﬁed random sampling
would not have added much value in terms of precision or generalizability, as the primary
objective was not to perfectly mirror the demographic proportions of the student
population but to garner a sample generally representative of that larger group. Despite
the fact that permission was given in only 16 of 1985 classes, a broad amount of majors

were expected to be represented in those 16 courses due to their interdisciplinary content.

114

During the data collection phase, the researcher and the subject matter was
introduced, and the fact that there would be no cost associated with participating except
for the time spent in composing a response. Students were verbally informed of the
conﬁdentiality and anonymity of the survey, that participation was completely voluntary,
and that it would take approximately 20 minutes to complete. They were also instructed
to refrain from revealing their name, student number, or any other identifying information
when ﬁlling out the questionnaire. In addition, potential respondents were made
cognizant that only group totals will be consolidated and released at the culmination of
the project. This was essential to protect the rights of the respondents, to encourage a
greater number of truthful responses, and to garner a reliable cross-section for measuring
the relevant constructs. As mentioned earlier, all of this information was also expressed
at the top of each questionnaire.

To note, the exact title of the study was not revealed to avoid predisposing
respondents to answer in a socially acceptable manner. The title of the instrument was:
“Questionnaire on Participation In and Attitudes Towards MP3s.” Once all of the data
were collected, the instructors and professors who allowed their classes to be surveyed
were asked to read a short debrieﬁng statement to their students. It detailed the exact
nature of the study and reconciled the mild deception previously used to garner unbiased
responses.

The study was restricted to undergraduate students because they are more
representative of traditional conceptions of the “college population,” and because one
might argue that they are categorically different in many ways than those in graduate

school. Nonetheless, the demographic question related to the respondent’s year of study

115

did include a “graduate school” answer choice in case a graduate student was enrolled in
a higher-level undergraduate class to eam elective credits. Those who identify
themselves as graduate students were removed from the analysis.
Pretest

Prior to its administration in classrooms, the instrument was pretested among a
select group of colleagues in this author’s department, as well as in two upper-level
undergraduate criminal justice classes (N=52). The retrieved data showed that a sizable
number of participants downloaded MP3s, and indicated that there would be sufﬁcient
variation in the dependent variable to facilitate statistical analyses. Furthermore, more
informed decisions were made possible as to which variables should be kept and which
should be removed (further explained below), and identiﬁed that a greater variety of
items measuring pirating behavior were necessary. For example, the instrument
employed in the pretest only asked about current participation in the phenomenon; the
revised survey inquired about participation in years past. This would enable a
longitudinal perspective of trends and patterns in participation to be captured. Through
the pretest, feedback on clarity, consistency, and content of the survey items was also
retrieved and considered. Many of these comments were subsequently incorporated to
preclude conceptual and operational problems from compromising the validity of the
retrieved data. This ﬁne-tuning of the instrument greatly assisted in the objective of
posing properly constructed questions and obtaining responses which most accurately
represented the primary concepts in the study.

As mentioned, the pretest was useﬁrl in reﬁning and paring down the number of

items on the questionnaire. Indeed, the primary sentiment revealed by students who took

116

the pretest was that the instrument was too lengthy. It initially consisted of 124
questions; however, upon reﬂection of the results stemming from the pretest, it was
determined that many items were superﬂuous for measurement purposes. As a
consequence, a concerted attempt was made to only include those that were most
statistically and theoretically relevant.
Speciﬁcation of Variables of Interest
S_tr2_1i_n

Following data collection, conﬁrmatory factor analyses (CF A) was conducted on
the six initial strain items to determine which items loaded together best for inclusion in
the statistical analysis”. A cut point of .5 was selected, which left three variables ﬁ'om
which a factor score would be created to measure strain in the statistical models; see

Table 1 for their factor loadings and reliability alpha.

 

Table 1: Strain Factor Score Factor Loading8_
Broke up with an intimate partner 0.546
Experienced weight gain or 1088 0.734
Had money problems (e.g., not being able to pay tuition, rent, bills) 0.635

 

(Chronbach’s a = 0.287; Eigenvalue = 1.240)

Self-Control
First, CFA was conducted on the initial six attitudinal self-control questions.
Respondents could select among “Strongly Disagree”, “Disagree”, “Neutral”, “Agree”,

and “Strongly Agree” as possible responses. Three items loaded above the selected cut

 

3° It should be noted here that the reliability of factors is dependent on the number of indicators per factor
and largely independent by the number of cases (or the interaction of N with the number of factors) (Marsh
& Hau, 1999). Also, there is general agreement that more indicators per factor is better than fewer. At
least three indicators are desirable, and that a large number per factor is not desirable unless N is sizable
(Marsh & Hau, 1999).

117

point of .500 (see Table 2), and were consequently used to create a factor score

measuring Attitudinal Self-Control.

 

Table 2: Attitudinal Self-Control Factor Score Factor Loadings
When things get complicated, I tend to quit or withdraw -0.634

I try to look out for others ﬁrst, even if it means making things difﬁcult

for myself 0.582

I feel better when I am on the move rather than sitting and thinking 0.588

 

(Chronbach’s a = -0.051; Eigenvalue = 1.087)

Second, a list of ﬁve “analogous” behavioral measures of self-control were
initially compiled; possible responses were true or false. CFA revealed only four
statements with factor loadings greater than .500. See Table 3 for the loadings and

reliability alpha statistic.

 

Table 3: Behavioral Self-Control Factor Score Factor Loadings
I have skipped more than 10 class periods in the past year 0.619

I have lied to a professor/instructor either via email, telephone,

or in person at least once in the past year 0.680

I have plagiarized on a school assignment at least once in the past year 0.598

I have driven a vehicle while under the inﬂuence of alcohol at least once

in the past year 0.595

 

(Chronbach’s a = 0.472; Eigenvalue = 1.557)

Control Variables

An assortment of questions were also included to ascertain the inﬂuence of
personal, social, and structural factors on the outcome measures. The answer set for the
ﬁrst subset included Strongly Disagree, Disagree, Neutral, Agree, or Strongly Agree, and
the items posed were as follows: “I am optimistic about my future”; “I have difﬁculty
maintaining long-term relationships”; “I actively expect the best ﬁom people and
situations”; “My emotional life is unstable”; “1 am able to express the feelings I have,
whether happy, sad, angry, frustrated, or confused”; “I am not comfortable with myself

when around others”; “I have difﬁculty achieving long term goals”; and “I am happy.”

118

Other variables sought to identify the extent of the respondent’s participation in
social activities, and his/her level of comfort in such an environment. These included:
“How many student organizations (like the Debate Team, Campus Crusade, Outing Club,
etc.) did you regularly participate in over the past year?”; “How many sports did you
regularly participate in (including running/working out) over the past year?”; “On
average each month, how many times do you participate in religious activities such as
attending a church, temple, or scripture study session?”; “I have a (very low, low,
moderate, high, very high) amount of ﬁiends in the area”; and “I would rate my self
esteem as (very low, low, moderate, high, very high)”

The third subset of control items attempted to roughly gauge the inﬂuence of
close family and ﬁiends on an individual’s mentality, well-being, and perspective. These
included: “On a scale of 1-5 (with 5 = “healthy and warm” and 1 = “cold, distant, and
completely dysfunctional”), how would you rate the quality of your relationship with
your parent(s) or guardian(8)?”; “On a scale of 1-5 (with 5 = very strongly), how strongly
have your parents shaped your personal perspective on life?”; and “On a scale of 1-5
(with 5 = very strongly), how strongly have your friends shaped your personal
perspective on life?”

Sociailearning Theory

Thirty-seven social learning theory variables were initially created, and CFA was
utilized to determine the appropriateness of the item groupings for each of the four
elements of differential association, imitation and modeling, deﬁnitions, and differential
reinforcement. To note, the imitation construct of social learning theory was particularly

difﬁcult to operationalize, due to its apparent theoretical overlap with differential

119

association. As mentioned earlier, imitation follows immersion in an environment where
socialization takes place, but such a concept was difﬁcult to capture in the construction of
survey items. CFA and reliability analyses were run on variables perceivably measuring
imitation. They revealed multiple factors, weak loadings, and extremely poor alphas —
posing a sizable problem. This author sought to ﬁnd a way to include imitation as a
separate concept because of its importance a distinct theoretical element of social
learning theory, and was hesitant to only measure differential association, deﬁnitions, and
differential reinforcement like other studies had done (e. g., Lersch, 1999). In Akers et
al.’s (1979) test of the theory, the component of imitation was found to be weakly related
to the frequency of alcohol and marijuana use. The researchers qualiﬁed the low levels
of variance explained by stating that modeling is the most narrow of the four empirical
phenomena and that “the interrelationships speciﬁed in the theory would indicate that
removing imitation has less effect because its impact is still reﬂected to some extent in
the remaining broader measures” (Akers et al., 1979:647).

From a practical perspective, though, MP3 participation truly seemed to be
inﬂuenced by individuals modeling the behavior of other music pirates, and thus its
individual inclusion seemed imperative. It appeared that imitation and differential
association could be highly interrelated, and so a determination was made to include all
variables of both of those groups in an exploratory factor analyses model. The two
primary factors seemed to differentiate the variables in a theoretically expected manner;
and measures concerning inﬂuences in real life (ofﬂine) clustered separately from those

measures related to online inﬂuences.

120

It may be that differential association takes place largely in an ofﬂine context —
where interaction is presumably more frequent, profound, and subtly persuasive in its
effect on the actions of a person. Imitation and modeling is more of a direct, perceptible,
and conspicuous process, as well-ensconced participants are concurrently online with
those who are new to the phenomenon, and the behavior of the former is immediately
observable for emulation by the latter. In sum, differential association was
operationalized by four variables speciﬁc to ofﬂine interaction: “My friends support my
MP3 usage”; “I associate with others in real life (ofﬂine) who are supportive of MP3
usage”; “I was introduced by another person in real life to MP3 8”; and “I have learned the
techniques of using MP3s from my friends” (see Table 4). Imitation31 was represented
by three variables endemic to the environment of cyberspace: “I have learned the
techniques of using MP3s from television or print media”; “I have learned the techniques
of using MP3s from online sources (web pages, chat rooms, etc)”; and “I associate with

others online who exchange MP3s with me” (see Table 5).

 

Table 4: Differential Association Factor Score Factor Loadings
My friends support my MP3 usage I 0.825
I associate with others in real life (ofﬂine) who are supportive of MP3 usage 0.821
I was introduced by another person in real life to MP3s 0.761

 

I have learned the techniques of using MP3s from my friends 0.698
(Chronbach’s a = 0.774; Eigenvalue = 2.421) .

 

Table 5: Imitation/Modeling Factor Score Factor Loadings
I have learned the techniques of using MP3s from television or print media 0.817

I have learned the techniques of using MP3s from online sources

(web pages, chat roonrs) 0.829

I associate with others online who exchange MP3s with me 0.576

 

(Chronbach’s or = 0.595; Eigenvalue = 1.686)

 

3 ' Skinner and Fream (1997) measured inritation by asking respondents about the sources from where they
might have learned computer crime techniques, such as from family, teachers, books or magazines, ,
television and movies, or computer bulletin boards. Their rationale was that even though such items do not
measure modeling exactly as it was proposed by Akers, they could Shed light on imitated sources that
provide a means for behavioral learning.

121

The ﬁnal set of items measuring social learning deﬁnitions were as follows: “One

of the reasons I download MP3s is because I will not purchase the music”; “One of the

reasons I download MP3s is because I feel the recording industry has been overcharging

the general public for music tapes and CD8”; “One of the reasons I download MP3s is

because many musicians and the recording industry make millions of dollars anyway, and

downloading MP3s of their songs does not really cut into their income”; and “One of the

reasons I download MP3s is because I think music should be free” (see Table 6). The

ﬁnal set of differential reinforcement items included: “It is a great beneﬁt to sample new

music through MP3s”; “It is a great beneﬁt to be able to transfer assorted MP3s onto an

audio/data CD or a portable MP3 player so that I can have music on-the-go”; “It makes

me feel good to download a song that I have wanted”; and “It is a great beneﬁt to me to

be able to access music freely” (see Table 7).

 

 

 

Table 6: Deﬁnitions Factor Score Factor Loadings

One of the reasons I download MP3s is because I *will not“ purchase the

music 0.656

One of the reasons I download MP3s is because I feel the recording industry

has been overcharging the general public for music tapes and CD8 0.724

One of the reasons I download MP3s is because many musicians and the

recording industry make millions of dollars anyway, and downloading MP3s

of their songs does not really cut into their income . 0.775

One of the reasons I download MP3s is because I think music should be free 0.656
(Chronbach’s or = 0.658; Eigenvalue = 1.986)

Table 7: Differential Reinforcement Factor Score Factor Loading§_

It is a great beneﬁt to sample new music through MP3s 0.889

It is a great beneﬁt to be able to transfer assorted MP3s onto an audio/data

CD or a portable MP3 player so that I can have music on-the-go 0.881

It makes me feel good to download a song that I have wanted 0.791

It is a great beneﬁt to me to be able to access music freely 0.803

 

(Chronbach’s a = 0.862; Eigenvalue = 2.836)

This ﬁnal set of ﬁfteen variables measuring the four tenets of social learning

theory did appear to warrant further examination. Could it be determined whether the

122

constructs can be distinctly operationalized and measured? Or, is overlap in their
measurement inevitable due to the tight interrelationships among the concepts? It was
useful to discover whether the data might lend additional understanding to this theoretical
complexity.

Empirically Disentangling the Tenets of Social Learning Theory

 

To begin, an exploratory factor analyses with varimax rotation was conducted on
these ﬁfteen variables presumed to accurately and separately measure the four tenets of
social learning theory. Four Eigenvalues over 1 were identiﬁed, but the scree plot
indicated that there were only two primary factors, the ﬁrst explaining 26.29% of the
variance, and the second explaining 13.23% of the variance. The rest of the explained
variation appeared to be statistical noise. Looking at the factor loadings across the four
dimensions and choosing a cut point of .45 or higher (as this threshold seemed to separate
those variables which loaded decently on a factor from those that did not), all four
differential association and all four differential reinforcement variables loaded on the ﬁrst
factor (see Appendix B, Table A). Furthermore, all three imitation variables and all four
deﬁnitions variables loaded on the second factor. Admittedly, this researcher’s
conceptualization of the tenets may be susceptible to error and consequently may have
confounded their distinct nature when creating the variables.

Notwithstanding that possibility, though, it may simply be that it is difﬁcult to
perfectly disentangle differential association from differential reinforcement and measure
them as two separate constructs. What are some reasons for this ﬁnding? The presence
of peers who support, inculcate, encourage, endorse, and share certain motives,

rationalizations, and drives related to an activity is intrinsically a positive and rewarding

123

element that is perpetuated through continued participation. Similarly, the presence of
these peer associations (and the attendant social and nonsocial beneﬁts) will perceivably
diminish if the behavior is reduced in frequency, and individuals may wish to preserve
that peer group and consequently continue an activity in order to avoid pain associated
with their possible loss. This, of course, is equivalent to negative reinforcement.

It is more difﬁcult to posit why the variables ostensibly measuring imitation and
deﬁnitions loaded on the same factor. Imitation refers to the modeling of actions of
others, and the transmission of knowledge, attitudes, beliefs, and techniques associated
with those actions. It appears to connote more of a physical replication of a person’s
behavior. If individuals download MP3s after watching a friend or family member in real
life, this would seem highly applicable. In cyberspace, however, such physical
replication is not possible simply due to the lack of material cues and the fact that all
interaction takes place through computer-mediated communication. As such, imitation
online would appear to result from observing and internalizing typed words and
commands that represent knowledge, attitudes, beliefs, and techniques of MP3
participation.

Once a person who wants to participate in the phenomenon meets and interacts
with an individual already well-versed with MP3s, the latter person can serve as a
palpable model to emulate solely through that which is recognizable and evident in their
textual content. Any demarcating line between content that induces imitation and content
that induces deﬁnitions is blurred, because the same actions might affect one, the other,
or both tenets of social learning theory. Deﬁnitions, as mentioned earlier, are evaluative

criteria designating a behavior as good or bad and thereby qualifying them as appropriate,

124

desired, or justiﬁed (Akers et al., 1979). Their learning and internalization may not occur
in a disparate moment from when imitation takes place. That is, the social setting on the
Internet which provides sources of imitation and deﬁnitions in support of the activity may
promulgate both through the same elements and individuals, thereby obscuring any
attempt to disconnect the two and measure each on an individual basis.

The next step was to run factor analyses with promax rotation, under the
assumption that the variables were correlated with each other. The ﬁrst run included the
variables measuring Differential Association and Differential Reinforcement. All of the
factors loaded relatively highly on the one factor that was identiﬁed (supported by the
resultant scree plot), further reinforcing the fact that these items may not measure
distinctly different concepts (see Appendix B, Table B). A reliability analysis revealed
an alpha of .878, indicating collective measurement of the same construct. When
separated into their two groupings, the reliability alphas fOr Differential Association and
Differential Reinforcement were .774 and .862, respectively. A decision was made, then,
to maintain the operationalized separateness between the two constructs by utilizing each
distinct set of variables rather than combining them.

The second run included the Imitation and Deﬁnitions variables. Contrary to
what happened in the ﬁrst run, two factors were identiﬁed with each set of variables
loading separately from the other. The reliability alpha value for these items together
was .659; their alpha values when separated into the two tenets was .595 for Imitation
and .658 for Deﬁnitions (see Appendix B, Table C). Unquestionably, these analyses

highlight some of the complexities inherent in attempting to empirically distinguish

125

between the four theoretically deﬁned elements of social learning theory, especially when
considering an Intemet-based activity.
Dependent Vagables

Thirteen dependent variables in the current research seek to measure the
frequency of an individual’s participation in music piracy. Respondents were asked to
indicate how many MP3 ﬁles they personally downloaded last week and in an average
week (0, 1-5, 6-10, 11-20, More than 20), and last month and in an average month (0, l-
25, 26-50, 51-100, More than 100) this year, last year, and two years ago. Furthermore,
they were asked to indicate how many they downloaded in totality during each of the past
three years (0, 1-10, 11-100, 101-1000, More than 1000), how many they had
downloaded over the course of their life thus far (0, 1-100, 101-500, 501-2000, 2001 or
more), and how many total complete music albums in MP3 format they had obtained
online (0, 1-5, 6-10, 11-20, More than 20).

These variables were factor analyzed using promax rotation, and the resultant
scree plot depicted a tremendous drop between the ﬁrst and second (and subsequent)
Eigenvalues. The ﬁrst explained 55.39% of the variance across the model, while the
second explained 14.03%. As such, a one-factor solution was forced; see Table 8 for the
detailed loadings. Additionally, the alpha value for these thirteen dependent items was
.930, indicating that if a respondent answered a certain way for one of these questions, it

was extremely likely that s/he answered the same way for the other questions.

126

Table 8: Music Piracy Dependent Variable Factor Score Factor Loadings

How many MP3 ﬁles downloaded in the last week? 0.590
How many MP3 ﬁles downloaded in the last month? 0.646
How many MP3 ﬁles downloaded since the beginning of 2003? 0.751
How many MP3s do you, on average, download per month? 0.744
How many did you download in an average week exactly one year ago? 0.810
How many did you download in an average month exactly one year ago? 0.814
How many did you download in an average week exactly two years ago? 0.772
How many did you download in an average month exactly two years ago? 0.776
How many MP3 ﬁles did you personally download in 2002? 0.819
How many MP3 ﬁles did you personally download in 2001? 0.783
How many MP3 ﬁles did you personally download in 2000? 0.673
How many total complete music albums in MP3 format have you obtained 0.604
online?

How many total MP3s have you downloaded over the course of your life thus 0.836
far?

 

(Chronbach’s or = 0.930; Eigenvalue = 7.201)
Note: one factor solution forced

Similar questions have been utilized in the descriptive studies on MP3s conducted
by various research ﬁrms (Angus Reid Worldwide, 2000a, 2000b; Jay, 2000; King,
2000a; Latonero, 2000; Leannonth, 2000; Pew Internet & American Life Project, 2000;
Reciprocal Inc., 2000a, 2000b; Webnoize, 2000). Determining immersion in intellectual
property theft through items inquiring about the frequency of participation in the activity
has been supported by research on software piracy (Rahim, Seyal, & Rahman, 1999;
Sims, Cheng, & Teegen, 1996; Solomon & O'Brien, 1990; Wood & Glass, 1995) .
Inquiring about the number of legal and illegal songs acquired by the individual has been
supported in the only MP3 study to date utilizing a behavioral perspective (Agrawal et
aL,2003)

Other questions in the survey ask the number of hours each week the respondent
spends looking for MP3s and what activities are done with MP3s (e.g., creating an audio
CD from MP3 ﬁles, made an MP3 from an audio CD or another sound source, listening

to them on a computer, listening to them on a portable MP3 player, bunting them to CD,

127

sharing them with friends, selling them). The individual is also asked to assess the ratio
of his or her download/upload time spent transferring MP3 ﬁles online (I do not
participate with MP38, 0% of the time downloading and 100% uploading, 25% and 75%,
75% and 25%, 100% and 0%) and the percent of MP38 possessed that are not personally
created from CD8 the individual owned, or are not of songs that the individual owns on
CD. Three questions then follow which seek to capture perceptions of the legality of
downloading and uploading MP38. These include: “Do you believe that receiving or
providing MP38 should be illegal?”; “As far as you know, is receiving or providing MP38
illegal?”; and “Do you refrain from obtaining MP3s because you believe it is illegal?”.
Finally demographic information was solicited through inquiries as to the
respondent’s race, gender, age, year of studies, major, parents’ annual household income,
employment status, living situation, type of Internet connectivity, variety of Internet use,
and proﬁciency of Internet use. A copy of the questionnaire is provided in Appendix A;

it can be consulted for additional details on the items utilized in this study.

128

Hypotheses
Hypothesis 1

Participation in music piracy varies based on the extent to which an individual is
proﬁcient with using the Internet, and on the range of online activities in which the
respondent has participated. Speciﬁcally, those who are highly skilled in performing
various Intemet-based activities, and those who take broader advantage of the
possibilities available online, will be more likely to engage in music piracy.
Hypothesis 2

The elements of social learning theory, self-control theory, and general strain
theory are all signiﬁcantly related to Internet music piracy. That is, each “general” theory
is appropriately named and has the capacity to explain variation in a crime that is highly
nontraditional both in content and in context”.
Hypothesis 3

The general theory that explains the most variation in online music piracy is
social learning theory, because above all this particular crime is learned from, and
supported by, the inﬂuence of individuals and institutions in society. Self-control theory
and GST are second and third in explanatory power respectively when considering the

three general theories.

 

32 A partial or completely new conceptualization of cybercrime may be required if these general theories
fail to demonstrate adequate predictive capacities.

129

CHAPTER 4: ANALYSES

This research seeks to capture a proﬁle of a typical music pirate in terms of their
demographic characteristics, and also assess the capacity of general strain, self-control,
and social learning theory to explain variance in MP3 participation. To accomplish this,
various statistical techniques will be employed: univariate analyses in the form of
frequency distributions; bivariate analyses in terms of crosstabulations, correlations, and
one-way analyses of variance (ANOVA) procedures; and multivariate analyses through
logistic and multinorrrial logistic regression models.

To test Hypothesis 1, two items were included in the questionnaire to measure
proﬁciency and variety of Internet usage. Examining the distribution of these categories
among music pirates will be possible through One-Way Analyses of Variance (AN OVA),
which assesses whether the population means are equal by calculating the signiﬁcance of
the difference of the sample means. Identiﬁcation of which category of the dependent
variables differs signiﬁcantly ﬁ'om the others in its power to inﬂuence music piracy can
occur through the Bonferroni Post Hoc test. To test Hypotheses 2 and 3, two types of
multivariate regression analyses will be conducted. Regression allows multiple
predictive factors to be examined together in the same model, to determine the inﬂuence
of one component while holding the others constant. Logistic regression analyses can be
performed to ascertain if general strain, self-control, and social learning theory elements
signiﬁcantly increase the likelihood that individuals will pirate music. Multinomial
logistic regression can be similarly performed to determine whether certain theoretical

components (as measured by factor scale variables) differentiate the intensity of

130

participation among music pirates. Rationale for the use of these analytic techniques are
provided with the statistical results.
Descriptive Statistics

Descriptive statistics allow researchers to summarize data in an easily
interpretable format, and can serve as a foundation for subsequent multivariate analyses.
The following text provides some basic distributions of the important demographic
characteristics of respondents in the sample, as well as crosstabulated percentages of how
general participation in music piracy varies across these groups. Speciﬁcally, the
variable of “total MP38 downloaded over the course of one’s life” - with response
categories of 0, 1-100, 101-500, 501—2000, or 2001 or more - was chosen to paint a
picture of which demographic groups participate in music piracy more than others. These
uniVariate measures, of course, do not take into account any other variables. These
ﬁgures are inclusive of a ﬁnal sample Size of 2032 individuals, following listwise
deletion of those cases with missing values.

To begin, 57.6% of the sample were female, and 7.8% of women sampled had
downloaded over 2000 MP38 in their lifetime (compared to 22.8% of men sampled).
Men, though, were more frequent participants in music piracy, with 22.8% having over -
2000 MP3 ﬁles (as compared to 7.8% of women). Interestingly, young males have been

identiﬁed as the population that disproportionately participates in software piracy (Rahim

131

et al., 1999; Sims et al., 1996; Solomon & O'Brien, 1990; Wood & Glass, 1995), which
mirrors these ﬁndings concerning music piracy33 .

The vast majority of the sample were Caucasian (77.9%), and the distribution of
total MPBS ever downloaded across racial groups was relatively similar (see Table 9). To
note, 10.1% of the sample were African-American, but almost 1/4th of that demographic
group (24.8%) had never downloaded an MP3. This ﬁgure indicates that African-
Americans do not participate in this activity to the same extent as other races. Almost
half of the sample were not employed at all (49.1%), approximately a ﬁfth worked 20
hours a week (19.5%), and only 3.2% disclosed that they worked 40 hours each week.
Interestingly, 22.7% of these full-time working students had downloaded over 2000
MP38, as compared to 13.6% of those who did not work at all during the week.

The majority of the sample was 19 years of age or younger (57.6%), and 11.5% of
this group had downloaded over 2000 MP38, compared to 18.1% of those 20 years of age
or older. With regard to educational level, 31.4% were freshmen, 28.9% were
sophomore, 24.2% were juniors, and 15.5% were seniors. A larger proportion of seniors
(19.7%) belonged to the heaviest group of MP3 participants than the other classes.
Slightly over half of those sampled lived on campus in a dormitory (55.2%) and 88.9%
had high-speed Internet access. Consonant with intuition, more MP3s were downloaded
by this group than those who connected to the Internet via dialup modem at home, and

those who did not have Internet access at their place of residence.

 

33 Interestingly, the largest purchasing consumers of software, however, are professionals usually past the
college age-group, while the largest purchasing consumers of music are generally those who are in
secondary and post-secondary education (Bhattacharjee et al., 2003). It seems that the piracy of music
would negatively affect the revenue Stream of that industry to a much greater degree than the inﬂuence of
software piracy on the producers and developers of applications and games.

132

The largest proportion of students stated that their major was housed in the
College of Social Science (24.8%); the heaviest downloaders belonged to the College of
Communication Arts and Sciences (21.8%) and the College of Engineering (20.7%).

This might be expected because those majors require more competence and participation
with computers than do some of the others, and a larger amount of proﬁciency may be
correlated with greater MP3 downloading“. Nevertheless, MP3 participation is generally
distributed similarly across majors, indicative of its prevalence across the entire student

body.

 

3‘ To note, proﬁciency as well as variety of Internet use are later tested in a bivariate analyses to determine
their predictive role in music piracy.

133

Table 9. Demggraphic Characteristics and Participation in Pirating (N=2032)

 

Total MP38 Ever Downloaded

 

Sample "/o 0 1-100 101-500 501-2000 2001+

 

 

 

S_ex_
Female 56.7 16.1 14.6 30.2 31.3 7.8
Male 43.3 7.2 9.0 22.6 38.5 22.8
Mr
White 77.9 10.7 11.1 27.2 36.3 14.7
Black 10.1 24.8 15.5 24.8 22.8 12.1
Asian 5.6 10.5 19.3 33.3 27.2 9.6
Other 6.4 13.2 14.0 20.9 35.7 16.3
Employment (hrs)
0 49.1 11.7 11.7 29.1 33.8 13.6
10 22.5 11.1 11.6 28.8 35.4 13.1
20 19.5 13.4 12.4 21.7 36.4 16.2
30 5.7 10.4 17.4 24.3 34.8 13.0
40 3.2 24.2 12.1 16.7 24.2 22.7
Ag;
19 or younger 57.6 10.4 14.3 30.0 33.8 11.5
20 and older 42.5 14.7 9.3 22.7 35.2 18.1
Educational Level
Freshman 31.4 13.3 16.1 31.2 28.5 10.8
Sophomore 28.9 7.0 10.5 28.6 41.2 12.8
Junior 24.2 13.4 9.8 23.6 36.2 17.1
Senior 15.5 18.2 10.8 20.4 30.9 19.7
Living Situation
On-Campus Dorm 55.2 10.6 13.6 29.9 33.4 12.4
Off-Campus Apt/House 38.7 13.9 9.9 22.9 35.8 17.6
On-Campus Apt 3.7 15.8 11.8 23.7 39.5 9.2
Other 2.4 18.8 14.6 27.1 27.1 12.5
Internet Connection at Home
High-speed 88.9 10.0 11.5 27.6 36.0 15.1
Dialup 8.3 27.4 22.0 20.2 20.8 9.5
No Connection 2.8 40.4 5.3 26.3 24.6 . 3.5
Mg‘or in the College of'
Social Science 24.8 15.3 12.5 25.0 34.0 13.1
Business 12.0 10.2 12.7 27.5 34.0 15.6
Natural Science 11.7 13.1 11.0 27.8 33.3 14.8
Comm. Arts/Sciences 10.6 6.5 10.6 20.4 10.7 21.8
Engineering 6.9 7.1 7.9 27.1 37.1 20.7
Human Ecology 5.7 16.5 11.3 35.7 30.4 6.1
Undecided 10.1 9.7 14.6 30.1 35.4 10.2
Other 18.2 14.3 13.5 27.8 31.8 12.7
Base Percentage of Sample: 100 12.3 12.2 26.9 34.4 14.3

 

134

Table 10 provides the distribution of thirteen survey items representing the
primary dependent variables employed in this study. The response categories differed
across these questions, and so the table demarcates ﬁve groups in a non-speciﬁc manner:
zero, low, medium, high, and extreme amount of participation. Individuals were asked to
recall their participation since 2000, and the data indicate that a larger amount of people
were introduced to, and partook in, the behavior with each subsequent year. For instance,
almost half (47.8%) had not downloaded a single MP3 ﬁle in 2000. In 2001, 34.8% did
not participate at all, and in 2002, only 1/5th of those surveyed (21.9%) did not download
any music. This trend is mirrored in the average number of MP38 downloaded per month
over those three years as well. In 2001, 63% downloaded at least one MP3 each month; in
2002 that percentage increased to 78.1%, and by 2003 it was 80.8%.

Though data collection took place over the course of three months, it is notable
that since the beginning of 2003, only 14.6% had never downloaded a music ﬁle and that
59.4% could be classiﬁed as “high” or “extreme” participants in the phenomenon.
Finally, it is intriguing that a sizable 61.6% disclosed that they had obtained at least one
complete music album in MP3 format online. This underscores the fact that MP3 ﬁle
downloading is not just a song here and a song there, but Often involves calculated
acquisition of the contents of entire CD8, presumably for the purposes of using those as a

substitute for purchasing the album from a store.

135

Table 10. Distribution of Music Piracy Variables (N=2032)

 

 

Zero Low Med High Extreme
% % % % %
How many MP3ﬁles have you downloaded:
inthe last week? 41.9 21.1 12.8 9.8 14.3
in the last month? 27.5 32.1 17.4 11.7 11.2
since the beginning of 2003? 14.6 6.8 19.2 34.4 25.0
on average per month? 19.2 48.6 18.8 8.2 5.1
on average per week one year ago? 23.4 25.2 21.3 16.2 13.9
on average per month one year ago? 21.9 31.7 22.8 14.8 8.7
on average per week exactly two years ago? 37.4 25.6 16.8 10.6 9.5
on average per month two years ago? 37.0 29.8 17.0 9.8 6.4
in 2002? 21.9 7.8 24.7 37.0 8.7
in 2001? 34.8 10.0 24.7 24.9 5.7
in2000? 47.8 10.8 21.1 16.5 3.7
over the course of your life thus far? 12.3 12.2 26.9 34.4 14.3
How may complete albums have you downloaded? 38.4 28.6 12.5 8.6 11.9

 

Though not utilized in any bivariate or multivariate analyses, it is enlightening to

discuss some other ﬁndings related to MP3 participation gleaned from the survey. For

instance, of those who do spend time each week looking for MP3s, 35.1% spend more

than 1 hour engaged in that activity and 9.9% spend at least 3 hours. It was also found

that while 36.7% download 100% and upload 0% of their participation time, 41.3 upload

at least 25% of that time. Two-ﬁfths of the sample (41.8%) have created an audio CD

from MP3 ﬁles, 5.2% have made an MP3 ﬁle themselves, and 33.6% have done both. In

addition, 64.8% of respondents listen to MP38 on their computer and listen to them after

burning them to CD or transferring them to a portable MP3 player. Finally, 31.4%

disclosed they share their MP3 ﬁles with others, 2.8% stated that they sell them, and

4.4% stated that they do both activities.

Three questions related to ethical, moral, and legal perceptions of MP3

participation also provided some valuable insight into the minds and motives of those

surveyed. A sizable 91.2% stated that receiving or providing MP38 should not be illegal.

136

More speciﬁcally, 54% believed that MP3 participation is completely appropriate on
ethical, moral, and legal grounds. Over one-ﬁfth (21.4%) felt that it is unethical and/or
immoral but still appropriate behavior, while 6.3% believed that the fact it is unethical
and/or immoral renders it inappropriate. A respectable 14.5% stated that from their
perspective, downloading or uploading MP38 is illegal but ethical and/or moral and
therefore appropriate. Only 3.9% felt that the activity was unethical, immoral, and illegal
and accordingly inappropriate. By extension, 49.4% stated that they participate because
they do not believe it is illegal, while 25.4% participate even though they believe it is
illegal. Among those who refrain from participating, 8% do sobecause they believe it is
illegal, while 6.9% do so for other unspeciﬁed reasons. These ﬁgures not only highlight
mass ignorance as to the existence and applicability of copyright law to the digital
domain and intellectual property found online, but also demonstrate how perceptions of
what is illegal, unethical, or immoral do not necessarily constrain certain behaviors.
Bivariate Statistics

Bivariate statistics are used to measure the presence and strength of a relationship
between two variables. Before discussing the ﬁndings from correlation and ANOVA
tests, a summary of the construction of the independent and dependent variables is
necessary. As mentioned, CFA with Promax rotation was employed on the subsets of
strain, attitudinal and behavioral self-control,and social learning theory variables to
ensure that each group of observable measures was speciﬁcally representative of the
unobservable construct it sought to measure. The resultant continuous factor score
variables were utilized as independent variables throughout the multivariate analyses.

With regard to the thirteen dependent variables, a summary scale variable ranging from

137

13 to 65 (as each question has answer choices of 1, 2, 3, 4, or 5) was created to use in
bivariate AN OVA procedures and a continuous factor score variable was used in the
multivariate models.

With the resultant factor score variables, a bivariate correlation matrix was
created to discover the existence, direction, and strength of relationships among variables
(see Table 11). The strongest relationship between predictor and outcome variables was
that of differential reinforcement and overall music piracy (r = .445, p s .01), followed by
differential association (r=.332, p s .01). Both the attitudinal and behavioral measures of
self-control theory were also signiﬁcantly correlated with music piracy. That is, a lower
amount of self-control was linked with a higher amount of MP3 downloading. The
correlations, though, were of a comparatively small magnitude.

Indicative of possible multicollinearity issues and a difﬁculty to clearly
distinguish between the four tenets of social learning theory was the correlation between
differential association and differential reinforcement (.682). This is not too alarming,
due to the fact that Akers himself has stated that there are interrelationships among the
SLT components and that they are not conceptually distinct (Akers, 1977; Akers et al.,
1979). Speciﬁc to the MP3 phenomenon, having peer associations who participate
would positively reinforce one’s own participation, and the possibility of losing that peer
group — who provide valuable social, emotional, and tangible rewards (MP3 ﬁles) to an
individual - would serve as negative reinforcement and thereby perpetuate that person’s
involvement. Thus, the theoretical overlap between the two measures is obvious. Strictly
concerning the theoretical variables, notable ﬁndings included a signiﬁcant correlation

between the attitudinal and behavioral measures of self-control (r=.070, p s .01), and

138

between strain and behavioral self-control (r=.187, p s .01). Causal relationships will be

explored below with multivariate analyses.

139

ass Bonsai 85 v P

DEE 0632

>

 

.33. I “nuanceomﬁom 3385wa 2N
43;. scam. I 802539 .X
..m:. 20.. ..mmN. I «8:88: «X
32. ammo. .88. envoy I conuMOOmm<3ueoco§Q 5%
..me Lino. ammo. some. some. I Ebeooﬁomﬁﬂogﬁom ex
.53. .38.. same. moo. Lama; .58. I ~obeo0¢3mﬁ£pﬁ€< ex
25.. ammo: Re. ovo. sumo: KS. ooo. I Eebm an
18. ..Noor m8. v8. ..vwor sumo. «No... ammo. I 52o 8 ON ax
scam. ammo. m8. come. .03. 3o. N3. ...:N.- Re. I 032 «x
33. Poo. one. ..er ..wE. 5o. 33.- .32.- 8o: oNo. I 33.3 an
> ex an .x RX ex “N J" «X «X an

 

ameuuaﬁﬁas» a he“: 5:23.80 surge .: as;

140

Apalvsis of Vam

Analysis of Variance (ANOVA) allows for the interpretation of mean differences
in the dependent variable across categories of the independent variable (Bachman &
Paternoster, 1997). Comparisons between actual (identiﬁed) and expected variation in
category averages is expressed in the F statistic, which is the ratio of the between group
variation and the within group variation. If the signiﬁcance of F is 205, we can
conclude that the variance in the dependent variable is the same irrespective of the
independent variable. Conversely, if the signiﬁcance of F is < .05, we conclude that the
variance is different due to the inﬂuence of the predictor variable.

Ho: all DV population means are equal (u, = p; = H3)

HA: at least one DV population mean differs from others (pl at 112 at p3)

To reiterate, ANOVA allows researchers to determine the signiﬁcance of the
difference between means in music piracy participation across values of the predictors.
Utilizing a factor score dependent variable provides incomprehensible results when
comparing mean levels. As such, the summary scale dependent measure was used, with
possible values between 13 (indicating that the respondent had selected “0” for each of
the thirteen questions) and 65 (indicating that the respondent had selected the highest or
largest choice for each of the thirteen questions). Signiﬁcant results were found across
Sex, Race, Educational Level, type of Internet connection at home, and the college in

which the respondent’s major was housed (see Table 12).

141

Table 12: ANOVA results between demographic variables and
overall music piracy (additive measure)

 

 

 

 

 

Mean SD. F-ratio
§e_x_ 200.252***
Female 30.445 1 1.269
Male 37.906 12.406
M; 6807*“
Caucasian 33.973 12.22
African-American 30.175 13.19
Asian 33.781 11.18
Other 35.473 12.54
Employment (hrs) .124
0 33.657 12.221
10 33.515 11.292
20 34.015 13.034
30 33.522 12.973
40 33.672 15.585
Agg
19 or younger 33.456 11.918 .853
20-older 33.967 12.887
Educational Level 2.909“
Freshman 32.679 12.453
Sophomore 34.264 10.873
Junior 34.581 12.707
Senior 33.159 13.891
Living Situation .075
On-Campus Dorm 33.691 12.029
Off-Campus Apt/House 34.276 12.268
On-Campus Apt 33.599 12.696
Other 33.479 19.941
Internet Connectiomﬂome 29.755 "*
High-speed 34.402 12.022
Dialup 27.399 12.811
No Connection 29.018 14.617
ﬂior in the College of: 8164*“
Social Science 32.424 12.437
Business 35.299 13.140
Natural Science 32.781 12.112
Comm Arts/Sciences 37.708 12.515
Engineering 37.271 1 1.917
Human Ecology 31.209 10.414
Undecided 33.131 1 1.778
Other 32.221 12.338

 

*Correlation is signiﬁcant at the 0.05 level (2-tailed).

I""‘Correlation is signiﬁcant at the 0.01 level (2-tailed).
*"Correlation is signiﬁcant at the 0.001 level (2-tailed).

142

Speciﬁcally, the greatest participants were male, and those who were neither
White, Black, nor Asian (and thereby collapsed into the category of “Other”). The mean
for Blacks was signiﬁcantly different from Whites (mean difference = -3.798 p 5.01) and
those in the “Other” category (mean difference = -5.298 p $.01). Majors in the College
of Communication Arts and Sciences scored the highest mean on this additive dependent
variable, and that value was signiﬁcantly different from those belonging to the Colleges
of Human Ecology (mean difference = 6.499 p 5.01), Natural Science (mean difference
= 4.927 p $01), Social Science (mean difference = 5.284 p $.01), Undecided (mean
difference = 4.577 p 5.01), and Other (mean difference = 5.487 p 5.01) according to the
Bonferroni Post Hoe Test, which determines which of the ﬁve predictor categories differ
signiﬁcantly from the others in their power to inﬂuence the dependent variable in the
population. As expected, mean piracy levels were signiﬁcantly higher for those with
high-speed access to the Internet at their residence.

To test Hypothesis 1, AN OVA procedures for Proﬁciency in Internet Use and
Variety of Internet Use were run against the additive measure of overall music piracy.
Mean levels of participation were different depending on levels of proﬁciency and variety
to a signiﬁcant degree (see Table 13). The Bonferroni Post Hoc test indicated that those
who used the Internet for 0-5 items were signiﬁcantly different than those who employed
the Internet for 6 or more items. The mean difference of music pirating activities
between those who used the Internet with the highest variety (9 or more items) compared
to those who used it for 1-2 items was 11.786 (p 5.01). Similarly, the test revealed that
those who were proﬁcient in 0-2 items were signiﬁcantly different from those who were

proﬁcient in 3 or more items. The mean difference between the highest proﬁciency level

143

and 1-2 items - was 13.564 (p $01). These analyses support Hypothesis 1 by
empirically demonstrating that individuals with a higher skill level when it comes to
using online resources, and those who take greater advantage of all that the Internet has to
offer, pirate music to a larger extent than their counterparts.

Table 13. ANOVA: Proficiency and Variety of Internet Use and Overall Music Piracy (N=2032L

 

 

 

 

Mean S. D. F-ratio

Proﬁciency in Internet Use 41.647***

0 items 29.94 1 1.90

1-2 items 27.94 11.82

3-5 items 30.50 11.77

6-8 items 34.71 11.70

9 or more items 39.72 12.34
ﬁrrietv of Internet Use 66.28?"

0 items 26.05 12.28

1-2 items 28.63 11.61

3-5 items 31.54 11.34

6-8 items 36.09 1 1.66

9 or more items 42.20 12.20

 

***p < 0001 (two-tailed tests).

Eta squared (n2) measures how much total variation can be attributed to the
variation that occurs between groups, and was obtained by dividing the Between Groups
Sum of Squares by the Within Groups Sum of Squares. For Variety of Internet Use, 112
was .082, and for Proﬁciency of Internet Use n2 was .131. That is, 8.2% and 13.1% of
variation in music pirating activities can be explained by variety in, and proﬁciency of,.
Internet use. As an assessment of strength, both of these variables indicate a weak
relationship between the explanatory and criterion variables.

Multivariate Statistics
Ordinary Least Squares (OLS) Regression
As a preliminary test, OLS linear regression analyses was performed, regressing

the GST, SCT, and SLT factor score predictor variables on the factor score outcome

144

variable of overall music piracy (see Table 14). Also included were control variables of
Race, Sex, and Age. With regard to the controls, a dummy variable of “White” was
created since 77.9% of the sample belonged to this racial group. In an ideal model, all
NonWhites would not be analyzed together. However, since such a comparatively small
amount of respondents belonged to a NonWhite group, this stands out as one limitation of
the current research. This type of analyses assists in determining the degree to which a
linear combination of certain predictors signiﬁcantly explains variance in the criterion
variable of music piracy. To note, by comparing the magnitude of the (standardized) beta
coefﬁcients in all of these regression techniques, identiﬁcation of which independent
item(s) are more important than others within each model can also take place.

Table 14. OLS Reggession: IV Factor Scores on Overall Piracy Factor Score (N=2032)

 

 

Variables B Std. Error
Constant -.291 .046
White .042 .046
Male .493*** .039
20 or older .107" .038
Strain .006 .020
Attitudinal Self-ControlA 068*" .019
Behavioral Self-Control" .125*** .019
Differential Association .068" .026
Imitation 094*“ .020
Deﬁnitions -.006 .020
Differential Reinforcement 380*" .026
R2 0.302

Adjusted R2 0.299

 

‘p < 0.05; "p < 0.01; "*p < 0.001 (two-tailed tests).
AGreater magnitude in these factor scores indicates lower self-control

With regard to the demographic factors, males and those over age 20 were
signiﬁcantly more likely to pirate music than females and those 19 and younger,
respectively, controlling for the effects of all the other variables. Strain was not
signiﬁcantly related to music piracy, while both the attitudinal and behavioral measures

of self-control theory, were. This is a preliminary indication that those with lower self-

145

control pirate music to a great degree than those with higher levels of that latent
dispositional trait. Concerning the four components of social learning theory,
Differential Association, Imitation, and Differential Reinforcement were found to each be
signiﬁcant predictors of music piracy when controlling for all other predictors, while
Deﬁnitions was found to be unrelated. The most inﬂuential variable, based on the size of
the coefﬁcient, was Differential Reinforcement”. This attests to the importance of
perceived and actual rewards and punishments stemming from engaging in digital
intellectual property theft. It also points to possible policy solutions that may be
implemented in an effort to shift the cost-beneﬁt ratio in favor of the law, so that
individuals “think twice” about participation in the activity. Some suggestions will be
discussed in a forthcoming chapter.

Tests for multicollinearity corroborated the notion of possible overlap in the
operationalization and measurement of social learning theory tenets (Long, 1997).
Tolerance and Variance Inﬂation Factor statistics for Strain, Self-Control, Imitation, and
Deﬁnitions were unproblematic, but Differential Association and Differential
Reinforcement revealed a notable issue. Variance proportions for the two revealed very
high loadings on the same dimension. To note, a comprehensive analyses of the social
learning theory measures and the difﬁculties inherent in conceptualizing and measuring
them as distinct elements was provided earlier in this text.

Limitations of Ordinary Least Squares (OLS) Regression
Speciﬁc regression models with ordinal-level variables have gained popularity in

their use since the mid 19808, due to biases inherent with Ordinary Least Squares

 

35 The coefﬁcients of the theoretical variables in the regression analyses can be conrpared with each other

146

regression analyses. Analyses using outcome measures that violate the assumptions of
OLS regression tend to result in biased standard errors (heteroskedasticity), abnormally
distributed residuals, and probability predictions that are difﬁcult or impossible to
pragrnatically reconcile (Long, 1997:39). Additionally, the functional form requirement
of a linear model — that a one unit increase in x results in a constant increase in y,
irrespective of the value of x — is not practical when considering probabilities as
outcomes. The inﬂuence of x tends to wane as probabilities approach 0 or 1 (Long,
1997). Finally, OLS is based on a linear relationship between variables, and assumes that
the dependent items are at the interval-ratio level. Also, it presupposes that the distance
between the categories of the criterion variable is equal.

The dependent variables representing participation in music piracy in this
analyses are all ordinal-level measures. Each one, in fact, is a transformation of an
underlying continuous variable, where neatly observable Categories are speciﬁed and
ordered but where the distance between them is unspeciﬁed and unknown (Winship &
Mare, 1984). It cannot be assumed that the values of an ordinal variable are equidistant
from each other; as such, additional parameters that represent those unknown distances
(thresholds) should be included in order to more accurately assess the effect of the
predictors on the criterion variable (Maddala, 1983; McKelvey & Zavoina, 1975).
Logistic Regression

The premise of this study is to test the applicability of general strain, self-control,
and social learning theories on participation in digital music piracy. The thirteen

questions created as dependent measures, all intended to reveal a particular frequency of

 

because they are factor scores and thereby standardized.

147

pirating activity over certain time periods or durations, have ﬁve possible response
choices. These include zero as the ﬁrst choice, and then incrementally advance higher -
depending on the question asked. For example, for the question “How many total IVIP3S
have you downloaded over the course of your life thus far?”, the response set includes
“0”, “1-100”, “101-500”, “501-2000”, and “2000 or more”. In general, it appears that
this research should not attempt to differentiate between very speciﬁc levels of pirating
activity based on certain theoretical tenets. That is, it is not useful theoretically or
practically to determine that negative affect stemming from a particular theoretical
predictor is not signiﬁcantly related to individuals who have downloaded 1-100 MP38
over the course of their life, but is signiﬁcantly related to those who have downloaded
501-2000 MP38. It is more useful, rather, to ascertain how these theoretical predictors
differentiate those that do not pirate from those that do.

Accordingly, the decision was made to collapse the second, third, fourth, and ﬁfth
response choices into one category, and create a dichotomous variable for each of the
thirteen dependent measures so that 0 = no participation and 1 = participation in digital
music piracy. Thirteen separate logistic regression analyses were then run with the seven
theoretical factor score variables of Strain, Attitudinal Self-Control, Behavioral Self--
Control, Differential Association, Imitation, Deﬁnitions, and Differential Reinforcement.

The logit model allows for the concluSion that a unit change in x will result in a
logit change in y by B, holding all other variables constant. The dependent variable
employed is generally dichotomous and can take the value of 1 with a probability of
success 0. The independent variable is not constrained by requirements to be linearly

related, normally distributed, or to have equal variances in the groups. Indeed, a

148

nonlinear relationship between the explanatory and criterion variables is tested using the
logistic regression function, and the model is ﬁt appropriately via maximum likelihood
estimation. The logit transformation of the probability of success 0 can be represented

with the following equation:

6(a+ [3111+ lel + Bixi)

6 = 1 + e(r.!+ [3111 +0111 +Bixi)

 

Understanding a logit change is not intuitive; therefore, odds ratios (Exp(B)
values) provide a more logical way Of interpreting the parameters. If the odds ratio is
greater than 1, the Odds of the predicted value of y=1 based on a one unit change in x are
Exp(B) times larger; if less than one, the odds of the predicted value of y=l based on a
one unit change in x are Exp(B) times smaller (Long, 1997:80-81). To reiterate, the logit
model speciﬁcally speaks to the underlying latent variable that stimulates the crossing of
the threshold from 0 to 1 — from no participation in music piracy to participation in that
act, and assesses the odds of one or the other value of the binary dependent variable (i.e.,
have not vs. have pirated music) occurring due to the values of the predictors.

The ﬁrst model utilized the dependent variable of “How many MPBs ﬁles have
you downloaded in the last week?” (see Table 15). Consonant with ﬁndings from the
OLS regression model, being male increased the likelihood of having downloaded at least
one MP3 in the last week by 1.999, controlling for the other variables. Attitudinal Self-
Control was signiﬁcantly related to the dependent measure, and indicated that a lower
self-control increased the odds that the individual had downloaded at least one ﬁle in the
last week by 1.147. Differential association was also identiﬁed as a signiﬁcant predictor,
increasing the odds of MP3 participation by 1.242. The strongest predictor by far was

differential reinforcement, and its inﬂuence increased the odds of MP3 downloading by

149

1.714. The results of the other twelve analyses are provided in Tables 15, 16, and 17 and

are further discussed below.

150

2258 you vac—Esau Owen?“ no do; at $52 E9: 26m .v
~88 B magmas as 8% Beanie ea 22 .88. son .m
55:2: ma— oﬁ E pupae—Esau BE 32 .288 Box .~

$003 «as 05 E pupae—:38 BE 32 .89: Bow .—

_ohnoo.=om 63o— moﬁoaoﬁ 838 88a.“ 82: E 38:.qu 3329..
see 355% 83 v hrs... :3 v a: £3 v A.

 

 

 

 

 

 

an. an. NH. 8N. ..m shanghai

8m. am. on. as. 2 :26 a .80

$25 .6332 55.8 Snag 82:95 mod N-
R; as. rack. 52 so. 53. a? 48. IE. :5 NS. :38. assesses. 3.88:5
Ra. Re. 8o: 2:. m8. .. SN. o3. So. 5.- 23 so. NS. 22.35
as. Re. a3- 8R. go. 1.3.- as. so. m8; 2.3 m8. c8. SEE
83 as. 2.3:. So.— §. INS. Q3 e8. :3. 9.2 m8. :5. Seance... 322.55
2 2 m8. IE. 32 :5. .22. 2 2 a8. 2:. 33 So. 28. Seacoeam assuage
22 ”8. INS. 82 as. item. an: R8. :3. 2.2 so. :3. (338.2% Ease...
88. as. So: on: :b. 43. we: was. me. 83 c8. as. seem
3e. .2. :Sn- E. 2:. $2.- 3m. 2 _. 252? Sn. NS. tame; .020 s 8
52 N3. :38. 33 §. rise. 32 2 _. fee. 83 2:. :33. 2“:
N8. e2. .5- 83 E. as. E. E. «.8- m8. 2:. on..- 233
23 N2. 52 Rod ME. 83 EA 3. 32 RE 5. N3. essence
Sign... new a 8::— Hm .— EEG Hm a Sign Em A

v m N _

 

Aancnuzv 5:9: comm—:23 .3 ES .33 ..e «En—sue.— oeﬁm .55:— aaﬂ £83 .2: 5 pupae—:32. 8:.— BE .8. 5.383%— onlguea— .3 93:.

151

wow.“ 930.» 030 .3088 5:08 0waB>a n0 5 vac—Esau so.» RU .308 30: .m
mama E00.» 95 .3083 x003 0300.6 :0 E 983300 :0.» 20 .308 Box <-
wowa 80.» 0:0 .3088 5.88 0920.6 :0 E 32:33 so» Ev .308 Box 6

News .80» 0:0 .3038 x003 095.6 5 E vac—:30“. :0» Eu .308 Ram .m

35:00-20.»- 0033 330.63 000000 0900.“ 00.05 E 003838 080000<
Ans. Baa-2:0 52. v 0:... ”:3 v 0: m8.0 v 0..

 

 

 

 

 

we. 3:. 0%. EN. .0 0023:0002
m3. m2. 2:. m3. .0 :80 a. .80
5082 $0.32 $05: 084% 08.03: 03 -
33 E0. :38. on: as. :38“. 32 2.0. :33. 32 to. 5.80. 05883300 3085005
ea. 20. «8.- mg. 03. «8. as. 08. 02.- mg. 08. $0.- 88305
NS; 2°. .32. Q: 03. :02. 3: 000. «8. E: So. So. 805?
E 2 as. 02. 2.3 $0. 08. 82 N8. :80. $2 08. ..NNN. 89.683. 3085-05
32 m8. :3. 22 m8. Ema. :3 80. 2.0:. 2.2 as. :E. (3080-200 0222300
32 so. :3. 2 2 so. as. :2 N8. :03. $2 08. .52. 28080-200 563?.
20. MB. 08.- «a. «8. 08.- 30. m8. 0:.- mwm. so. ..Nﬁc 5.5
$3 2:. 3.0%. «a: 2:. 5:0. 23 g. .E. 32 a _. .08. 520 3 8
50 02. 2:5. 3 Z we. 55.. ~80 ca. 2.50:. E: «S. :35. 2.2
Go. 0.2. so; :a. ms. 80.. 32 NE. N2. 8: 5. 5. 333
on: S. 03. $3 5. 08. «n2 3. 23 32 02. $0. 8588
SEE dd 0 SEA.— .m-m ._ Even.— .Hm. ._ SEE-.— .Hm a
a .- 0 m

 

Amman-luau." 0.80» 0.5 an: cww .30» 0:0 52. .55.: ES 300.: 0mmu0>a E 603253.. 00E mm: ..8 5.00003”— 0uﬂw3 .3 05:.

152

sham v.35 a: So» be 858 on. 5>o Ego—:38 no» 022 mmmE :32 >58 >53 .2
muE—co c9553 so» 0%: EFF—8 mm: 5 £55.“ 89:: 829:8 .88 .39: 303 .m—
SSN a 32:38 3882 a; E. 86 22 .62: 3o: ._ _

SOON E vac—=38 3283a .5» Eu BE 22 >52: 302 .3

308 E vac—=33 33523 ac» Eu moE 9:2 bar: 302 .m

_ob:8..=um 530— 83065 358 58$ 30:. E 35::me 5.8%?

 

 

 

 

 

 

.332 B.§-§é :55 v as: :3. v at. ”3° v 9.

SM. as. E. SN. RN. Na exhoﬁumaz

m

5. :8. as. mi. Na. :25 a. xmu

82:95

:32: M382 $0.22 $.82 Mama: m3 N-

EuEuBQEoM

33 v8. :35. 83 8o. 52m. 33 as. 2:23. 23 NS. $3.... 23 ms. 33%. 5.553

5. 2:. :3? m2: 8°. 2°. wa. mg. mg; ca. 9.8. 36.. :w. So. :2? 255.55

3». Mac. .33.- :2 m8. :8: 82 NS. :33. em: o8. :3. :3 So. :o. more?

GOﬂMCOmm<

53 2:. 59:. 3: So. Re. 33 So. N8. 33 EV. N3. 22 :3. :E. Eaeuma

soﬁoozom

:2 So. 58. :2 so. 232. Q: o8. 358. 82 ﬁe. :5. :2 8o. :5. Eosﬁum

2225025

:2 So. Emma. £3 25. :3. NW: So. .32. 82 NS. 3.0:. a: so. :3. 35253

”8.— So. So. 83 5. 8o. 8a. so. 98.. ea. v8. 8a.- v3. So. So; 5.2%

ms. of. SN; 33 Mac. 35. 5: Mac. .232. 3.3 2:. ENE. 82 m2. tam. 32058

Ed 5.. ES: SS §. ism. 82 8o. 23%. was 2:. 33m. $2 5. :13. 222

82 ME. 8m. 8m. OS. 87 NE. 2a 89. 82 v2. 2:. en: ~:. «8. 223

83: 3. EN :2 o:. gm. 8”. ME. 3w- o3. me. So; we: 3;. 96. 2528
SEE A; a Erna :3 a SEQ .3 a 8:5 .m.m a SEQ .3 a

2 2 : 3 a

 

~8~uz 0.: ya... 5:... 332:2... ”22 E; .255: 82 32 ES .5... .82 .88 .82 5 332:2... m2: .2 :22: oz 2.: 3 .2 as:

153

The ﬁndings for the demographic variables were relatively consistent across the
models. While race (as measured by the “White” dummy variable) did not affect the
outcome variable, being male increased the likelihood of having pirated across all of the
models; the effects of age were signiﬁcant in 11 of 13 models but the direction of the
relationship alternated depending on the speciﬁc dependent variable. As such, nothing
conclusive can be said about the predictive capacity of age in differentiating those who
do not pirate music online from those who do.

Overall, strain was found to be signiﬁcantly but negatively related to MP3
participation in one of the thirteen models: “How many did you download in an average
week exactly one year ago?” One cannot take much stock in this ﬁnding, however, for
two reasons. First, it is counterintuitive and the presence of a signiﬁcant relationship was
not found in twelve models. Second, the general strain questions posed to respondents
asked them to reﬂect back only upon the last six months from the date of the survey and
reveal strain-inducing experienced in that time period. Strainful circumstances, and the
negative affective states they tend to generate, cannot be linked to participation in the
MP3 phenomenon due to time ordering. The signiﬁcance of the strain factor score for
this question, then, must be due to at least one spurious and unaccounted element. This
contradicts Hypotheses 2, which stated that each “general” theory is appropriately named
and has the capacity to explain variation in music piracy. General strain theory,
according to the analyses of this sample, is not a signiﬁcant predictor of the unlawful
activity in question.

When controlling for the predictive effects of all the other variables, Attitudinal

Self-Control was signiﬁcant in twelve of the thirteen models, and Behavioral Self-

154

Control was signiﬁcant in eleven. They were also all in the expected direction, and
indicated that lower self-control increases the odds of music pirating activities. When
considering the last model, “How many total MPBs have you downloaded over the course
of your life thus far?” and a response set of either “0” or “1 or more”, an increase in in
Attitudinal Self—Control increased the odds of belonging to the “1 or more” group by
1.334 (see Table 17). Similarly, an increase in Behavioral Self-Control increased the
odds of having pirated at least one MP3 by 1.221. This supports Hypotheses 2, which
predicted a positive signiﬁcant relationship between the explanatory and criterion
measures. In accordance with Hypotheses 3, the elements of social learning theory were
most strongly related to music piracy. Speciﬁcally, Differential Association was
signiﬁcantly and positively related to MP3 participation in eight of thirteen models, and
Differential Reinforcement was signiﬁcant in all thirteen.

The last model revealed ﬁndings that are perhaps most representative of music
pirating activity: “How many total MP38 have you downloaded over the course of your
life thus far?” It had the largest Nagelkerke R2 of any of the models (36.5% explained
variance) and had the best model ﬁt (-2 Log Likelihood = 1079.751). While general
strain was not found to be a signiﬁcant predictor, both measures of self-control were
signiﬁcant but negatively related. Each of the four elements of social learning were
signiﬁcantly related, but Imitation and Deﬁnitions both decreased the odds that the
respondent had pirated music. Differential Association, though, increased the odds by
2.182, while Differential Reinforcement increased the odds by 2.048, that the individual
had downloaded at least one MP3 over the course of his or her life. To note, Imitation

and Deﬁnitions were signiﬁcant in some models but the direction of their relationship

155

was not consistent. Furthermore their predictive effect was comparatively small. It is
difﬁcult to therefore make conclusive statements about their role in affecting
participation in music piracy.
Multinomial Logistic Regression

While it is not theoretically or practically useful to determine how the theory
factor scores explain variance at each of the ﬁve specific levels of MP3 participation
provided as possible response choices, it may be instructive to examine the
discriminating effect that the theories have on general levels of involvement in music
piracy. Therefore, the next stage of analyses involved taking the initial thirteen
dependent measures and recoding them to better understand the theoretical elements that
differentiated those individuals who engage in “low”, “medium”, and “high” amounts of
MP3 downloading. This also brought about more proportionate distributions among the
categories than the previous analyses in which dichotomous variables were created. To
construct these three-category variables, the choices of 1 or 2 were coded as “Low”, the
choices of 3 and 4 as “Medium”, and the choice of 5 as “High”. The decision to combine
1 - which equaled zero MPBs, and 2 - which equaled the lowest number of MP3s
downloaded among the possible responses for each question, was made because some
individuals who have merely “dabbled” in the activity and experimented with the

technology should arguably not be grouped with those who have downloaded a Medium

156

or High amount of digital music”. This then facilitated thirteen multinomial logistic
regression models.

Multinomial logistic regression is an extension of logistic regression model when
the nominal dependent measure has more than two levels (Hosmer & Lemeshow, 2000;
Long, 1997). For this analyses, music piracy is measured at three levels — “Low”,
“Medium”, and “High”, and those in the “Low” group are the reference category for the
calculation of odds ratios. MNL can be utilized to predict the log odds of one outcome as
compared to a baseline category by producing two logits simultaneously (where there are
three levels) (Hosmer & Lemeshow, 2000; Long, 1997). In this research, the two logits
will represent Low as compared to Medium, and Low as compared to High.

If Low is represented as J, the equation for the ith category (i.e., Medium) is as
follows:

Log (Pi/Pj) = 3.0 + Bi1X1+ Bax2 + ... + 13.,xp

The coefﬁcients that result must be interpreted as a change in log-odds resulting
from a one-unit change in the predictor variable. This is not very intuitive, and so odds-
ratios are used in this research. The odds of i rather than j occurring due to the inﬂuence
of x can be represented as follows:

Qm I n (Xi) = Exp(xiﬂli - 151])

The relationship of general strain, attitudinal self-control, behavioral self-control,

differential association, imitation, deﬁnitions, and differential reinforcement and the

 

3‘ Some might posit that a differentiation must be made between “experimentation” and “occasional” use.
It seems appropriate to consider those who have downloaded the lowest possible number of MP3s (apart
from zero) as experimenters, while those who have downloaded any larger number should at least be
considered as “occasional” participants.

157

likelihood of belonging to groups representing different amounts of MP3 downloading

are assessed and presented in Tables 18 through 30.

158

Table 18. Multinomial Logistic: How manLMP3s downloaded in the last week? (N=2032)

 

 

Medium High

Variables B S.E. Exp(B) B 8.13. Exp(B)
Constant -.979 .138 -2.028 .187
White -.047 .140 .954 .240 .179 1.272
Male 590*“ .118 1.803 1.107*" .146 3.026
20 or older -.766*** .120 .465 -.804*** .147 .447
Strain .067 .060 1.069 -.020 .073 .980
Attitudinal Self-ControlA .132“ .057 1.141 .031 .070 1.032
Behavioral Self-Control" .132“ .058 1.141 .228" .069 1.257
Differential Association -.020 .079 .981 .056 .097 1.058
Imitation .032 .059 1.033 289*" .068 1.335
Deﬁnitions .049 .060 1.050 .074 .071 1.077
Differential Reinforcement 538""Ml .087 1.713 .568'"'* .111 1.764
Chi-Square 319.735
-2 Log Likelihood 3360.114
Cox & Snell R2 .146

ﬂgelkerke R2 .174

 

"p < 0.05; "p < 0.01; “*p < 0.001 (two-tailed tests).
“Greater magnitude in these factor scores indicates lower self-control

Table 19. Multinomial Lﬁistic: How many MP3s downloaded in the last month? (N=2032)

 

 

Medium High

Variables B S.E. Exp(B) B S.E. Exp(B)
Constant -.712 .131 -2.164 .203
White .022 .132 1.022 -.016 .191 .984
Male 621*" .110 1.861 1288*“ .166 3.624
20 or older -.786"""" .111 .456 -.822*" .164 .440
Strain .116“ .056 1.123 .041 .081 1.042
Attitudinal Self-ControlA .077 .054 1.080 .137 .077 1.147
Behavioral Self-ControlA .120“ .054 1.127 .213” .076 1.237
Differential Association .078 .074 1.082 .166 .108 1.180
Imitation .043 .056 1.043 350*" .076 1.419
Deﬁnitions .002 .056 1 .002 .052 .079 1 .054
Differential Reinforcement 502*" .080 1.653 627"” .127 1.872
Chi-Square 354.988
-2 Log Likelihood 3355.270
Cox & Snell R2 .160

ﬂgelkerke R2 .191

 

‘p < 0.05; "p < 0.01; "*p < 0.001 (two-tailed tests).
“Greater magnitude in these factor scores indicates lower self-control

Note: Reference group is low lifetime music piracy (either zero participation, or the lowest possible amount
of participation). Medium equals the two largest penultimate MP3 totals in the response set, while High
equals the ultimate (and largest) response choice - the highest number of MP3s downloaded as available in
the response set.

159

Table 20. Multinomial Logistic: How many MP3s downloaded since the beginnimf 2003? (N=2032)

 

 

Medium ﬁgh

Variables B S.E. Exp(B) B S.E. Exp(B)
Constant 1.013 .151 -.159 .185
White .241 .149 1.272 .329 .181 1.390
Male .479‘" .137 1.615 1055*" .159 2.871
20 or older -.292* .128 .747 -.538"* .153 .584
Strain .014 .067 1.014 .063 .079 1.065
Attitudinal Self-Control“ 247*" .066 1.280 .291‘" .077 1.338
Behavioral Self-Control“ .1 13 .070 1.120 .371 **"‘ .079 1.449
Differential Association .470*** .087 1.601 438"” .103 1.549
Imitation -.124 .073 .883 .049 .082 1.050
Deﬁnitions -.151* .075 .860 -.006 .085 .994
Differential Reinforcement 034"" .082 1.886 1064"” .1 10 2.898
Chi-Square 564.715
-2 Log Likelihood 3543.040
Cox & Snell R2 .243

Milena: R2 .280

 

‘p < 0.05; ”p < 0.01; *"p < 0.001 (two-tailed tests).
“Greater magnitude in these factor scores indicates lower self-control

Table 21. Multinomial Logistic: How many MP3s do you download on average per month? (N=2032L

 

 

Medium Hijh

Variables B S.E. Epr) B S.E. Exp(B)
Constant -1.140 .134 -3.815 .321

White -.078 .133 .925 .041 .267 1.042
Male .717*"”" .111 2.048 1673"" .256 5.331
20 or older -.314"”" .111 .730 -.306 .223 .736
Strain .147“ .057 1.159 .057 .113 1.059
Attitudinal Self-Control“ .127“ .054 1.136 .127 .106 1.135
Behavioral Self-Control“ 288*" .054 1.333 .196 .106 1.217
Differential Association .062 .075 1.064 -.178 .141 .837
Imitation .129“ .055 1.137 .435""' .101 1.545
Deﬁnitions .103 .056 1.109 .173 .108 1.189
Differential Reinforcement 532*" .083 1.702 1075*" .199 2.931
Chi-Square 342.173

-2 Log Likelihood 3124.355

Cox & Snell R2 .155

Nagelkerke R2 .197

 

"p < 0.05; "p < 0.01; "*p < 0.001 (two-tailed tests).
“Greater magnitude in these factor scores indicates lower self-control

Note: Reference group is low lifetime music piracy (either zero participation, or the lowest possible amount
of participation). Medium equals the two largest penultimate MP3 totals in the response set, while High
equals the ultimate (and largest) response choice - the highest number of MP3s downloaded as available in
the response set.

160

Table 22. Multinomial Logistic: How many MP3s downloaded in an average week exactly one year
_ago? (N=2032) ‘

 

 

Medium High

Variables B S.E. EHKB) B 8.13. Exp(B)
Constant -.480 .125 -l.855 .186

White .011 .125 1.011 -.034 .178 .967
Male .436*** .105 1.547 778*" .149 2.176
20 or older .126 .103 1.135 .337“ .145 1.400
Strain .046 .053 1.047 .092 .075 1.097
Attitudinal Self-Control“ .126“ .051 1.134 .016 .072 1.017
Behavioral Self-Control“ 210*” .053 1.233 349"" .071 1.418
Differential Association .078 .070 1.081 .027 .099 1.028
Imitation .l 19* .054 1.127 .175* .072 1.191
Deﬁnitions .051 .054 1.052 .188“ .074 1.207
Differential Reinforcement .461 *** .073 1.586 .770'”" .1 16 2.160
Chi-Square 283.103

-2 Log Likelihood 3750.904

Cox & Snell R2 .130

Nagelkerke R2 .151

 

‘p < 0.05; "p < 0.01; "*p < 0.001 (two-tailed tests).
“Greater magnitude in these factor scores indicates lower self-control

Table 23. Multinomial Logistic: How many MP3s downloaded in an average month exactly one year
_ago? (N =2032)

 

 

Medium - High

Variables B S.E. Exp(B) B S.E. ExﬁBl
Constant -.668 .125 -2.815 .234
White .062 .124 1.064 .067 .212 1.069
Male 397*“ .103 1.488 1089"" .184 2.971
20 or older .206“ .102 1.229 .373“ .175 1.452
Strain .062 .053 1.064 .197“ .090 1.218
Attitudinal Self-Control“ .053 .050 1.054 .178" .085 1.195
Behavioral Self-Control“ .190*" .052 1.210 .370‘" .083 1.448
Differential Association .184IMI .070 1.201 -.035 .116 .966
Imitation .099 .052 1.105 304*” .085 1.356
Deﬁnitions .034 .05 3 1.035 .176“ .089 1.193
Differential Reinforcement .470*** .074 1.600 .837'""' .141 2.309
Chi-Square 314.337
-2 Log Likelihood 3398.457
Cox & Snell R2 .143

_l\l_agelkerke R2 .171

 

*p < 0.05; "p < 0.01; "*p < 0.001 (two-tailed tests).
“Greater magnitude in these factor scores indicates lower self-control

Note: Reference group is low lifetime music piracy (either zero participation, or the lowest possible amount
of participation). Medium equals the two largest penultimate MP3 totals in the response set, while High
equals the ultimate (and largest) response choice — the highest number of MP3s downloaded as available in
the response set.

161

Table 24. Multinomial Logistic: How many MP3s downloaded in an average week exactly two years
ago? (N =2032)

 

 

Medium HiLh

Variables B S.E. Exp(B) B S.E. Exp(B)
Constant -1.252 .133 -3. 120 .228
White -.051 .131 .950 -.203 .198 .816
Male 447*" .110 1.564 l.107"* .175 3.026
20 or older 560*" .108 1.751 1328*" .172 3.775
Strain .073 .056 1.075 .008 .086 1.008
Attitudinal Self-Control“ .108“ .053 1.1 14 .058 .083 1.060
Behavioral Self-Control“ .171" .054 1.186 .354‘" .080 1.425
Differential Association .022 .074 1.022 -.068 .112 .934
Imitation .139“ .055 1.149 .276" .082 1.318
Deﬁnitions -.013 .056 .987 .058 .086 1.060
Differential Reinforcement .487*** .079 1.627 .635*** .126 1.888
Chi-Square 298.745
-2 Log Likelihood 3236.440 ‘
Cox & Snell R2 .137

_1\1_age1kerke R2 .166

 

‘p < 0.05; I""‘p < 0.01; “*p < 0.001 (two-tailed tests).
“Greater magnitude in these factor scores indicates lower self-control

Table 25. Multinomial Logistic: How many MP3s downloaded in an average month exactly two years
go? (N=2032)

 

 

Medium High

Variables B S.E. Exp(B) B S.E. Exp(B)
Constant -l.514 .137 -4.415 .315
White .010 .132 1.010 .131 .252 1.140
Male 573*" .110 1.773 1516*“ .227 4.554
20 or older 684*" .109 1.982 1.470""‘ .211 4.348
Strain .052 .056 1.054 -.002 .104 .998
Attitudinal Self-Control“ .111“ .054 1.117 .165 .099 1.179
Behavioral Self-Control“ .107* .054 1.1 12 358*" .095 1.430
Differential Association .017 .074 1.018 -.174 .132 .840
Imitation .136“ .055 1.146 377"" .096 1.458
Deﬁnitions -.062 .056 .940 .042 .101 1.043
Differential Reinforcement .545 *** .081 1 .724 1026*" .172 2.791
Chi-Square 340.118
-2 Log Likelihood 2903.048
Cox & Sne11R2 .154

ﬂgelkerke R2 .193

 

‘p < 0.05; I"“p < 0.01; "*p < 0.001 (two-tailed tests).
“Greater magnitude in these factor scores indicates lower self-control

Note: Reference group is low lifetime music piracy (either zero participation, or the lowest possible amount
of participation). Medium equals the two largest penultimate MP3 totals in the response set, while High
equals the ultimate (and largest) response choice - the highest number of MP3s downloaded as available in

the response set.

162

Table 26. Multinomial Logistic: How many MP3s did you ﬁrsonally download in 2002? (N=2032)

 

 

Medium mh
Variables B S.E. Exp(B) B S.E. Exp(B)
Constant .206 .127 -2.568 .252
White 316* .128 1.372 .425 .229 1.529
Male .672"* .1 16 1.957 1605*" .200 4.976
20 or older .352" .112 1.422 .460“ .187 1.584
Strain -.023 .057 .977 .015 .096 1.015
Attitudinal Self-Control“ .1 18" .055 1.125 .185 * .092 1.204
Behavioral Self-Control“ .164" .058 1.178 364*" .091 1.440
Differential Association .256" .074 1.291 .100 .124 1.105
Imitation -.062 .060 .940 . 177 .093 l . 194
Deﬁnitions -.104 .061 .901 -.046 .096 .955
Differential Reinforcement .590'""'I .073 1.804 .929*** .143 2.531
Chi-Square 404.036
-2 Log Likelihood 3136.235
Cox & Snell R2 .180
Nagelkerke R2 .219

 

‘p < 0.05; "p < 0.01; ""p < 0.001 (two-tailed tests).
“Greater magnitude in these factor scores indicates lower self-control

Table 27. Multinomial Logistic: How many MP3s did you personally download in 2001? (N=2032)

 

 

Medium HigL

Variables B S.E. Exp(B) B S.E. Exp(B)
Constant -.645 .123 4.108 .325
White .316‘ .122 1.371 .272 .263 1.312
Male 702*“ .104 2.017 1.764‘" .242 5.835
20 or older 579“" .103 1.784 1399*" .221 4.051
Strain -.086 .052 .918 .002 .111 1.002
Attitudinal Self-Control“ .109“ .050 1.1 15 .142 .106 1.152
Behavioral Self-Control“ .206" .052 1.229 .207 .106 1.230
Differential Association .067 .069 1.069 -.064 .141 .938
Imitation .089 .053 1.093 378"" .103 1.459
Deﬁnitions -.084 .053 .919 -.165 .107 .848
Differential Reinforcement 607“" .072 1.834 1.127""‘ .179 3.087
Chi-Square 403.304
-2 Log Likelihood 3133.201
Cox & Snell R2 .180

ﬁagelkerke R2 .218

 

*p < 0.05; “p < 0.01; "*p < 0.001 (two-tailed tests).
“Greater magnitude in these factor scores indicates lower self-control

Note: Reference group is low lifetime music piracy (either zero participation, or the lowest possible amount
of participation). Medium equals the two largest penultimate MP3 totals in the response set, while High
equals the ultimate (and largest) response choice — the highest number of MP3s downloaded as available in

the response set.

163

Table 28. Multinomial Logistic: How many MP3s did you personally download in 2000? (N=2032)

 

 

Medium High

Variables B S.E. Exp(B) B S.E. Exp(B)
Constant -.978 .124 —4.215 .361 ‘
White -.007 .123 .994 . 100 .303 1.105
Male 852*” .102 2.345 1493*" .275 4.448
20 or older .309" .101 1.362 .860" .251 2.363
Strain -.057 .052 .945 -.004 .128 .996
Attitudinal Self-Control“ .092 .050 1.097 .206 . 121 1.228
Behavioral Self-Control“ .164" .051 l . 178 .237* . 120 1.267
Differential Association .083 .069 1.087 -.238 .157 .788
Imitation .199*** .052 1.221 458*“ .122 1.580
Deﬁnitions -.102 .052 .903 -.157 .126 .855
Differential Reinforcement .485*** .073 1.624 .829*** .189 2.291
Chi-Square 298.323
-2 Log Likelihood 2967.455
Cox & Snell R2 .137

ﬂgelkerke R2 .171

 

‘p < 0.05; ”p < 0.01; *"p < 0.001 (two-tailed tests).
“Greater magnitude in these factor scores indicates lower selfecontrol

Table 29. Multinomial Logistic: How many total complete music albums in MP3 format have you
obtained online? (N=2032)

 

 

Medium Hm
Variables B S.E. Exp(B) B S.E. Exp(B)
Constant -1.169 .139 -2.224 .189
White -.206 .138 .814 -.205 .178 .814
Male 597*" .118 1.816 956*" .154 2.601
20 or older -.281"‘ .118 .755 .084 .149 1.088
Strain .053 .060 1.054 .007 .077 1.007
Attitudinal Self-Control“ .089 .057 1.093 .146“ .073 1.157
Behavioral Self-Control“ .197" .057 1.218 .188“ .073 1.207
Differential Association -.074 .079 .929 —.025 .101 .976
Imitation 222*" .059 1.249 328*" .072 1.388
Deﬁnitions .078 .061 1.081 .059 .075 1.061
Differential Reinforcement 338“" .082 1.401 .625*** .1 18 1.868
Chi-Square 223.921
-2 Log Likelihood 3230.245
Cox & Snell R2 .104
Nagelkerke R2 .128

 

‘p < 0.05; ”p < 0.01; "*p < 0.001 (two-tailed tests).
“Greater magnitude in these factor scores indicates lower self-control

Note: Reference group is low lifetime music piracy (either zero participation, or the lowest possible amount
of participation). Medium equals the two largest penultimate MP3 totals in the response set, while High
equals the ultimate (and largest) response choice - the highest number of MP3s downloaded as available in
the response set.

164

Table 30. Multinomial Logistic: How many total MP3s have you downloaded over the course of your

life thus far? (N=2032)

 

 

Medium High

Variables B S.E. Exp(B) B S.E. Exp(B)
Constant .496 .135 -2.073 .224 be
White .456“ .137 1.577 .511“ .205 1.668
Male 677*" .128 1.969 1657*” .180 5.246
20 or older .148 .121 1.160 .729"* .170 2.073
Strain .030 .062 1.031 .050 .088 1.051
Attitudinal Self-Control“ .183" .060 1.201 .265" .085 1.303
Behavioral Self-Control“ .205 ** .064 1.228 .391 *” .086 1.479
Differential Association .414"* .080 1.512 .344" .1 14 1.41 1
Imitation -.072 .067 .930 .162 .087 1.176
Deﬁnitions -.O36 .068 .964 -.061 .090 .940
Differential Reinforcement 551*" .077 1.736 1200*“ . 134 3.320
Chi-Square 549.928
-2 Log Likelihood 3196.982 _
Cox & Snell R2 .237

_liagelkerke R2 .282

 

‘p < 0.05; “p < 0.01; "*p < 0.001 (two-tailed tests).
“Greater magnitude in these factor scores indicates lower self-control

Note: Reference group is low lifetime music piracy (either zero participation, or the lowest possible amount
of participation). Medium equals the two largest penultimate MP3 totals in the response set, while High
equals the ultimate (and largest) response choice — the highest number of MP3s downloaded as available in
the response set.

165

To begin, the Chi-Square was signiﬁcant across each model and indicated that the
group of independent variables is signiﬁcantly linked to the respective dependent
variables in the analyses. Conceming the theoretical measures, general strain was
signiﬁcantly and positively related in only three of the thirteen models. The dependent
variables in these models were: “How many MP3 ﬁles have you downloaded in the last
month?”, “How many MP3s do you, on average, download per month?”, and how many
MP3s did you download in an average month exactly one year ago?”. Nonetheless, the
strength of general strain in predicting music piracy was comparatively unimpressive, and
so overall not much can be said about the capacity of that theoretical factor to explain
MP3 participation.

When considering the number of MP3 ﬁles were downloaded by the respondent
each week, having low attitudinal self-control increased the odds of an individual’s
participation in “Medium” as compared to “Low” amounts of music piracy, while having
low behavioral self-control increased the odds of MP3 participation in both “Medium” as
compared to “Low”, and “High” as compared to “Low” amounts. Imitation signiﬁcantly
differentiated those who downloaded “Low” amounts of MP3s from those who
downloaded “High” amounts, increasing the odds of belonging to the “High” group by
1.335. Differential Reinforcement was the strongest predictor, and had the greatest effect
when evaluating its inﬂuence on the High group; that theoretical tenet increased the odds
of belonging to the “Medium” group by 1.703 and increased the odds of belonging to the
“High” group by 1.764. This trend was consistently found throughout all of the

multinomial logistic analyses.

166

Surprisingly, differential association - a predictor found to be Signiﬁcantly related
to music piracy in the previous binary logistic analyses, was not in the following models:
“How many MP3 ﬁles downloaded in the last week?”; “How many MP3 ﬁles
downloaded in the last month?”; “How many MP3s do you, on average, download per
month?”; “How many did you download in an average week exactly one year ago?”;
“How many did you download in an average week exactly two years ago”?; “How many
did you download in an average month exactly two years ago?”; “How many MP3 ﬁles
did you personally download in 2001?”; “How many MP3 ﬁles did you personally
download in 2000?”; and “How many total complete music albums in MPB format have
you obtained online?”. It was signiﬁcantly related for “How many MP3 ﬁles
downloaded since the beginning of 2003?” and for “How many total MP3s have you
downloaded over the course of your life thus far?” in both the “Medium” and “High”
categories, and for “How many did you download in an average month exactly one year
ago?” and “How many MP3 ﬁles did you personally download in 2002?” in the
“Medium” category.

The last model - “How many total MP3s have you downloaded over the course of
your life thus far?” - indicated the best ﬁt based on the Chi-Square statistic (see Table
30). Being male increased one’s likelihood of belonging to the “Medium” and “High”
group, as compared to the “Low” group, while being age 20 or older increased one’s
likelihood of belonging to the “High” group as compared to the “Low” group.
Attitudinal and behavioral self-control were both Signiﬁcantly and positively related.
That is, lower self-control (measured either through attitudinal or behavioral variables)

increased the odds that an individual participated in “Medium” or “High” amounts of

167

MP3 downloading. Interestingly, behavioral self-control had a stronger inﬂuence than
did attitudinal self-control. With regard to the social learning theory variables,
Differential Association was a Signiﬁcant predictor and increased the odds of belonging
to the “Medium” group by 1.512 and the odds of belonging to the “High” group by 1.411.
Differential Reinforcement was also Signiﬁcant and increased the odds of belonging to
the “Medium” group by 1.736 and the “High” group by 3.320.

Interestingly, Imitation was Signiﬁcantly and positively related in 10 of the 13
models. Unlike the ﬁndings from the binary logistic analyses, the multinomial models
consistently indicated that that Imitation Signiﬁcantly increased the likelihood of
belonging to the “High” group as compared to the “Low” group (evident in 10 models)
and between the “Medium” group and the “Low” group (evident in 6 models). The role
of models providing sources of behavior, attitudes, techniques, and beliefs to observe and
emulate on the Internet does appear to differentiate music pirates among their quantity of
music piracy participation, but not between non-pirates and pirates.

Deﬁnitions was Si gniﬁcant in only three of the thirteen models, but the direction
of its inﬂuence was not constant and the coefﬁcients were quite small. Accordingly, not
much can be said about its inﬂuence - and this ﬁnding resonates across all of the

statistical models in this research.

168

CHAPTER 5: DISCUSSION

The preceding analyses has clariﬁed the relationship between the three general
criminological theories and digital music piracy. Two hypotheses were supported; one
was rejected. A host of questions stemming from the ﬁndings merit discussion in an
attempt to crystallize the statistical relationships into concrete knowledge. To begin,
what are some possible reasons as to why GST was not a signiﬁcant predictor of music
piracy? On the surface, this makes intuitive sense. Individuals who experience strain
from certain negative life experiences are probably not more likely to venture online to
download intellectual property from Internet sources. Though asserted as a “general”
theory and purportedly universal in its explanatory power to all forms of crime and
deviance, this was not the case when examining frequency of online music pirating
behavior as the outcome variable in this study. This could be for a variety of reasons.
GST may not be extensible to cyberspace and to research that commingles real-life and
the online realm. It also may be that negative affect ensuing from strain is reconciled
through behaviors or actions in real-life, rather than reserved for manifestation when the
person is in front of a computer connected to the Internet.

In addition, these traditional measures of strain involving the experience of real-
life stressors and misfortunes may be inappropriate for analyzing this particular
phenomenon. Strain perhaps needs to be speciﬁcally measured as resulting from the
inability to obtain or purchase the desired commodity of music, rather than as resulting
from problems in one’s day-to-day living experience. This smacks of initial
conceptualizations of Robert Merton’s (193 8) strain. Merton argued that persons are

limited in their access to socially approved goals and the means to achieve those goals.

169

Music is valued by the vast majority of Americans, and meets a variety of psychological,
emotional, and social needs. Possessing certain songs or recordings is a largely universal
desire - indicative of the fact that such a goal is strongly emphasized in American culture.
The legitimate method to obtain these tracks is via purchase from a retail establishment,
which requires ﬁnancial resources. Accordingly, restricted access to music can lead to
feelings of strain that are compounded by the strain resulting from a dearth of funds
requisite to resolving the initial strain.

Five modes of adaptation were proffered by Merton (193 8) in order to resolve the
resultant dissonance: conformity, innovation, ritualism, retreatism, and rebellion.
Innovation concerns the acceptance of culturally- and socially-promulgated goals but
have been thwarted in their attempts to achieve them in a legitimate fashion, thereby
inducing them to design another (unethical or unlawﬁil) method to do so (Merton, 1938).
The criminological literature base tends to utilize generic measures of strain that are not
Speciﬁc to the dependent variable, and the current research followed that trend.
However, perhaps greater consideration of the uniqueness of the crime is necessary when
conceptualizing and operationalizing predictors. As such, four questions included in the
survey but not initially designated as strain measures may be useful in more precisely
(but not perfectly) measuring strain in general as a predictive inﬂuence. These included:
“I would be more likely to download/upload MP3s if I could not afford the purchase
price of the music on CD?”; “I would be more likely to download/upload MP3s if I
needed the music and wouldn’t be able to obtain it any other way?”; “I would be more
likely to upload/download MP3s because I can’t afford to waste money on a music CD

that might only have 1 or 2 good songs?”; “I would be more likely to upload/download

170

MP3s because without the ability to evaluate the music, I will not be able to determine if
I really want to purchase it on CD?”. These focus on strain induced by ﬁnancial
limitations — likely the most prominent type among university Students.

To test the utility of these more speciﬁc measures, a factor score variable was
created from these items, and had a solid reliability alpha value (.766) and factor loadings
over .7. This variable was run through the same logistic and multinomial regression
models, and while stronger support was found for general strain theory as a positively
related predictor of music piracy, variability and mixed results were still found to a small
extent across the models. As such, a relationship between general strain and MP3
participation is identiﬁable but not consistent among the various measures of the
phenomenon. General strain, then, must be operationalized in a more informed manner,
based on the inconclusive results of the current analyses.

Finally, copyright infringement through MP3s may be considered by some as a
white-collar crime, simply due to its technologically-advanced and comparatively
sophisticated nature. Under that assumption, it is useful to compare music piracy with
another white-collar crime - embezzlement - since the latter involves traditional strain to
some degree. Donald Cressey (1953) argued that embezzlement occurred when an
individual had an unsharable ﬁnancial problem and could use his or her position of trust
to " get back in the black", could justify the behavior as acceptable and necessary, and was
presented with the right opportunity. Music piracy does not seem to result from the
"unsharable ﬁnancial problem" of not being able to afford a music CD, even though
rationalizations are used and an attractive opportunity is presented. Indeed, obtaining

digital music to satisfy a personal desire cannot be equated to the misappropriation of

171

funds to address a ﬁnancial need for survival. Strain, then, may only play a determinative
role in wrongdoing that is related to a signiﬁcant and life-impacting necessity.

The second question arising from the results is as follows: Why was attitudinal
and behavioral self-control strongly related to digital music piracy? Ceteris paribus, one
would think that an individual’s inability to regulate and constrain his or her behavior in
the real world would translate into an inability to refrain from participation in
questionable online activities when presented with an attractive opportunity. This was
depicted in the analyses, which generally showed that lower self-control - both measured
behaviorally and attitudinally - increased participation in the phenomenon. Interestingly,
mean scores created from both types of self-control measures indicated that the
distribution of responses were negatively skewed. This is inconsistent with general
estimations of self-control among university students; one might believe that because of
their enrolhnent and participation in higher learning, they possess more self-control than
their counterparts who are not enrolled in a university. Based on the measures employed
in this study, however, the attitudes and behaviors of students seem to decidedly indicate
a lack of self-control — which then seems to manifest itself in music piracy to some
degree. Why, then, does this lack of self-control result in pirating music online, rather
than shoplifting it from a retail store? This question warrants some discussion.

Gottfredson and Hirschi (1990) remark that criminal acts require little skill or
planning. It seems, though, that MP3 participation involves methodical actions which -
while quickly routinized - still necessitate a certain amount of skill and planning.
Conversely, shoplifting a music CD from a retail establishment involves only the

Spontaneous seizing of an attractive opportunity. Prior to these ﬁndings, one might

172

conclude that the former would be favored by those with more self-control, and the latter
by those with less self-control. The sample involved in this study, however, has revealed
a disproportionate amount of low self-control (based on the response choices in the
survey items). Furthermore, it is likely that all members are attracted by the possibility of
acquiring music at no cost. Music is an extremely valued commodity for the emotional,
psychological, and relational beneﬁts it provides among the college-aged population, and
this demographic group typically has little discretionary income to purchase CDS at their
whim.

If one assumes that individuals are equally motivated to obtain free music,
perhaps one of the primary variables that distinguish those of the college age-group (17-
24) who would engage in larceny from a store from those who would partake in digital
music piracy online is their level of low self-control. While this particular sample of
university students indicated a disproportionate amount of low self-control, it is possible
that they still have more self-control in general than those who are not in school but are
still of the same age. As such, those enrolled in higher education — though lacking in
self-control because of their age - are more inclined to learn the methodical actions to
download music from the Internet for free than to shoplift it because they do have more
self-control than their non-college-going peers.

The third question is relevant to the strongest theoretical predictor: why does
Differential Reinforcement increase the likelihood and amount of music piracy among
respondents? Evidently, the consequences that result ﬁ'om the behavior are very potent
inﬂuences in its perpetuation and perhaps even its augmentation. It is highly likely that

the beneﬁcial outcome of receiving valued goods at no cost (MP3 ﬁles) increases

173

participation in downloading. Similarly, the relative lack of punitive repercussions in the
form of detection, apprehension, and penalty promotes the behavior as well. Intrinsic and
extrinsic rewards, and the comparative lack of punishment, commingle to create the
predominant contributive element in music piracy. Sanctions do not seem to effectuate
much in the way of deterrence; this will be discussed later in the text. Policy intended to
reduce the frequency and prevalence of this phenomenon must somehow decrease the
rewards that accompany participation.

The fourth question concerns the reasons behind the role of Differential
Association as the second strongest predictor. As previously stated, MP3 participation is
a wildly social event online, and techniques, motives, rationalizations, and beliefs in
support of the activity are taught through peer associations. Indeed, the prevailing
atmosphere on the Internet is incontrovertibly pro-MP3, and thousands of venues exist
where computer-mediated communication between individuals can encourage and foster
the behavior. As increasing amounts of new participants internalize the attitudes and
actions that contribute to music piracy, they then fulﬁll an instructive role to others who
might be fresh to the “scene”. This consequently and continually enlarges the circle of
individuals who partake in the phenomenon. Differential association was Signiﬁcant in
eight of the thirteen binary logistic models, but only three of thirteen multinomial logistic
models. That is, the tenet largely differentiates between nonpirates and pirates, but not
among the amount of music piracy committed by pirates. This is consonant with the role
that differential association is believed to play in introducing individuals to wrongdoing,
and the fact that it is the ﬁrst occurring component of social learning theory (Akers,

1979). Peer associations, then, introduce people to the behavior and reinforce its

174

appropriateness. Once a person becomes acquainted with MP3s, though, it appears that
the inﬂuence of this tenet attenuates.

The ﬁfth question is related to the contributory role of Imitation in differentiating
music pirates in their amount of MP3 participation (e.g., Low from Medium, Low from
High), but not differentiating persons who pirate from those who do not. This is contrary
to the ﬁnding with differential association, and makes intuitive sense. Once individuals
are introduced to the activity, their frequency of participation would presumably increase
upon observing and learning other methods of obtaining MP3 ﬁles, and other sources and
venues on the Internet from where they might be downloaded. Thus, people do not begin
to pirate music due to the modeling of someone else’s pirating activities - that appears to
stem from differential association. Rather, a person’s initial involvement is deepened
after emulating the actions of those who they encounter in cyberspace. While this
corroborates the general chronological ordering of the SLT elements, it is contrary to the
ﬁndings in Akers et al. (1979), who presumed that imitation has its greatest effect in the
initial phases of participation rather than in maintaining the activity.

The Sixth and ﬁnal question is: why were Deﬁnitions largely irrelevant in
explaining any variation in music pirating activity? To remind the reader, Deﬁnitions are
evaluative criteria designating behaviors as good or bad and thus qualifying them as
acceptable or appropriate (Akers et al., 1979). They are similar to Sykes and Matza’s
(195 7) techniques of neutralization, but are shared by a group and are a byproduct of
social interaction. It can be speculated that deﬁnitions are not necessary to deﬁne MP3
participation as “right” and “justiﬁable” Simply because of its ubiquity. The presence of

hundreds of thousands of individuals online might serve to preempt any questions as to

175

the acceptability of the activity. That is, if so many people do it, how could it be wrong?
No deﬁnitions, then, must be actively embraced in order to resolve or overcome any
qualms or misgivings about participation. One might participate simply because MP3

ﬁles are available, easily obtainable, and provide great beneﬁts with little to no cost.

176

CHAPTER 6: LIMITATIONS

Limitations are inherent in any social science research endeavor, particularly
because the capricious nature of human behavior, and also because of the virtually
unlimited number of inﬂuences that may play predictive or determinate roles. General
strain, self-control, and social learning theories cannot perfectly explain criminal activity
and scholars have pointed out some vulnerabilities in their constitution. Furthermore,
methodological choices related to survey instrument design and data collection
sometimes lead to over- or underspeciﬁcation of the signiﬁcant ﬁndings. These do not
invalidate the results, but provide a caveat for cautious interpretation and generalization.
In the following text, possible and actual limitations of the general theories and the
research methods of the current study are discussed.
Limitations of General Strain Theory

Some weaknesses have been identiﬁed with general strain theory. Broidy (2001)
discovered that the Speciﬁc negative emotion experienced largely determines the
legitimacy of the coping mechanism employed. That is, anger led to illegitimate
outcomes, while other negative emotions (e.g., frustration, disappointment, sadness,
loneliness)37 were associated with legitimate coping. In the current study, strain-induced
frustration was hypothesized to impel individuals towards pirating music in order to
obtain relief. However, the speciﬁc negative emotion that respondents might experience
when their desire to achieve a positively valued goal is thwarted was not determined
through any items in the questionnaire. They might have experienced frustration, or

anger, or depression, or a host of other subtly different feelings. In line with Broidy’s

 

37 The variables measuring negative emotions were combined into a scale in Broidy’s (2001) research.

177

(2001) ﬁndings, if the strain that followed one’s inability to obtain music in a legal
manner was any negative emotion other than anger, legitimate coping might have taken
place. Additional research would have to corroborate Broidy’s (2001) ﬁndings before
any blanket statements can be made about the relationship between different types of
negative affect and prosocial or antisocial behavior. Nonetheless, the failure to identify
the speciﬁc emotional response(s) experienced by individuals who feel strained further
blurs the already complex relationship between Agnew’s primary predictive concept and
criminal outcomes.

Though individuals may be similar in their strainful experiences, many do not
engage in deviance, and social science scholars have attempted to point to the factors that
condition the relationship between negative affect and wrongdoing. Morality, self-
efﬁcacy, personality, coping resources, social support, social control, and peer
associations all moderate the proclivity towards delinquent behavior following strain
(Agnew, 1997; Agnew et al., 2002; Agnew & White, 1992). Questions to determine the
presence or absence of many of these factors were not included in the survey instrument
due to length constraints. They may have proven useful to more accurately tease out any
conditioning effects and to lend greater insight into the causal chain. GST also argues
that delinquency as a response to strain can be escapist (e.g., self-destructive behaviors),
instrumental (property crimes), or retaliatory (violent crimes) (Agnew, 1992). It is
conceivable that strain amongst a college-aged population may ensue in escapist
behaviors such as drinking or even drug use, rather than digital intellectual property theft.

Furthermore, GST contends that strain, negative coping, and delinquency are

proximately related in time. Agnew (1992) has indicated that strain and delinquency

178

should occur concurrently, and that states that stressful events occurring more than three
months ago are largely unrelated to the dependent variable in question. Nonetheless,
research following the initial assertion of the theory has demonstrated that delinquency
following strain also occurs often (Agnew & White, 1992; Paternoster & Mazerolle,
1994). In the current research, individuals were asked to reﬂect on strainﬁrl experiences
over the past six months, and to indicate their current and past participation with MPBS
over a three-year period. It may be that any attempt to capture how MP3 participation is
affected by strain through such retrospective questions will result in null or conservative
ﬁndings because of the requisite temporal contingency.
Limitations of Self-Control Theory

Gottfredson and Hirschi’s general theory of crime, which posits that crime and
other Similar behaviors can be explained by a lack of self-control originating from
ineffective parenting in early childhood, has been faulted for many reasons. Critics have
pointed to the inherent tautological nature of its explanation, its underdeveloped
conceptualization of the role of opportunity and reward in the etiology of crime, and its
superﬁcial portrayal of the nature of the behavior that the theory is designed to explain
(e.g., Sellers, 1999).

Also, Gottﬁedson and Hirschi (1990) address the importance of socialization
(e.g., familial) in rearing individuals in the way they Should go. Even if such social
institutions effectively develop a bond to convention among persons and a consequent
ability to repress drives to ﬁrlﬁll immediate self-interest, it is plausible that they may still
participate in deviance. For example, neutralization theory (Sykes & Matza, 1957) holds

that individuals can still maintain adherence to dominant mores of society but temporarily

179

 

place them in abeyance in order to facilitate commission of a crime, and also to prevent
any imputation of a deviant identity onto themselves. Thus, while self-control is of
crucial import, there are other factors that necessitate recognition. Additionally, it is
possible that an upstanding individual with deep-seated law-abiding principles could
potentially be worn down and induced to engage in criminality if immersed in a culture
where such activity was standard and esteemed practice.

Finally, the deﬁnition of crime employed by Gottfredson and Hirschi (1990) arose
from classical thinking in which individuals were believed to act in ways that maximized
pleasure and minimized pain. There is a panoptic assumption that human beings perceive
pleasure and pain in exactly the same way. That is, no allowance is made for varying
degrees of each of these constructs among persons, which might result from different
biological, psychological or socio-cultural factors. These all shape the way an individual
perceives what is pleasurable and what is painﬁrl — and while it can be argued that there
are extant overall generalities among all humans, it is important that individual and
cultural nuances be taken into account when considering any criminal activity, including
one as unique as music piracy.

Limitations of Social Learning Theory

One limitation of Akers’ theoretical paradigm is that he does not factor in the
element of spontaneity and quick, spur-of-the—moment decisions to engage in an act. If
some aspect of opportunity theory were afﬁxed to the current ﬁamework, it would
account for those types of crime and deviance which are predominantly the result of a
lack of capable guardians at a locale. Additionally, social learning theory seems to

presume that all individuals can be likened to small vessels being tossed about on the seas

180

of change and circumstance - inevitable pawns with no control over being inﬂuenced by
others. No consideration is given to elements of individuality or unbiased and rational
thinking among people, and each of the four tenets of the theory appears to necessitate a
passive acceptance by the individual in order for the deviance to be internalized. Also,
what makes people decrease or terminate their deviant activities as they grow older?
Maturation is not covered. Moreover, is there a point where those who have appropriated
the values, motives, and rationalizations become part of a separate delinquent subculture
antagonistic to moral propriety, or are they principally adherent to law and convention
but Simply - on occasion - allow themselves to be inﬂuenced and taught to engage in
wrongdoing?

Akers does not provide points around which to develop policy solutions for
varying forms of deviance and crime. He Simply takes speciﬁc actions and demonstrates
how they can be learned according to the theory’s dictates. Once we know how behavior
is learned, then what? Do we target the maladaptive inﬂuences or do we concentrate on
preventing the associations and internalization of deﬁnitions favorable to committing
crime? How do we proceed? Does antisocial learning need to be stymied, or can it rather
be overcome through an abundance of prosocial learning? By extension, how can
education - both moral and legal - be used to inculcate normative principles and
consequently diminish crime if there is always a chance of exposure (and equal
vulnerability) to deﬁnitions favorable to crime?

Some other questions are left unanswered in Akers’ explication of social learning
theory. For example, are all individuals equally able to learn, and are some groups more

susceptible than others? Since nonsocial and social inﬂuences vary in frequency and

181

intensity, are all behaviors reinforced similarly”? Do structures of inequality and
privilege play a role? Does proper socialization and instilling positive beliefs about
individuality and agency in some people reduce the possibility of learning criminality? IS
there any hope for implementation of proactive strategies to counter this trend of learning
deviance?
Methodological Limitations

Inherent in the research methodology are some noteworthy limitations. The
techniques employed do not facilitate precise generalizations to the universe of college
students in the United States, as a probability sampling technique was not utilized.
Further, any inferences made to all undergraduates at the university studied can also be
deemed questionable because the classes chosen for surveying were not randomly
selected but were chosen based on this author’s knowledge of the population. While
Hagan and McCarthy (2002:127) point out ﬂaws associated with collecting data from a
sample of students in a classroom - such as their potential difference from those who do
not attend school and their assumption of a stable familial situation at home - the very
crime under study is disproportionately committed by this population, and no attempts
are made to generalize to all college students across the nation. That said, replication
with samples from other universities of varying student and regional demographics would
be useful to corroborate the accuracy of the current ﬁndings. This author hopes to
complete a second wave study to use in a comparative analysis in the near future, and

will shortly thereafter make the results available to the academic community.

 

38 Akers (1985) mentions nonsocial reinforcers but does not fully develop their role.

182

Another weakness involves the source of the data. A caveat is found in the
research of Brownﬁeld and Sorenson (1993), who compared self-reported delinquency
with ofﬁcial records and determined that the latter is associated with low self-control to a
stronger degree. As such, this methodological limitation may have been replicated in the
current work, as it also utilizes a self-reporting instrument. Furthermore, music piracy
participation may have been underreported because of the tendency of individuals to
provide socially desirable answers, especially about a topic that is so hotly contested and
widely discussed in many social circles.

The survey itself suffered from a few methodological weaknesses. First, it had
124 total questions, all of which were deemed necessary to accurately measure the
theoretical constructs, participation in the criminal activity, and to provide various
demographic and social structural controls. Indeed, previous versions of the instrument
had more items for each theory, but conﬁrmatory factor analyses assisted in paring down
these measures. The number of questions may have been somewhat onerous for
respondents, even though most were able to complete the survey within twenty minutes
and many informally commented that they found it interesting. Additionally, a higher
number of items were employed to measure social learning theory than self-control or
general strain theory because of the necessity to accurately assess the four individual
tenets of the former. The stronger ﬁndings for social learning theory may reﬂect this
disproportionate number of items, but also may not.

AS mentioned earlier, scantron forms were given to respondents to “bubble in”
their answers to save on data entry costs and to preclude data entry mistakes. In their

effort to complete the survey quickly, some individuals may have been hasty while

183

“bubbling in” their response choices — thereby introducing some inconsistencies into the
data. Furthermore, individuals may have failed to erase initial answers completely and
properly, or may have otherwise left some stray marks on the answer Sheet. All of these
could have confounded the scoring machine in some capacity and rendered the results
slightly less than perfectly accurate. These machines, however, are relatively robust in
their scanning and reading of forms, and so it is unlikely that this issue was a problem.

A ﬁnal point relevant to the theories is worthy of mention; Campbell (1988)
eloquently states that “theories are tested only by imperfect exempliﬁcations of their
parameters, which means that deﬁnitional operationism is impossible, but multiple
operationism using designations that are ﬂawed in different ways, is not.” Accordingly,
while multiple items were employed to most accurately measure the theoretical
constructs, it is not possible to perfectly represent them in concrete and direct ways. The
operationalization of GST, SCT, SLT variables, despite this researcher’s best intentions,
may have been less than ideal. Indeed, the Chronbach’s alpha coefﬁcients for the GST
and SCT variables were problematic, and their factor loadings were not as high as those
for the SLT variables. A decision was made to conduct some supplemental analyses with
three singular variables representing strain, attitudinal self-control, and behavioral self-
control, rather than the previously created factor scores for those theories. These
variables were selected based on how evenly distributed the responses were across the
categories. The strain measure utilized was, “Over the last six months, I have received a
bad grade in a class.” The attitudinal self-control measure was “I lose my temper easily,”
and the behavioral self-control measure was “I have driven a vehicle while under the

inﬂuence of alcohol at least once in the past year”. The strain and behavioral self-control

184

measures were already dichotomous; in order to dummy-code the attitudinal self-control
measure, Strongly Disagree, Disagree, and Neutral were combined, and Agree and
Strongly Agree were combined.

In the preliminary OLS regression model, Male, 20 or older, Strain, Behavioral
Self-Control, Differential Association, Imitation, and Differential Reinforcement were
signiﬁcantly and positively related to the factor score dependent variable of music piracy
participation. In the Logistic Regression model, White, Male, Strain, Differential
Association, and Differential Reinforcement were signiﬁcantly and positively related,
while Imitation and Deﬁnitions were signiﬁcantly but inversely related. Finally, in the
Multinomial Logistic Regression model, Male, Strain, Behavioral Self-Control,
Differential Association, and Differential Reinforcement all signiﬁcantly increased the
likelihood of a respondent pirating Medium and High amounts of music, as compared to
Low amounts”. Overall, it is evident that the inclusion of a Singular variable of Strain
predicts piracy to a signiﬁcant extent, while the Strain factor score does not. This is the
only consistent conclusion that could be made with these secondary models; ﬁrture
analyses should attempt to clarify the relationship between strain and music piracy.

Data collection procedures also fostered some issues that deserve comment. AS
previously stated, a total of 169 emails were sent out to professors teaching 185 classes
detailing the nature and arguable importance of the study, specifying the fact that IRB
approval had been obtained, and requesting permission to take no more than 20 minutes

of their class to administer the survey. The number of accommodating responses from

 

39 Being white increased the likelihood of pirating High amounts as compared to Low amounts, but not
Medium amounts as compared to Low amounts. Not much conﬁdence can be taken when analyzing Race
because of the heavily skewed distribution.

185

non-Criminal Justice professors and instructors was quite discouraging. The vast
majority declined, citing a variety of reasons - most frequent of which concerned the
inability or unwillingness to devote 20 minutes of class time to a study that had no
relevance to the subject matter they were teaching. Criminal justice professors - perhaps
because they felt obligated to accommodate a fellow colleague - were more than happy to
acquiesce and offer their time so that data might be gathered from their students to aid in
answering the research questions.

After more emails were sent out to other professors, and more rejections were
received, it was decided that some of the introductory integrative studies classes would
have students from a variety of disciplines. Admittedly, this might be Simply viewed as a
justiﬁcation which was employed to reconcile the unsuccessful attempt to creating an
optimal sampling frame of classes from all colleges across the university. There appears
to be a ﬁne line between purposive and convenience sampling; it seems that the line has
not been crossed, though, because concerted effort was still made to obtain respondents
in a multiplicity of majors — and that goal was achieved.

Concerning the research sample, the demographics of participants were somewhat
skewed in terms of race and socioeconomic status, even though they largely represented-
the overall undergraduate population of the university where the study took place. In the
Fall of 2003, 75.1% of the total student population at the university were Caucasian,
8.1% were African American, 7.4% were International, 5.1% were Asian American, and
4.3% were either Chicano/Mexican-American, American Indian/Alaska Native, Hispanic,
or Unknown. As mentioned previously, the sample in the current research consisted of

7 7 .9% of the sample in the current research were Caucasian; 10.1% were African

186

American, 5.6% were Asian, 3.3% were Hispanic, and 3.1% were classiﬁed under the
collapsed category of “Other”. The parents of 67.8% of respondents made $50,000 or
more each year, and so comparisons of pirating behavior across socioeconomic groups
was not readily forthcoming. These inferential limitations must be kept in mind during
any attempts to generalize the ﬁndings to a larger population.

Recall bias may have also affected the data provided by participants in this study.
Individuals who were prompted to remember their music pirating behavior ﬁom years
past may have been unsuccessful in accurately doing so (or doing so at all), and may have
coddled their memories by adding erroneous information or altering previously stored
information in order to recollect and reveal it in a manner appropriate to what the survey
questions asked. Some scholars argue that data which stems ﬁ'om individuals’
recollection about the past - “retrospective data” - is inherently unreliable not only
because of the tendency for individuals to misrepresent or distort facts from a previous
time period. The errors that are made also tend to be nonrandom, as persons generally
recall facts about past actions in a manner that collimates with their current behavior
(Hirnmelweit, Biberian, & Stockdale, 1978; Horvath, 1982; Morgenstem & Barrett,
1974). Applied to the present study, those who pirate a signiﬁcant amount of music in
current times might state that they pirated similar amounts in past years. Similarly, those
who have stopped participating in the MP3 phenomenon now may report less
participation in the past than is accurate because of the desire for temporal consistency
and behavioral consonance. Indeed, the very fact that MP3 participation may be a

commonplace behavior signiﬁes that it may not stand out in a person’s mind as a

187

noteworthy event when compared to other momentous experiences or occurrences. This
may also compromise or render dubious the accuracy of responses.

Furthermore, the associations found in this study may have resulted from a
reporting bias because questions related to GST, SCT, and SLT were asked as the same
time as those concerning frequency and scope of music pirating behaviors. An individual
who responded with normative answers to the theoretical questions may have been more
likely to provide normative responses measures than someone who did not. Also,
nonrespondent bias may have occurred in that those who had pirated music may have
been less forthright in their responses than those who did not because of its inherently
questionable nature (Seale, Polakowski, & Schneider, 1998). Finally, self-serving bias —
where individuals demonstrate a tendency to view themselves more favorable than not —
may also have been evident among respondent’s choices (Babcock & Loewenstein, 1997;
Cross, 1977).

The Grasmick et a1. (1993) scale, from which items in the current study were
taken, was previously subjected by other researchers to an application of item response
theory (IRT) for the purposes of ﬂeshing out differences between respondent
characteristics and item characteristics, something that CF A does not do (Piquero,
MacIntosh, & Hickman, 2000:898-899). Speciﬁcally, it was important to determine if
the only factor that inﬂuences an individual’s choice among the answer set was in fact
their level of the theoretical construct being measured by that scale and not something
speciﬁc to the person’s ability to objectively respond to an item. [RT analysis provides a
difﬁculty item statistic that indicates how accurately the construct is being assessed

among an individual.

188

The analysis revealed that the validity of a person’s response to items in the
Grasmick et a1. (1993) scale depends on their level of self-control, which calls into
question the purported unidimensionality of the construct (Piquero et al., 2000). Hirschi
and Gottfredson (1993) had also asserted that the Grasmick et a1 (1993) scale suffered
from this methodological weakness and explicitly recommended the use of noncriminal,
analogous behaviors as an appropriate behavioral proxy for low self-control (as
mentioned earlier). This is another reason why the current research utilized both
attitudinal and behavioral measures of self-control in assessing the theory’s relevance.

Participation in music piracy — the dependent variable — was positively skewed,
and as such violated the assumption of a normal distribution for regression analyses. As
such, a transformation of the crime variables may have been useful; perhaps taking the
natural log of the dependent variable scale may have changed the relationship between
the predictor and criterion variables. In addition, CF A on a pretest of ﬁfty-two criminal
justice students was the deciding factor in determining which variables to remove in an
effort to pare down the number of questions in each theoretical scale. The pretest sample
may not have been adequately representative of the ﬁnal sample, though, and as such the
ﬁnal sets of questions may have not been optimal measures of GST, SCT, and SLT
among the larger group.

Finally, it must be mentioned that data collection may have been affected by a
historical issue; the RIAA ﬁled 261 civil lawsuits against individuals who were
exchanging copyrighted digital music ﬁles over P2P networks while undergraduate
classes were still being surveyed. This made national headlines and had the potential to

affect perceptions of music piracy and drastically reduce immediate (and possibly future)

189

participation, and so a ﬂag was set on the data collected after the news of the lawsuits
was made public. Independent samples T-tests revealed statistically signiﬁcant
differences in the responses of those surveyed before and after the lawsuits, which was
initially alarming. Closer look at the means, though, revealed that in with every single
dependent measure, individuals surveyed after the lawsuits participated with a higher
ﬁ'equency in music piracy. This was counterintuitive, and calls into question the efﬁcacy
of ﬁling legal action for the purposes of effectuating deterrence amongst this population -

at least when it comes to digital intellectual property theft.

190

CHAPTER 7: POLICY IMPLICATIONS

The preceding analyses have attempted to contribute to the theoretical
knowledgebase by testing the purported universal applicability of three general
criminological theories. Speciﬁcally, Gottfredson and Hirschi’s (1990) self control
theory and certain tenets of Akers’ (1985) social learning theory were determined to be
causally related to online intellectual property theft in the form of digital music piracy.
Technical and social policy considerations stemming from these ﬁndings merit discussion
for the purposes of suggesting responses that can be implemented by private- and public-
sector institutions. It is hoped these will curtail the pervasiveness of copyright
inﬁingement and lead to increased protection of, and respect toward, the value of
intellectual property.
Policy Implications of Self-Control Theory’s Relevance

According to Gottfredson and Hirschi, self-control is a personality characteristic
largely developed and reﬁned when a person is a child. They also argue that it is age-
invariant - or tends to stay constant over the course of a person's life. Unfortunately, this
does not assist much in the way of suggesting policy solutions that societal institutions
can enact to increase self-control among its members, apart from appropriate and
adequate parenting of children. The role of the nuclear family in developing self-control
notwithstanding, other factors seem to escalate the likelihood of a university student
disregarding any internal or external constraints and proceeding to commit music piracy.

An example can be provided to illustrate two of these factors. Many college
students know of the illegality associated with driving under the inﬂuence of alcohol.

They can viscerally understand why it is a criminal offense, and they are accustomed to

191

stories - perhaps on television or in print - of intoxicated individuals causing their own
death or the death of others. As such, a tangible loss or harm is visible to potential and
actual drunk drivers. Second, almost everyone is aware of the harsh punishment (ﬁnes,
attorney and court costs, incarceration, loss of driving privileges, etc.) that follow a DUI
arrest and conviction, and the fact that the chance for arrest and conviction is respectable.
This points to issues related to certainty and severity of punishment (Beccaria, 1968).
Both of these notions underscore elements that are nonexistent with the MP3
phenomenon. There is no tangible and visible harm associated with participating in
downloading copyrighted music from the Internet, which might serve as an internal
constraint. There is also no substantive threat of detection, apprehension, and
punishment to serve the role of an external constraint.

The former might be tackled through increased use of music artists and bands
speaking out against piracy because of losses incurred to them and the industry.
Recently, both motion picture actors and musicians have spoken out in advertisements
against movie and music piracy, but any positive effect has not been identiﬁed. To note,
South Park - an animated sitcom on Comedy Central - parodied the ﬁscal harm that MPS
downloading has caused musical artists and bands. An FBI detective takes a few of the
main characters — who are elementary-school children - to visit the homes of famous
musicians, and tells them that their music piracy has prevented one artist from purchasing
a gold-plated shark tank to be installed by his pool, another from purchasing a
“Gulfstream IV” jet and making her settle for a “Gulfstream III”, and another from
buying his son a tropical island as a birthday present. This seems to indicate that the

public cannot believe that multimillionaire musicians are ﬁnancially suffering as their

192

songs are downloaded online. Consequently, these deterrence tactics lack legitimacy and
are even laughable in the eyes of ordinary individuals.

The latter would presumably require increased vigilance by law enforcement and
other regulating entities. Intellectual property theft is an act subject to civil and criminal
penalties, and is expressly prohibited by the law. This negative deﬁnition by itself does
not, however, appear to deter people from downloading unauthorized MP3s. Perhaps the
inability of the criminal justice system to actively enforce the law due to resource
limitations is the reason why individuals are not affected by such anti-MP3 stipulations.
To be sure, when the RIAA ﬁled 261 civil lawsuits“) in the second half of 2003 and 532
in January 2004, piracy participation on P2P ﬁle exchange networks dropped off
substantially - but began to creep up relatively soon thereafter as the likelihood of a
music pirate’s detection and apprehension was identiﬁed as miniscule amongst the
millions of persons who took part in the phenomenon. This example is only one of many
initiatives that have not been largely successﬁrl in curtailing the behavior online. Perhaps
more private sector companies in the copyright industries must collectively engage in
such endeavors in order to provide external constraints on the behavior, but the results
may be the same - a sharp decrease followed by an increase to levels prior to policy
implementation. Indeed, attempting to constrain self-control does not appear as the most
viable solution; those that combine strategies steeped in social and technological change

are endowed with much more promise.

193

Policy Implications of Social Learning Theory’s Relevance

It appears that peer associations are quite inﬂuential in an individual’s acceptance
of, and participation in, digital music piracy. These associations presumably provide
behavioral models to emulate, and also champion a belief system that support the activity
while minimizing or ignoring those that do not. It is not practical or functional to attempt
to control interactions between individuals with the intent of preventing the manifestation
of negative inﬂuences that stem from peer association or imitation. What appears to be
more utilitarian and effective is a concentrated effort to address the contributive role of
differential reinforcement, which then may serve to attenuate the inﬂuence of differential
association.

The salience of differential reinforcement can be addressed in two general ways.
Not only must technology he enlisted to conform and direct behavior to adhere to lawful
standards, but general social and individual sentiment towards the appropriateness of
piracy must be modiﬁed through cognitive restructuring endeavors. More Speciﬁcally,
the following initiatives may prove valuable toward this end: the industry’s adaption of a
new business model that takes advantage of digital dissemination of music; the
employment of copy protection schemes to restrict the uncontrolled distribution of songs;
the proactive countering of cognitive factors that impel or induce participation in piracy,
and the creation of a normative culture where legal mandates and moral sentiment do not
contradict or clash, and where individuals feel individually and socially compelled to
abide by those legal standards. These are discussed in detail in the following text with

the intent of depicting their capacity to decrease the rewards and increase the possibility

 

4° Civil, rather than criminal, lawsuits were ﬁled because it was difﬁcult to muster the attention of law

194

of punishment associated with MP3 participation. Indeed, the most useful approach
might be to incorporate elements from all of these policy suggestions to most aptly
produce the desired change in a person’s thoughts and actions.

Embracing a New Business Model

Since the ﬂourishing of MP3 technology, the producers of music have struggled
to develop ways to continue generating revenue while combating piracy. For example,
the top ﬁve major record companies and a plethora of smaller labels have offered a
selection of music downloads — some free and some at a price — for digital audio
enthusiasts. Also, many partnerships have been established between the top labels and
Internet companies such as Bertelsmann’s CDNOW.com, Yahoo’s Launchcom, and
RealNetwork’s Listen.com in order to capitalize on the beneﬁts (marketing, promotion,
distribution, innovation) associated with the MP3 invasion of cyberspace, real space, and
popular culture (RIAA, 2000d). This has lent itself to the continued viability of the
respective companies, and has served to meet the listening needs (and win the allegiance)
of a respectable number of people. As such, embracing the potential of digital music and
assimilating it into a business model may actually prove to be a wise and lucrative
strategy.

In 1999, the ﬁrst year in which MP3s gained a large following, the record industry
experienced an 8% growth in revenue (from $13.7 billion to $14.6 billion) (Ploskina,
2000). Acolytes of the technology contend that these ﬁgures would have been higher
had the industry embraced the new paradigm, rather than seeking to quell the “music

revolution.” If a new model for promotion, reproduction, and distribution for online

 

enforcement.

195

music had been created and implemented - incorporating the promise of MP38 - the
industry may have beneﬁted in sizable ways.

Considering how much appeal digital music has for individuals, there are many
ways that recording labels could provide a valued product and capitalize on existing
demand in a manner that meets the needs of both producers and consumers. For instance,
they could offer promotional singles for free download and discounts on buying an album
after downloading a free song. They could make their entire music database available in
digital format to the online consumer, including those artists and tracks from decades ago
which never beneﬁted from CD capabilities. The release of new music albums using the
Internet as the sole vehicle for dissemination may be a proﬁtable strategy as well. New
revenue models could be implemented for music, including digital distribution,
subscription access, personalized radio, and pay-per—listen webcasts. Marketing
strategies could be better tailored to those who will most likely purchase a particular
artist’s work because of the panoptic nature of the World Wide Web (Breen, 2000). For
example, emails might be sent to individuals who sign up at an artist’s web site in order
to apprise them of that artist’s media or public appearances (including concerts), or news
about a current or forthcoming album. Accordingly, promotional campaigns driving
visitors to particular websites for artists’ music, merchandise, concert tickets, and special
contests can aid in amassing a giant database of users most interested in the music and
most likely to purchase the products available.

To illustrate, musician Tom Petty made available full-length MP3 tracks from a
soon-to-be-released album on his web site in 1999, requiring only that a visitor provide a

valid email address so that news and information related to Tom Petty could be delivered

196

to that person’s inbox every so often (Kibbee, 1999). This provided the musician and
record company with a Sizeable database of Tom Petty fans, who could then be targeted
for concert ticket and merchandise sales, and who would presumably be the most likely
individual to purchase such items. David Bowie, another popular rock artist, was also a
trailblazer in using the Internet to solidify his relationship with fans, expand his
popularity on a global scale, and to market his creative talent. Bowie offered free
downloads of songs from upcoming albums and even live concerts to the visitors of his
web site (Robertson, 2000). Additionally, he gave fans the opportunity to write lyrics to
one of his songs to be selected for inclusion on his new album. Robertson (2000)
compares Internet music pioneers such as Bowie to “drug dealers” who give fans free
“stuff” such as MP3s in the hopes of winning their allegiance and their future business in
the form of album purchases, concert tickets, and merchandise“. An increasing number
of musicians and bands are providing free preview MP3s for digital download on their
web sites to promote their records. Perhaps this needs to be the role that all musicians
should adopt with increasing frequency in order to capitalize on the ubiquity of the Web
and its users’ hunger for information and multimedia.

Furthermore, there appears to be much potential for artists to have more personal
freedom to create music without submitting to the demands and constraints imposed by
record companies, thereby maintaining more control and creative license in their work.
Concurrently, there is an ever-present need to bridge the gap between themselves and

their fans through an online presence. As such, web sites offering space and advertising

 

‘1 Indeed, some MP3 pundits argue that the old music business model should be replaced by a new
ﬁ'amework that concentrates on selling merchandise (clothes, posters, stickers) and concert tickets
associated with the artist, and making music more of a service than a commodity (Philips, 2000).

197

for independent artists to demonstrate their wares have proliferated — some of which
include emusic.com, greenwitch.com, goodnoise.com, kick.com, artistone.com,
mubu.com, gigabeat.com, emikolo.com, and live365.com, (Harari, 1999). In April 2000,
MP3.com had postings from 58,000 unsigned artists and musical groups. In August 2003
(since being purchased by a major record label), more than 250,000 artists were featured
and'over 750,000 songs were available either for download or streaming.

Another way that the music industry might embrace MP3s would be through the
widespread adoption of ID3v2, a labeling system allowing for extra information such as
lyrics, song ratings, copyright information, encrypted ﬁles, hyperlinks, CD cover art, and
the artist’s web page, to be embedded into individual music tracks (Nilsson, 2000). Upon
playback of a track, this meta information would be capable of directing the listener to
visit the web site of a musical artist. It can also be tied to retailers who sell the music or
band merchandise, concert tickets, and a variety of other goods and services that fans
might appreciate.

A more harmonious relationship with the consumer population may even result if
the industry demonstrates they are willing to work with the public to satisfy their music
needs by utilizing MP3 technology, rather than opposing any change to the status quo.
Wholehearted adoption of the technology may also reduce costs to both business and the
environment as manufacturing, packaging, and physical distribution costs are largely
eliminated. Finally, creativity and innovation may be further encouraged because a
global market is now readily available. As the entire process iS Simpliﬁed, positive
outcomes should result for musicians and consumers in the short term, and for the music

industry in the long run.

198

Copy Protection Schemes

The duplication of music prior to the MP3 phenomenon was prohibitively
difﬁcult for the average consumer, and a retrospective look at the evolution of media on
which music is sold attests to this fact. When the predominant medium was the vinyl
record, the general public simply could not afford equipment to reproduce them. The
cassette tape introduced the possibility of duplication at low cost, but quality noticeably
degraded with each successive generation, and was time-consuming due to the need to
play from the source and record to the destination cassette in real time. A similar trend
was evidenced in the evolution of video players. Prior to the introduction of the video
cassette recorder, reproduction of movies was next to impossible; following its
mainstream adoption, duplication did take place with greater frequency but also suffered
from the problem of quality loss“. In both these cases, many individuals came to the
conclusion that obtaining the best listening experience was worth purchasing the ofﬁcial
recording, and a general consensus arose that pirated recordings were of inferior quality.

Digital audio tapes (DATS), following their introduction in 1987, became popular
in the professional recording industry due to their relatively affordable price and storage
features. However, they were never fully embraced by the consumer population because
of the prohibitive cost of DAT players and the fact that a tax was added with each DAT
sold in order to compensate the record labels for losses stemming from piracy facilitated
I by the product (Amter, 2001). Further, digital copies of recordings could only be made

from an original source, as a bit or ﬂag was set on each duplicate digital recording

 

‘2 To note, the very legality of video cassette recorders was challenged in front of the U. S. Supreme Court,
where it was determined that they had “substantial noninfringing use” and afforded time-shifting so that

199

signifying that it was, in fact, a clone. Terrned Serial Copying Management System
(SCMS), this prevented the creation of another generation of copies from that clone.
Sony Minidiscs were another technological advance that failed to catch on among
consumers in the United States, despite their popularity overseas. They allowed for
portable digital recording onto a small disc, incorporated SCMS and provided
functionality previously unknown (Woudenberg, 2003). Both DATs and Minidiscs did
not engender mass music piracy Simply because of the difﬁculties associated with easy
duplication and their comparative lack of mainstream adoption.

Audio compact discs (CDS) were brought to market in 1982, became popular in
the late 1980s, and have Since been the medium of choice in terms of recorded music
(Amter, 2001). Videos on digital video discs (DVDS) were introduced in 1996 and have
acquired considerable market Share from video cassettes Since the beginning of the 21St
century. Their ubiquity has provided relatively insecure digital sources of music and
movie data that were similarly difﬁcult to duplicate in their initial years of existence.
With the continued exponential growth in technological capabilities, increasingly easy-to-
use methods to extract media content and preserve the ﬁdelity of the audio and/or video
are emerging. The Size of data ﬁles is also not as relevant as in the past, due to the
growing pervasiveness of fast connections and large hard drives“. When considering the
previous mediums in light of the current advances, it is interesting to note that the digital

music and movie phenomenon is the ﬁrst time that the general populace has possessed

 

individuals could view a previously-aired program at a later time ("Sony Corp. v. Universal Studios,"
1984)

‘3 As of December 2003, brand name hard drives with capacities around 200 megabytes were available for
$150 or less.

200

the ability to copy and propagate high—quality creative works - and thereby dictate their

own experience of audiovisual media.

As such, two competing interest groups need to reach common ground in terms of
their demands and perceived rights in order for the controversy that surrounds digital
music to be resolved. Consumers -— who have become accustomed to obtaining an
incredibly large amount of high-quality music from a variety of time periods and genres
at no cost — want this trend to continue. Furthermore, they desire no limitations that
inconvenience their ability to transfer songs to portable players or burn them to
recordable CDS. Producers — who had been accustomed to signiﬁcant control over the
distribution, marketing, and cost of music prior to the MP3 phenomenon — desire
adequate compensation and revenue generation for their talent and investments, and to

continually to maintain a fanbase that will perpetually be a source of income.

It is incontestable that many artists, primarily those independent and unsigned, are
quite willing to advance the distribution of digital copies of their music on the Internet,
Simply because of their love for music and their desire to promote their musical efforts.
At the same time, there are hundreds of artists (primarily those used to obtaining royalties
whenever their music is sold or used) who vehemently discourage the illegal copying of
their music online. They argue that while those artists who openly allow and support
online dissemination of their work should have the freedom to continue in that vein, those
who disagree with the practice should be able to protect their creations from unlawful
duplication, and a mechanism Should be in place to afford this defense. With this in
mind, individuals have demonstrated that they increasingly prefer the convenience of

obtaining high-quality music online rather than through a retail establishment due to

201

continued growth in broadband availability and computer technology to the general
public. In order to align with the interests and objectives of the producers and to meet the
consumer demand for downloadable tracks online, some constraining factor must compel
individuals to purchase music through authorized Intemet-based distributors, as opposed

to freely downloading music from P2P ﬁle-exchange networks.

Digital rights management (DRM) is one possible solution being advanced by
several IT and media companies such as Microsoft, Sony, and Xerox Corporation. It
seeks to restrict the uncontrolled distribution of digital ﬁles by embedding protective
code inside the music ﬁle. This allows the media to be used through DRM-enabled
software or hardware, or for a limited time, or solely for one computer system, rendering
itself unusable if transferred elsewhere. DRM-encoded ﬁles are not only protected in
transfer to the end user, but also are protected from use beyond what is authorized. This
is termed “persistent protection” — as the content is secured continually due to inherent

control mechanisms (Stamp, 2002, 2003).

DRM allows for the creation of a digital music infrastructure that allows the
music labels to have more control over the consumer’s listening experience. Through
unfettered music piracy, consumers have been unfairly bringing the intellectual property
of others into the public domain. As such, DRM may be useful in restoring the balance
between the interest of the public and the rights of creators and owners. In the ideal state
for the recording industry, then, music will no longer be shared with impunity but will be
delivered to those who pay for the right to listen to it, and will only be playable by
approved, DRM-compliant devices (MP3.com, 1999; Weekly, 2000). In terms of ﬁscal

incentives, DRM has also been billed as the recording industry’s saving grace due to its

202

potential to thwart piracy and to increase their ability to dictate how music will be
obtained and utilized — presumably in a way that generates revenue in as proﬁtable a way
as traditional sales from brick-and-mortar retail stores. By way of illustration, DRM was
requisite before Apple went public with its iTunes digital music store in May 2003
(Long, 2003; Zeiler, 2003)“. Apple knew that if the music ﬁles they offered were
insecure and easily duplicated between individuals online, their ﬁnancial proﬁt would

greatly decrease.

Typically, the security of valuable digital content has relied on the “honor”
system where delivery to an authorized customer takes place using cryptographic
methods but can be accessed and then saved in an unsecure method on the recipient’s
hard drive — which of course allows for the unfettered and illegal re-distribution of that
unprotected content to other individuals. Piracy will continue on an incredibly large
scale, then, unless the content — e-books, movies, music —.is indissolubly integrated with
a protection scheme that controls its use. One might argue that the implementation of
technical measures to protect content is largely Sisyphean because all software-based
limitations will inevitably be broken. This notion has merit, but does not invalidate the
use of rigorous DRM methodologies to secure content. The overarching goal is to make
the reverse-engineering and cracking of security controls a more difﬁcult alternative than

Simply purchasing the content. Also, the security control must “fail well” (Schneier,

 

4‘ iTunes has been a terriﬁc success, selling two million songs to Macintosh users in its ﬁrst sixteen days of
existence (Zeiler, 2003). It is supported by many recording artists and the music industry because it
encodes music in the more secure Advanced Audio Coding (AAC) ﬁle compression format, restricts usage
to three corrrputers authorized by a single individual, and because its inherent “sharing” functionality is
restricted to ﬁve users on a local area network.

203

2000, 2003). This means that when it is compromised, the loss and damage is isolated
and localized so that greater loss and damage is precluded.

A brief description of a typical DRM technology tied to the distribution of digital
music can illuminate this technical policy solution. Rather than providing insecure,
easily reproducible MPB ﬁles to interested individuals online, the music is packaged in an
encrypted form — rendering it unplayable unless a key is lawfully obtained (i.e.,
purchased) from the owner or authorized distributor of the content. Public-key
encryption can facilitate the secure exchange of keys via the Internet between parties”.
Apart from the key, a license must also be obtained which speciﬁes the rights an
individual has with the digital content at hand. Perhaps the person can only play the song
a particular number of times, or for a certain number of days (Cravotta, 2000; de
Fontenay, 1999; Weekly, 2000). Perhaps the person has unlimited ability to listen to the
music when, where, and how S/he desires. This would be deﬁned in the license and
would be related to a price that intuitively increases according to the freedoms afforded.
Indeed, the license generally contains the key that decrypts the secured music ﬁle for
playback, and is often hosted on separate Intemet-based servers which coordinate
delivery to a DRM-enabled software application (such as Windows Media Player) on the
end-user’s system“.

When the encrypted music ﬁle is downloaded, it cannot be enjoyed unless the

corresponding key is obtained via purchase of the appropriate license. In the “license

204

predelivery” method, the software that is performing the request and download of the
media ﬁle is also granted delivery of the license and key concurrently in a seamless and
transparent manner. Alternatively, “license postdelivery” involves the acquisition of the
license and key in a separate and additional process from the initial music ﬁle download,
and often requires the provision of payment and personal information (e.g., age group,
musical preferences, purchasing habits) to the producer. As is evident, the former is the
most convenient and unobtrusive, but does not provide a wide range of options for
delivery and information gathering.

Another distinction concerning the secure delivery of digital content is also
relevant. In a “tethered” scheme, the DRM-based media player contacts a particular web
server and requests a decryption key to the music ﬁle every time an individual attempts
its playback — consequently requiring a dedicated connection to the Intemet“. The
decryption key is destroyed after playback ensues, which provides robust security at the
expense of inconveniencing the user to always be connected. More corrrmonly employed
is the “untethered” scheme, where the DRM-based media player requests the appropriate
key once and keeps it with the media ﬁle on the local machine, rather than obtaining it

each time the ﬁle is played.

 

‘5 For technical details related to this process, please see (Diffle, 1988; Difﬁe & Helhnan, 1976).

‘6 It should be noted that the allegiance of the purchasing population must be maintained and their interests
and preferences considered prior to the implementation of any DRM scheme that might not resonate
positively with them These solutions vary in the level or degree of restrictions imposed upon end users,
and may need to be reevaluated in light of the desires of consumers to enjoy music in a hassle-free and
uncomplicated manner.

‘7 A similar practice is identiﬁable in the software industry, as developers are writing their applications and
games so that their execution triggers a small amount of data to be sent to the corrrpany or business which
released the product (“phoning home”, so to speak) in order to verify the legitimacy of the serial number
used to unlock the full capabilities of the program (Microsoft, 2000).

205

Incorporating both predelivery and postdelivery methods in the distribution of all
secure music online appears to be the best technological solution. The initial license
provided during predelivery can be limited in its scope, thereby providing a evaluation
period of sorts to an individual. Once that evaluation period expires, the DRM-based
media player can connect to a server that dispenses a license allowing for unlimited and
unrestricted playback upon provision of two pieces of data: information necessary for a
purchase transaction, and information related to demographics and listening preferences
for future marketing purposes. If the end user has enjoyed the music ﬁle during the
initial period, S/he can legitimately buy it and continue to derive its beneﬁts. Conversely,
if the end user did not, s/he can choose to decline the invitation for purchase and move
on. Consumer desires of quick, easy acquisition of a valued commodity with unrestricted
usage can thus be balanced with the creator and producer demands of control over, and
adequate remuneration for, their digital intellectual property.

A major obstacle to the universal adoption of DRM is the fact that many
companies are devising protective technologies that are not interoperable with one
another. Unprotected, commonplace MP3 ﬁles are playable on a variety of operating
systems with a variety of software media players, while DRM-encoded ﬁles often require
proprietary software for playback. Efforts are underway to develop a standardized
language of interaction48 so that license, key, and usage data in secure music ﬁles can be
extracted and utilized across applications and operating systems without action or even

awareness on the part of the end user. However, it may be that open standards with

 

‘8 The XML-based languages currently under development for secured digital content include ODRL
(http://www.odrl.net) and XrML (http://www.xrrnl.org), and the reader is encouraged to visit their
respective web sites for more information.

206

exhaustive documentation may allow for the creation of software and music players that
ignore the metadata that restricts functionality, thereby allowing for circumvention of the
DRM technology. If it becomes easy to implement DRM in a standardized way across
platforms, it may be similarly easy to devise a method to enjoy the media content without
adherence to the restrictions in place. The veracity of this point has yet to be determined.
Some interim solutions have been brought to market due to the delay in a
standardized DRM adoption. Recording labels have released new albums from their top
artists and bands on copy-protected CDS, which take advantage of the differences in how
audio CD players and computer CDROM drives read data from discs. The way in which
the data is written to the CD allows for its playback in the former but not in the latter. In
theory, this strategy had potential for success, but in reality has infuriated and frustrated
consumers by causing operating systems and software to lock up and crash (Mariano,
2002; Oakes, 2000). Additionally, issues relating to the degradation of consumer rights
also surface, as many individuals contend they Should be able to listen to the music they
purchase in any capacity without restrictions. It remains to be seen what impact DRM
will have on the digital music trade. As of the beginning of 2004, DRM-encoded
Windows Media Audio (WMA), Apple iTunes, and Napster 2.0 ﬁles have all garnered
some popularity among certain groups of consumers, but no single DRM implementation

has captured the lead as a viable overarching solution to the problem of music piracy .

Incidentally, Weekly (2000) voices an intuitive concern that proponents of secure
digital music must address before copy-protected formats have any potential to become
standardized. A software application, when reading an encrypted or otherwise-secure

audio track, must decrypt it and direct it to the computer’s sound card in an unprotected

207

raw format for it to be played. If a secure sound recording can be outputted to a sound
card, it must at some point be rendered insecure before playback can ensue. At that
point, it is vulnerable for usurping and copying into a digital ﬁle format by software on a
computer system. As a consequence, individuals so inclined can easily develop a piece of
software that can intercept and duplicate the contents of that ﬁle as it is passed off from
the operating system to the sound card. Then, they can save the data in a format once
again suitable for unrestricted duplication and dissemination without any limitations or
perceivable negative repercussions, thereby invalidating all of the security and copy
protection measures that were implemented. Some programs that perform these ﬁrnctions
are available commercially at a very reasonable cost”. Also, newer sound cards such as
those made by Creative Labs allow consumers to digitally record any audio played
through the sound card with no loss in quality. Finally, other individuals can still exploit
the “analog hole” by utilizing a device to record music from the audio speakers
themselves — which admittedly reduces the ﬁdelity of the recording but still provides an

acceptable duplicate of the music at no cost (Wikipedia, 2003)

The DRM facilities heretofore mentioned are primarily software-based, and
circumvention of protective controls — through techniques such as those mentioned above
— remains a distinct possibility. To augment the difﬁculty inherent in breaking the
controls, joint software and hardware initiatives are underway, such as Microsoft’s
forthcoming “Next-Generation Secure Computing Base for Windows” operating system

core, which works in conjunction with Intel’s Trusted Computing Platform Alliance”. In

 

’9 Total Recorder (www.totalrecorder.com) and Super MP3 Recorder Pro (www.audio-mp3—recorder.com)
are two examples.
5° Microsoft’s Next-Generating Secure Computing Base was previously given the code name Palladium.

208

basic terms, digital content is linked with the unique hardware in each individual’s
system, effectively binding it to one location and one person (Carroll, Juarez, Polk, &
Leininger, 2002). By extension, breaking the security of one digital ﬁle on one system
will not open it up for exploitation and misappropriation by others. The security, then,
fails “well” (Schneier, 2000, 2003).

Microsoft has emphasized that insecure and unprotected content acquired prior to
the introduction of their secure computing base will still be playable on their new DRM-
enabled systems (Carroll et al., 2002). Nonetheless, once the majority of new music and
movie releases are DRM-encoded and distributed solely online, individuals will be forced
to participate in the secure schema through legitimate purchase in order to obtain and
enjoy the commodity. While the ramiﬁcations for users of other operating systems such
as Linux and BSD have yet to be determined and demand consideration, this seems the
most promising technical approach to effect the protection of copyrighted content, and to
stem the tide of unfettered intellectual property theft. Nonetheless, past experience
underscores the very real potential for compromise in technological solutions, and
therefore a complimentary initiative that addresses cognitive, psychological, and
sociological stimuli is warranted. Steve Jobs, CEO of Apple Computer, has echoed the
same sentiment: “Piracy is a behavioral issue, not a technological one” (Taylor, 2002).
Cognitive Restructuring Initiatives

The proactive countering of cognitive, behavioral, psychological, and sociological
inﬂuences of music piracy is as important as the institution of reactive technological
measures. For instance, when the boundaries of lawful behavior are clearly deﬁned, it

will presumably be more challenging for potential offenders to justify their deviant

209

actions. Deviance, then, may be reduced in severity and frequency with the use of laws,
legal sanctions, or threats of sanction (Tittle, 1980). By extension, if acceptable and
unacceptable computing behavior is plainly Spelled out by university administration
through the use of ethical codes substantively similar to laws and legal sanctions, the
incidence of piracy among students may be attenuated. Engendering a respect for
intellectual creations and property among students is an essential function of higher
learning, particularly when it involves a networked environment where duplication and
dissemination of works without the author or owner’s permission can proliferate easily
and with great celerity.

Tittle (1980) has stated that levels of wrongdoing may be decreased if laws are
crafted and made known deﬁning the behavior as illegal and prescribing penalties for its
violation. Online intellectual property theft - inclusive not only of music but also of other
forms of digital content - must somehow be designated as completely unacceptable in
order for individuals to abstain from participation. The Campus Computing Project
found in 2003 that 80 percent of public, and 78 percent of private, universities have
policies in place explicitly prohibiting individuals from downloading copyrighted content
(Word, 2003). Indeed, at the university where the research was conducted, a substantial
migration to a new email system was being made over the course of 2003. The
“upgrade” process was requisite for all students who desired to continue to receive email
at their university account, and stipulated that downloading and uploading unauthorized
copies of digital music was not acceptable use of the computing resources they were

provided. Speciﬁcally, it stated:

210

“Examples of unacceptable use of your network account
include sharing copyrighted ﬁles through ﬁle sharing or
peer-to-peer software such as KaZaA, Morpheus, Gnutella,
or other Similar program. If you do not adhere to this
policy your network access and e-mail account may be
suspended.” (Michigan State University, 2003)

The presumed goal of such a declaration is to increase awareness of the illegal
nature of the activity despite its prevalence. For instance, at the university where this
study took place, a feature article on the unacceptability of downloading MP3s was
printed in the school newspaper on the ﬁrst day incoming freshmen in Fall 2003 were
able to check into their residence halls. The article stated that the university receives
approximately 35 complaints each day from music and movie industry representatives
who have scanned the network utilizing software which identiﬁes the host IP of the
computer facilitating the data transfer of copyrighted material (Frank, 2003).

Furthermore, all students who registered their computer to use the broadband
network resources on campus were required to indicate that they would comply with the
following statement:

“I acknowledge if I share copyrighted material from my
network connection using a program like Kazaa, Morpheus,
Gnutella or other ﬁle sharing program or method I will be
subjected to disciplinary action which will minimally
include the loss of my network connection. I also risk

losing more than my campus Internet connection. Owners

211

of copyrighted material may sue me, or press a criminal
complaint against me which could face [Sic] heavy ﬁnes-—

or even imprisonment.” (Rondeau, 2003)

Such statements are often utilized in conjunction with campus-wide Acceptable
Use Policies (AUPS) that delineate general appropriate and inappropriate use of computer
systems and software. For example, they prohibit the unlawful access, inﬁltration,
disruption, and damage of systems or networks belonging to others. Further, they
typically specify adjudicatory measures such as disconnection of individual Internet
connections in preference to the examination of the contents of user data due to strict
adherence to principles supportive of academic freedom and privacy (Michigan State
University, 1998; Middle Tennessee State University, 2001). This practice promotes an
environment conducive to the advancement of knowledge through its unrestricted

expression and dissemination, while still designating penalties for transgressors.

Civil and criminal prosecution are speciﬁed as possible sanctions, but most
matters - at least at the university studied - are resolved internally after contacting the
infringing individual and requiring their discontinuation of the unacceptable activity
(personal correspondence, August 3, 2003). First-time violators are sent an email by
network authorities on campus to immediatelycease their activity; second- and third-time
infringers are subject to loss of their Internet connection, academic suspension or even
expulsion (Frank, 2003). To note, though, this piecemeal, case-by-case approach of
addressing computer-related infractions does not help to inform the entire student
population of the wrongfulness of the action, and a more panoptic initiative which can

precipitate widespread change in the conceptual paradigm of individuals towards the

212

activity may be more useful. As an example, incoming freshmen in Fall 2003 at the
University of California in Berkeley who desired to use their dorm-room high-speed
Internet connection were required to attend a 30-minute orientation session that focused

heavily on the illegality and punitive repercussions of ﬁle-sharing (Brand, 2003)5 1.

Finally, the RIAA has developed a campaign to work in conjunction with colleges
and universities to increase awareness of copyright issues on the Internet and to foster a
respect and appreciation for the intellectual property and creative output of others”.
University administrators can order educational materials from the association and can
institute programming to encourage ethical and lawful conduct on the Internet, and it is
hoped that such a partnership will curb the rate of copyright violations among college
student populations nationwide53 . Irrespective of who plays the role, it is clear that some
social institution or authority must teach individuals that stealing a product from a retail
store and stealing a product over the Internet are both examples of theft - illegal activity

necessitating prohibition and penalty.

 

5 l A technical measure was also introduced to complement the cognitive initiative - a limit of ﬁve gigabytes
of data can be transferred both upstream and downstream by students each week (Brand, 2003). To note,
though, this would still allow approximately 1,250 downloads of four megabyte MP3s to be obtained and
will perhaps only rarely become an issue amongst students.

52 See http://www.soundbyting.com

53 In order to determine the efﬁcacy of codes, ethical training, warning signs, disclaimers, and entreatrnents
to students from university personnel in reducing the frequency of music piracy through increased
awareness and sensitivity, longitudinal studies must be performed following such policy implementation.

213

External to the university environment, law enforcement has not developed any
campaign to inform individuals about copyright infringement and intellectual property
theft, and may never view these acts as signiﬁcant enough to warrant such policy. The
recording industry and certain high-proﬁle musicians and celebrities have spoken out
against the unauthorized transfer of digital music ﬁles, but are perhaps too distant and
remote for individuals to actually relate to and agree with. Curbing music piracy through
personal admonishments and informative initiatives may be fruitful, but more individuals
must perceive the problem as serious and contribute to such efforts.

The private sector has also come under ﬁre for accidental or willful ignorance of
copyright inﬁingement occurring on corporate networks. The Recording Industry
Association of America, the Motion Picture Association of America, the National Music
Publishers’ Association, and the Songwriters Guild of America have jointly crafted and
disseminated letters to businesses to increase their awareness of the phenomenon, request
that steps be taken to monitor and block digital intellectual property theft over their
networks, recommend software tools that will aid in this endeavor“, and even threaten
lawsuits if no action is taken.

Unfortunately, the distinction between right and wrong among copyright
infringing behaviors appears amorphous, unclear, and susceptible to varying
interpretations. AS such, society must step up and address this issue through the speciﬁc
and conspicuous delineation of appropriate and inappropriate computing behavior. With
regard to ﬁrture research, it would be instructive to compare statistics of music piracy

before individuals were made familiar (or reminded) of its intrinsic unlawfulness, to see

 

5‘ Appendix C provides a sample letter for the reader.

214

if any general deterrence resulted from the conscience-raising effort. It might be
determined that the inability to cultivate cognitive restructuring among those particularly
prone to engage in online deviance has in some respects fostered and perpetuated the
problem of music piracy.

Shaping Morality to Cohere with Legality

Tyler (1996) provides an interesting commentary on the ineffectiveness of
deterrence measures in facilitating compliance with law among individuals, and instead
points to the relevance of two related concepts: morality and legitimacy. He argues that
intellectual property law is, in itself, impotent to dissuade copyright infringement because
it is economically and pragrnatically impossible to implement mechanisms to raise
probabilities of detection, apprehension, and punishment past a threshold where they will
induce conformity. Regulation on the Internet has only recently been attempted through
the ﬁling of civil lawsuits by the RIAA in 2003 and 2004 who offered copyrighted music
for unauthorized downloading from their computer systems (Bowman, 2003; CNN .com,
2004; Dean, 2003a).

The Department of Justice has hinted toward criminal prosecution and
incarceration of music pirates for the purposes of “preserving the viability of America’s
content industries,” but this has yet to occur outside of the reahn of commercial software
(McCullagh, 2002). Unquestionably, priorities are set by the general public, and law
enforcement must cater to the demands of social and political pressures. Further, the
police simply do not have the resources (as of yet) to expend on combating this form of
theft, and societal members will generally rail against privacy-compromising intrusions

into their lives. Even if such resources were allocated and such “big-brother” practices

215

allowed - thereby increasing the likelihood of penalty for transgressors - the impact of the
initiatives would wane and eventually cease to exist as individuals come to a realization
that the perceived risks are either less signiﬁcant than previously imagined, or that there
are other methods to circumvent the model of justice in place. When coupled with the
widespread opportunities for copyright inﬁingement, threats of punitive action designed
to stimulate abidance will remain unproductive. As mentioned earlier, in the current
research a ﬂag was set differentiating data collected prior to the news release of the 261
civil lawsuits in September 2003 and those collected after (but prior to the ﬁling of 532
civil lawsuits in January 2004). An independent samples t-test countered intuition by
ﬁnding that participation in music piracy actually increased following media coverage of
the lawsuits, rather than fostering general deterrence and resulting in a decrease of the
activity.

The objective, according to Tyler (1996), Should be to gain voluntary cooperation
among the citizenry through cognitive restructuring of their conceptions of morality and
legitimacy. The former concerns a person’s beliefs towards wrong and right behavior,
while the latter concerns a felt obligation to abide by the law. Tyler (1996) asserts that
individuals largely engage in behavior not because of the degree or chances of risks and
rewards, but because it aligns with their sense of morality. By way of illustration, the
role of morality in determining actions - to a greater degree than the threat of sanctions -
has been identiﬁed in a host of research studies (e.g., Christensen & Eining, 1991; Eining
& Christensen, 1991; Grasmick & Bursik, 1990; Grasmick & Green, 1980; Grasmick et

al., 1993; Nagin & Paternoster, 1993; Paternoster, 1989; Paternoster & Simpson, 1996).

216

By extension, dissonance in moral beliefs is apparently not experienced by
individuals who contemplate and then participate in copyright infringement. Law is most
effective when it coheres with the moral consensus of its subjects, and a climate must be
created where individuals not only experience moral qualms when considering or
partaking in intellectual property theft, but one in which overall law conformance is
expected and culturally obligatory. This notion is not new; centuries ago, Jeremy
Bentham ([1781] 1970) stated that the restraining power legal sanctions have stems in
large part from their connection to social sanctions. In the online realm, there are not yet
any viable social sanctioning mechanisms in place. This consequently places the onus for
the promotion and preservation of Internet propriety not only on the courts to develop
legal sanctions, the private sector to develop technological safeguards, and law
enforcement to execute the law, but also on society to engender an appreciation and
respect for digital property and copyright.

Second, individuals must view the laws, its process of creation, and the authority
ﬁgures in charge of its promulgation and administration as legitimate, and thereby feel
compelled to respect and obey them. Both morality and legitimacy contain a notion of
justice and fairness, as citizens will volitionally subscribe and support only that which
they believe was devised and implemented in a manner they deem evenhanded and
proper. Future research Should identify what individuals conceive as fair when it comes
to downloading, copying, and distributing intellectual property, and policy makers should
attempt to embrace and then modify those public perceptions to collimate with the

written law.

217

In sum, Tyler (1996) maintains that deterrence and preventive measures (e.g.,
through technology) only hold short-term worth, while affecting societal conceptions of
morality and legitimacy have long-term implications, and is decidedly the policy road to
travel to engender lasting positive change in this area. The deterrence approach in this
subject area appears with Acceptable Use Policies and similar warnings that vilify music
piracy, while the preventive approach is observable in DRM schemes. Both of these have
been discussed, and the astute reader will be able to identify inherent vulnerabilities and
weaknesses in their actual implementation. Tyler’s (1996) suggestion to create a law-
abiding utopia appears overly idealistic, and no practical steps are articulated towards this
end. In accordance, perhaps the combination of deterrent and preventive measures —
though admittedly short-term — can serve as Stepping-stones in the direction of
widespread positive change in shaping behavior through informal social control.

The preceding research sought to analyze the predictive capacity of certain
cognitive, behavioral, psychological, and sociological factors in determining participation
in music piracy. The hope was that each of the three general theories from the social
sciences would determine - to some substantive degree - the behavior under study.
Individually, they would explain some variation in music piracy; collectively, they would
explain a moderate amount and would truly clarify the stimuli associated with the act.
Unfortunately, a great deal of variation in music piracy is as yet unexplained, despite the
inclusion of intuitively powerful and inﬂuential theoretical predictors.

To a large extent, the majority of criminological research results in similar
ﬁndings —— a great proportion of variance in the dependent variable is due to elements that

are not accounted for in the research, despite the fact that most studies purpose to

218

empirically verify the expected inﬂuence of theoretically- and conceptually-relevant
independent variables. Notwithstanding the constant variability in the human condition
and the consequent unpredictable nature of any action — criminal or otherwise, are social
science researchers to be content with small R2 values and little clarity as to the strongest
contributive elements of a phenomenon? Is there more to it than our weak-to-moderate
ﬁndings, with which we seem so content? If so, the “big picture” may not be as elusive
as the statistics Show (or, for that matter, do not Show). It appears obtainable with a two-
pronged approach: 1) by altering traditional conceptualizations of the makeup of criminal
behavior and 2) by rallying elements from other disciplines outside the social sciences to
improve our predictive models.

For the purposes of ending this work in a progressive and optimistic manner, a
profound transformation in the way in which social scientists perceive and respond to
crime is proposed. The following chapter introduces and explicates a new theoretical
model for criminology, borrowed from the sociology of science and technology. It
allows us to elucidate all of the possible causal factors associated with digital music
piracy in an effort to most comprehensively understand it. The discussion builds upon
the initially conceptualized model in a substantive and radical way by combining
predictive social and technical elements into an interdisciplinary “network” paradigmatic

framework to fully capture the etiology and reality of the phenomenon at hand.

219

CHAPTER 8: THE HETEROGENEOUS ENGINEERING
OF INTERNET CRIME

Two fundamental conceptual positions appear well-entrenched in the
criminological enterprise, and perhaps even deﬁne and demarcate the standardized way in
which research problems are approached. First is the disproportionate focus on
individual, interpersonal, and socio-structural elements. Indeed, the reader may be
wondering, “What other predictors of crime could there possibly be?” Such a reﬂexive
thought epitomizes the issue. Second, social science scholars throughout recent history
have vacillated in their allegiance between the determinate roles of agency as compared
to structure. Certain epochs have been characterized by a greater acceptance of one or
the other depending on social and political consensus at the time, but perceptions and
policy have been reﬂective of either agency or structure as causal inﬂuences. These
points are further explicated below to provide the backdrop for the introduction of a novel
theoretical approach.
The Social and the Technical

Criminologists undoubtedly distinguish between people and things in their
attempt to understand antinormative behavior, and devote incommensurate attention to
the former. Some might contend that focusing on individual and socio-structural factors
has served the discipline well in its goals to explain variation in traditional criminal
activity. However, others might posit that criminology has not been as successful or as
expedient in understanding the nature, causes, and correlates of crime as it could be.

Many scholars in various disciplines have expressed the importance of one or

more of the technical components of crime, but no prominent effort has been made to

220

comprehensively subsume all elements under an explanatory rubric. Scientiﬁc progress,
however, hastens such action. Novel forms of criminality that have surfaced in the last
decade - such as computer crime and Internet crime — take obvious advantage of many
more “things” in their execution. Additionally, the inability of traditional authorities such
as law enforcement and legislators to competently curtail high-tech deviance has
compounded the problem (Borrowman et al., 2001; Leibowitz, 1999; Manning, 1997).
These changes demand a paradigm shift that understands the equally contributory
relevance of a vast number of nonsocial components that are intertwined and linked with
the social - ultimately effectuating the phenomena at hand. It can even be cautiously
asserted that a greater understanding of how human and nonhuman elements commingle
and interact with each other might, in fact, enable research endeavors on traditional
crimes to be more successful in their explanatory goals and even in the efﬁcacy of
derived policy solutions.

It appears evident that disciplinary boundaries, an unyielding commitment to
traditional methods of inquiry, ignorance and lack of awareness (partially self—imposed
and partially intrinsic) of other ﬁelds, and subject area specialization have largely
preempted most joint endeavors of understanding. If two or more disciplines have
actually worked together, it has been in a sequent and temporally laborious manner,
rather than with Simultaneously thriving, reﬂexive synergy between their representative
ideas. Productive and efﬁcient interdisciplinary approaches are far too uncommon due to
the establishment of occupational specialization that has continued ever since the

Industrial Revolution introduced the division of labor.

221

With regard to sociotechnical phenomena, information technology-based
conceptualizations tend to focus on instrumentation and applied proactive or reactive
methods of response. For example, there is much epistemological confusion about which
disciplines are to be summoned to address the complexities of Internet crime. Even if
both social scientists and computer scientists see merit in perceiving the activity through
their specialized paradigmatic frameworks, they do so ﬁom separate spheres of inﬂuence
and very rarely attempt to coordinate a joint response - despite obvious overlap. Social
scientists concentrate on elucidating and comprehending the predictors of the behavior,
while computer scientists devote their efforts to devising technology-based strategies to
monitor, thwart, reduce, or confound the activity or the actors.

To be sure, partnerships between the disciplines are occurring in increasing
frequency in very recent years55 — but this is the vast exception rather than the norm.
Research and grant proposals from each discipline vary so pronouncedly and
categorically in their content, and it may be discovered (off the record, of course), that
neither truly believe in the efﬁcacy of their individually-constructed solutions because of
a shared unstated notion that some “part,” some “element,” some “component” is
missing. In this author’s opinion, that which is missing from the social perspective is the
technical, and that which is missing from the technical perspective is the social.

Elements of social reality, it seems, cannot be clearly or accurately understood
through disproportionate concentration on the “human,” which is all too often done in the

social sciences — particularly when attempting to explain criminal phenomena. Indeed,

 

55 At this author’s doctoral institution, a cybersecurity partnership has been formed between a number of
disciplines, and led by criminal justice and computer science. The partnership has led to much productivity
in joint teaching, research, and outreach efforts.

222

the preceding study exempliﬁes the type of exploration and analysis performed regularly
in the “soft” sciences and — like the rest — is unable to thoroughly identify and Specify the
exact correlates of one type of crime. If criminologists are responsible for studying the
causes, nature, and extent of crime, does that charge restrict them to devoting their
research to humans? Is that written in ﬁne print on the job description? If not, and if
criminal justice is considered an interdisciplinary ﬁeld, what prevents the incorporation
of a signiﬁcant nonhuman (material) component in conceptualizations of crime?

Michel Callon and John Law, two preeminent scholars in the sociology of science
and technology, have speciﬁcally asked, “Why should we start out by assuming that
some of these [technical objects] have no active role to play in social dynamics?”
(1997:168). The heterogeneous nature of any manifestation of social reality demonstrates
that multiple “nonhuman” factors are not to be viewed as separate considerations, but as
equally important and relevant. They mediate interaction between individuals and, to
some extent, Shape and dictate the manifestation of that interaction. Furthermore, the
animate does not exclusively dictate how the inanimate function, nor does the inanimate
dictate exclusively how the animate function (Law, 1992). As such, no mandate or
necessity exists to separate the two and then analyze one to the exclusion of the other.
Agency and Structure

The classical school of criminological thought, exempliﬁed by the tenets of free
will, hedonistic calculus, and rational choice, aligns with the sociological precept of
“agency,” where individuals are completely responsible for their actions. The positivist
school of criminological thought ﬁnds its place at the opposite end of the spectrum, and

coheres with the sociological precept of “structure”, where normative, cultural, legal, and

223

social characteristics constrain and determine one’s actions, excluding entirely the role of
individual volition. Latour (1997) asserts that the social sciences have historically
alternated levels of analyses between the micro-level and the macro-level. That is,
scholars have focused on the low-level characteristics of a situation, event, or person —
the veritable manifestation of a host of converging background variables. Conversely, at
other times they have sought to step back, acknowledge, assimilate, and appraise the
contributions of societal and systemic factors to that manifestation. Both, however, have
left these scholars with an unsettled and incomplete feeling. Empirical studies over the
last three centuries have failed to adequately clarify whether agency or structure is more
salient in determining criminal behavior, and it appears that no resolution to this debate is
imminent.

Supplementing this point, Latour (1997) states that a phenomenon can never
become completely micro-level (because the micro-level instance does not exist in a
vacuum and has been inﬂuenced by other entities and relationships) and can never
become completely macro-level (because there are always instances where grouping or
“summing” takes place in a localized, endemic area). The phenomenon, then, vacillates
between the two extremes therein, and can be considered a “circulating entity” (Latour,
1997:17) rather than a ﬁxed macro- or micro-level construct. As such, its research must
occur in a method that takes note of all heterogeneous components, as well as the
capacity of each to affect others amidst the interconnections of a network which
represents its existence. Accordingly, the agency/structure debate may simply be
misguided. Research which concentrates on the determinate value of one are invariably

limited by the failure to invoke or consider the other. Even integrated theories that

224

incorporate both appear incomplete. A departure from these traditional
conceptualizations appears requisite to make theoretical and explanatory headway in the
reﬁnement of our knowledgebase.

Actor-Network Theory (ANT) corrects for these two limited paradigms by
imbuing nonhuman actors with the capacity to act, focusing on the role of interactions in
determining action, and by avoiding certain elements of traditional ﬁxation and
reevaluating their signiﬁcance when viewing the “big picture” and their interconnections
with other heterogeneous artifacts within a network-based architecture. Relationships
exist between various human and nonhuman entities that produce agency and structure.
Volition and context, then, are not the unit of analysis since they are eﬂects or outcomes
of interactions (Law, 1999:3); rather, the interactions themselves are what determine any
and all behavior.

In the following text, ANT is introduced to the ﬁeld of criminology as a viable
and holistic theoretical perspective — accounting for both the social and the technical - in
which to view crime and deviance. It is then applied to the phenomenon of digital music
piracy in an attempt to underscore the importance of studying and analyzing each of the
components (and their interrelationships) which make up the reality of that illegal
behavior. Deﬁnitions of human and nonhuman actors, as well as descriptions of the
relationships between those actors that create and perpetuate the existence of the
phenomenon, are also presented. Finally, policy implications that stem from the theory
are discussed, along with general suggestions as to how future research Should proceed.

Actor-Network Theory

225

Stemming from the sociology of science and technology, ANT views any and all
phenomena as the product or effect of patterned networks of heterogeneous materials —
both social and nonsocial in nature (Law, 1997). Heterogeneous engineering, then, is the
activity that produces the form and structure of a phenomenon, and this form is best
understood by studying its constituent elements (Law, 1987b:113) Furthermore, pieces
of social reality are to be viewed and interpreted not through a reductionist lens which
focuses solely on the human and social element, but also the nonhuman and nonsocial,
which include: technical elements (such as material objects); organizational, situational,
and economic elements; informational elements (such as the press, the media, cultural
and normative standards); and even contextual and perceptual elements (Law, 1992).
These nonhuman entities play an active rather than passive role, and just as humans do,
they act and interact with other elements in the network.

Law (1992) points out that the actions of actors should not be solely ascribed to
their individual beings, but should be conceived of as a derivative outcome of a network
pattern. This then clariﬁes the meaning behind the term “actor network,” as an actor is
inherently a network (Law, 1992). More speciﬁcally, “an actor network is
Simultaneously an actor whose activity is networking heterogeneous elements and a
network that is able to redeﬁne and transform what it is made of” (Callon, 1987:93). To
illustrate, a law is not a law unless there are lobbyists to suggest it, correspondence to
discuss it, legislators to create it, logic to support it, civil liberties to undergird it, a justice
system to record it, police to enforce it, judges to interpret and apply it, law books and
legal documents to display it, governmental buildings and computer systems to house it,

databases to archive and index it, and a populace to respect it. These constituent

226

 

components interrelated in a network fashion demonstrate that law is a compound reality
(Callon & Law, 1997).

A network can be mapped and understood, then, by following actors as they
“deﬁne and distribute roles, and mobilize or invent others to play these roles” (Law &
Callon, 1988:285). These roles may be social or nonsocial, individual or corporate - and
all are interwoven into a “seamless web” (Hughes, 1983). Such interconnections are
often marked with a tension stemming from the disassociative properties of each
individual component. Law (1992) asserts that these constituent elements would like to
each “make off on their own,” but their networked integration creates a force that
overcomes any resistance and serves to reify the phenomenon. The collectivity of these
elements must hold up against indifference and hostility from all interacting network
forces (i.e., nonhuman and human actors). Its composition, then, “reﬂects the power and
the nature of both the forces available and the forces with which the network collides”
(Law, 1987b: 121). As such, the network must maintain the relationships that support its
persistence while assimilating or rejecting those that might prove detrimental. This is
part of the dynamic, continually changing nature of the network, and ensures its
survivability.

It is important to note that following construction of the network, stability is
maintained through a continued process of reappraisal and redesign, and new or modiﬁed
actors (having autonomy) and artifacts (without any autonomy) can be incrementally

brought in to preserve its general form and structure. Each element contributes either

227

directly or indirectly (through other components) to the overall goal(s) of the network56
(Hughes, 1987). Interestingly enough, each is the simpliﬁcation of a smaller network of
other elements. They have no inherent properties; such properties are instead acquired
through relationships with other entities. As they deﬁne and shape one another, entities
are also performed in, by, and through those relationships, which provide durability and
stability (Law, 199924).

The goal of the sociology of science and technology is “to make sense of [the]
processes and understand the social consequences of the artefacts [$10] that are thereby
generated [through the interaction and juxtaposition of various technical, political,
economic, and social complexities]” (Law, 1987a:406). While one might consider the
contents of their domain to be largely nonhuman, scientiﬁc and technological
developments are inextricably intertwined with societal factors; many examples in the
sociological literature base attest to this (see, e.g., Hughes, 1983; Law, 1987a). For
instance, Law (1987a) points to how automated machine tools on the shop ﬂoor
controlled from a remote location via programmed numerical controls was largely due to
a desire for social control. The mechanization of typesetting seemingly due solely to
economic interests was also ushered in by a desire to allow craftsmen to reassert
dominance over the printing process and over the millions of women who work at
keyboards (Law, 1987a).

As a ﬁnal example, in a piece analyzing the technological and sociological
components of a British military aircraft project, Law and Callon (1988) assert that both

are inextricably intertwined and Simultaneously determine the ultimate development of

 

5° Hughes actually chose to use the word “system” instead of network; network is used in this piece for the

228

any phenomena. Sometimes the technical features exert inﬂuence over the demands of
the social (when the weight of the aircraft dictates the speciﬁcs of its takeoff procedure
and functionality), and sometimes social, political, and bureaucratic factors are more
cogent (when the requirement for a long-range aircraft and a smooth ride for the pilot
dictates the design of the wings). Similarly, computer crime might seem to be primarily
technologically-based, but it necessarily implicates a host of other elements which must
be considered in any thorough attempt to understand it.
Application to Digital Music Piracy

An example is useful to lend clarity to this idea and to better understand its
salience to criminological phenomena. Indeed, the lucid representation of the constituent
elements in any activity is key to understanding its complexities - such as how it is
created, how it stabilizes, and how relations between the social and technical are
determined and mediated. Any computer crime can be considered a product or practice
resulting from a patterned network of heterogeneous elements. It can be conceptualized
and studied as the amalgamation of a host of varied components ordered and organized in
a manner that creates a cohesive whole. Moreover, it cannot occur without each of these
components; that is, the network would weaken or fail if one or more is absent or
malfunctioning. In this architecture, every “actor” (a term used for both human and
nonhuman elements) is dependent on every other actor, and its integrity is only realized if
all remain participatory and in working order (Callon & Law, 1989). Failure to consider
any constituent parts - and excluding their place in relationships - will result in a failure

to fully comprehend the phenomenon. One type of crime - intellectual property theft on

 

purposes of consistency and uniformity.

229

the Internet in the form of digital music piracy — is extremely suitable for explanatory
purposes to illuminate the crux of ANT. A supplementary conceptual map of the
phenomenon is provided for visual support in Appendix D.

MP3 (an abbreviation for MPEG-1 Layer Audio 3) is an audio compression
format that enables audio ﬁles to be compacted into relatively small ﬁle Sizes, while
maintaining near perfect ﬁdelity when played back. Compression occurs through the use
of perpetual coding techniques, where auditory information from large digital multimedia
ﬁles that exceeds the perceptual range of human hearing is removed, resulting in smaller
ﬁle Sizes (Crawford, 2000). The technology allows for giant repositories of music to be
easily accumulated by individuals from online sources, circumventing the previous
requirement to purchase the recording on a physical CD from a retail establishment in
order to enjoy the valued commodity.

According to the US. copyright law (Title 17 U.S.C. Section 101 et seq., Title 18
U.S.C. Section 2319), however, only the owner has the lawful ability to distribute,
reproduce, adapt, perform, or display their creative works. Intellectual property is
copyrighted as soon as it is expressed in a ﬁxed, tangible form (Copyright Ofﬁce of the
United States, 2000a), which consequently renders it protected under the aforementioned
legislation. Anyone or anything that directly or indirectly contributes to unauthorized
dissemination or duplication of another’s copyright work, then, is committing a federal
offense subject to imprisonment and ﬁnes (Copyright Ofﬁce of the United States, 2000b;
RIAA, 2000c).

The most frequently-employed method of obtaining digital music online involves

the utilization of ﬁle exchange software and participation in a peer-to-peer network

230

hosted by the company. It appears that use of the technology and the programs that
facilitate it stems primarily from word-of-mouth and media coverage. Participation
ensues when individuals whose interest has been piqued obtain, install, and run the
software that provides them with a user-friendly interface and the point-and-click
functionality to search for and download music ﬁles. These ﬁles come from the
computer hard drives of other individuals online who are Similarly connected to the same
service; the software and service, then, serve as the mediator between the two parties and
broker the activity. Nonhuman actors such as technological devices coordinate the
interactions of human actors, and this occurs in a social, organizational, Situational, and
economic context that facilitates the activity. A greater elaboration of this process now
follows to more accurately depict the theory’s explanatory potential.

To begin, the crime cannot occur without the existence of music as a valued
commodity. It is created and produced by a variety of human actors such as artists,
bands, and their respective music labels, who are a part of the recording industry57 as a
whole. Nonhuman actors such as instruments, sound engineering and mastering
software, recording studios and equipment, and even design and artwork for CD cases
and packaging are also largely essential. Presumably, the music industry acts so that their
product is purchased legitimately rather than obtained non grata. This, of course, is so
that they receive a certain amount of compensation and reward for developing the
creative work. Their unwitting presence in the network is requisite for music piracy to

occur. Music as a commodity and the recording industry that creates and produces it,

 

57 For the purposes of parsimony, the “recording industry” refers to the overseeing association, participating
recording labels, the commercial artists and bands signed to these labels, and the infrastructure that
encompasses all of these entities.

231

when set in a relationship with the technology of MP3, collectively provide the object
which is Situated at the conceptual center of the phenomenon.

Additionally, it is obvious that the crime of music piracy cannot occur without
some actions taken by humans desiring to beneﬁt from the availability (in digital form) of
the valued good. The participant58 can be considered the primary agent of the activity as
S/he instigates the action and “gets the ball rolling.” Having the motivation and
inclination to pirate music, however, is not enough to actually access and download the
desired good; other individuals (secondary agents) on the Internet (secondary agents)
must somehow meet the needs of that participant.

These co-participants59 (also human actors) who host and provide the intellectual
property must be accessible for interaction, in either temporal synchronicity - where
active interaction between parties occurs on a real-time basis (e.g., peer-to-peer ﬁle-
exchange programs, chat channels, and instant messagingiprograms) or asynchronicity —
where data transfers can occur between participants even if the co-participant is away
from the computer system (e.g., web servers and ﬁle servers)6°. Both the participant and
co-participants are linked to: the technology of MP3s; a hardware, software, and network
infrastructure; the criminal justice system, and social, organizational, and economic
systems. The piracy of digital music only occurs when all of these components act upon

and among each other.

 

58 Throughout this piece, “participant” always refers to the person initiating digital music piracy.

59 “Co-participants” always refers to those who aid the initiator of music piracy (the “participant”) in
accomplishing intellectual property theft, generally by providing digital music ﬁles to download.

60 To note, there are exceptions to this statement; peer-to-peer ﬁle-exchange programs might facilitate
downloads in an asynchronous manner if co-participants are passively connected to the network and have
available their library of ﬁles for any active primary user to access.

232

As alluded, the heterogeneous network is not complete if one considers only
human elements. Various technological actors are also implicated in the architecture, and
their absence would render music piracy infeasible. For instance, a computer system, the
relevant peripheral equipment, and an Internet connection is required. The hardware and
software that reside in and on the computer system interact with the Internet connection,
and are also vital for certain actions which facilitate online intellectual property theft.
Software for web browsing, newsgroup reading, chatting, instant messaging, and peer-to-
peer ﬁle downloading are the primary nonhuman actors and interfaces through which
human actors access digital music. Software to extract and decompress archived (e.g.,
.zip, .tar, .rar, .arj) ﬁles and to piece together ﬁles separated into “dictionaries” for easier
transfer is sometimes necessary. Playback software to listen to ﬁles is indispensable.

Even the participant’s operating system is a nonhuman actor in the network as it
allocates and manages the hardware and software resources, and provides a way for any
application or data to interact with the hardware in a relatively consistent and
standardized way. To note, the computers and Internet connections of the co-
participants are also automatically implicated. In addition, software that resides on both
parties’ systems are unexpendable brokers of any data communication and transfers that
occur. All of this hardware and software mediates and shapes the social reality of the
phenomenon by initiating, facilitating, monitoring, and closing data exchanges.
Curnulatively, these computer-related objects are linked to the human actors and the

commodity of music through the vehicle of MP3 technology.

233

The criminal justice system“ is also a crucial part of the network. In its most
perfect form, it would effectively deter or prevent potential and actual transgressors
through effective surveillance, detection, apprehension, and enforcement capabilities.
Furthermore, its efforts — when partnered with those of the prosecution — would enable
the state or federal government to construct strong cases that can aptly demonstrate proof
of guilt in court. In such a utopian scenario, the court system would clearly interpret
traditional copyright law, and would apply them to issues that arise online in a consistent
manner, so that no unjust variability allows some criminals to escape punitive sanctions.
The legislators would contribute towards this end by creating relevant laws and statutes
through informed decision-making processes, intelligently considering the range of
ramiﬁcations before their promulgation and enactment. These laws would clearly deﬁne
what constitutes appropriate and lawﬁrl behavior online, would logically cohere with
previous jurisprudence and societal standards, would prescribe appropriate penalties, and
would preempt the possibility of arbitrary and biased decisions handed down by the
courts”. Software would be created and utilized to police digital information networks
for the purposes of monitoring, tracking, identifying, and responding to those who break

the law“. Unfortunately, though, such ideals are far from what has actually materialized.

 

6‘ For purposes of simplicity and because the term is commonly used in this author’s discipline, “system” is
used to describe an organized array of elements that represent a larger component that is part of the
network. Traditionally deﬁned in the social science literature, a largely social system consists of the
“patterned activities of a number of individuals. . .complirnentary or interdependent with respect to some
common output or outcome” (Katz & Kahn, 1966). Such a conceptualization of a system is applicable in
this context, but ANT gives equal attention to the contributory role of nonhuman actors playing a part in the
network architecture, rather than viewing a system solely as an organized collective of human relationships.
62 A bill was proposed in July 2003 that clearly designates as a felony the act of uploading a ﬁle to a peer-
to-peer network. If passed, possible penalties that can result include imprisonment up to ﬁve years, and a -
ﬁne up to $250,000 (Dean, 2003b).

63 A bill was proposed in June 2003 to errrpower federal agents to create software to deter copyright
infringement online, to develop FBI-related warnings to send to violators, and to coordinate greater

234

As such, many of these criminal justice elements are weakly linked to other actors in the
network (such as the participant, the co-participants, the recording industry, and the
technology of MP3), and are represented by dashed lines on the ANT conceptual map in
Appendix D.

These criminal justice actors all provide examples of ineffectual controls; the
veritable “absence” of informed and directed action is equally as important as the
“presence” of purposeful action by other actors (Law, 1994). By extension,
strengthening or reinforcing these links to and from criminal justice entities would
presumably aid in crippling the phenomenon. It would also seemingly lead to a
symbiosis of sorts among the elements of the criminal justice system. For example, if
legislators crystallize their deﬁnitions of online copyright inﬁingement and the
consequent penalties for wrongdoing, the process of enforcement by the police will
become easier, as will the job of the prosecutors who initiate formal legal proceedings
against offenders. Thus, despite the large number of weak links among and between
these nodes, improvement among some components will lead to stronger relationships
among others. To note, the converse is also true — a point which should hasten immediate
attention and response by those with the ability to change the current state of affairs.

Social, organizational, situational, and economic elements are brought to the
forefront by the human and nonhuman actors who comprise the phenomenon. Socially,
one or more individuals (or the institutions they represent) must engage the participant by
making him or her aware that certain items of socially ascribed value (e.g., music,

movies, software, books) are available to download from Internet sources such as web

 

information-sharing among law enforcement, ISPS, and the victims of intellectual property theft (HR2517,

235

pages, bulletin boards, ﬁle servers, and even from other individuals concurrently online.
These actors must also highlight the rewards associated with the activity and the
comparative lack of punitive repercussions that presumably will result. Finally, they
must ascribe value to the digital intellectual property, and champion it as a commodity
worth having. Knowledge of how to participate in the environment of acquisition is also
important - such as which web sites to visit, and how to enter into any communities that
support or facilitate the practice. The participant must also know the technical steps to
initiate a ﬁle download to his or her computer system, and the steps to execute or run the
ﬁle for the purposes of playback. These can be learned as the relationships between
human actors are mediated by nonhuman objects (e.g., technology) in the network. Most
importantly, all of these instructional processes which serve to facilitate and further the
behavior are brought about by, and manifested in, the actions of participants.
Involvement by individuals in music piracy indicate and inform others of the possibility,
acceptability, and proﬁtability"4 of the action. As a consequence, these actions can
consequently introduce and ensconce an individual in the phenomenon.

Actors also create an online organizational context that is fundamental to the
phenomenon of music piracy. For example, multiple stable venues‘55 are provided for .
participants to congregate, converse, and exchange MP3 ﬁles, and are developed by
individuals with knowledge of technology and availability of computing resources such

as hardware, software, and Internet connectivity. In these contexts, human actors through

 

2003)

236

their actions also overtly endorse or subtly condone the practice of uploading and
downloading digital music ﬁles without appropriate authorization from the owners of the
respective copyrights“. Without this organizational backdrop, the interaction of other
human and nonhuman actors would have no supportive environment in which to occur.

Criminal justice scholars would agree that one Situational component is requisite
in order for a crime to occur - the opportrmity to partake in the wrongdoing must be
presented to the offender. A host of elements contribute to the inﬁastructure of
opportunity and either increase or decrease the probability of the activity occuning.
Opportunity is not, however, an actor that relationally interacts with other nonhuman or
human actors. Rather, the opportunity to pirate music comes into existence due to the
action (e.g., technological devices which enable the creation, distribution, and access of
digital music ﬁles) or inaction (e.g., law enforcement’s inability to successfully combat
hi gh-tech deviance) of other actors.

Finally, economic aspects are brought into the fold via human actors. To begin,
software for exchanging MP3s is available at no cost online“. AS such, there is no
prohibition to participating for anyone who owns a computer and can connect to the
Internet. This greatly increases the number of potential pirates; it is presumable that if

peer-to-peer software were not freely downloadable, both the amount of participants and

 

6‘ Proﬁtability is by no means restricted to ﬁnancial gain through the avoidance of purchasing legitimate
music recordings . Rather, many nontangible and tangible beneﬁts unrelated to saving one’s monetary
resources are derived from music piracy.

65 These include interactive web sites, chat channels, bulletin boards, newsgroups, mailing lists, and peer-
to-peer ﬁle exchange programs.

66 The reproduction and dissemination of a digital music ﬁle without explicit permission from the copyright
holder of that creative work is a federal offense and is subject to both civil and criminal penalties (RIAA,
2000c).

237

music would be severely reduced. In addition, wealth is unequally distributed among
individuals in the current US economic system, and the extant capitalistic enterprise
exclusively champions goals of proﬁtability that may indirectly or directly lead to the
exploitation of consumers. Individuals are affected in two ways. First, the price of
retailed recordings, as set by human actors, enables only the more afﬂuent to legally
acquire as much music as they might want, as most people can only afford to Spend a
certain amount of money on CDS each year. Second, due to the recording industry’s
arguable monopoly on music, the consumer might be unfairly overcharged when
purchasing their product. This would allow the reaping of tremendous ﬁscal rewards
from the industry’s business model to the detriment of the populace who unwittingly
supports it.

When considering the role of economic factors, it is instructive to view the
phenomenon from another vantage point through the lens of ANT. Can an argument be
made that music piracy is encouraged by the actions of those who represent the recording
industry and its constituent components (e.g., recording labels, artists, bands, producers)?
Due to being overcharged for music recorded onto physical media such as CDS,
consumers might bring to the network a particularly negative conception of the industry.
This conception may have fostered feelings of antagonism which percolated in a
subterranean manner for years among some individuals“. Passive acceptance of the

situation and continued consumption of the industry’s valued product might have

 

67 Software that provides the capacity to create MP3s from CDS is also available online at no cost. CD-
recording software to burn MP3s to CD is relatively inexpensive and often comes preinstalled on many
new computer systems. These factors expand the already pervasive reach of the phenomenon.

68 Also, it is important to note that a motive to retaliate at the music industry's business practices might not
have existed among most people.

238

continued indeﬁnitely without any interruptions. Soon to surface, though, was the perfect
opportunity to move from exploitation to increased equitability in the marketplace.

Whether this vantage point is more factually representative than the former is not
the issue; the example is provided only to stimulate thought and discussion as to how
ANT can aid in comprehensively understanding all sides of a phenomenon. As was the
case in the prior example, all of these social and nonsocial actors are essential for music
piracy to occur. The technology to compress CD-quality music into a ﬁle size small
enough to transfer over an Internet connection was developed and subsequently
introduced to the music-loving population. The infrastructure of hardware, software, and
networks — coupled with the lack of an entry cost - greatly simpliﬁed the process of
creating and exchanging MP3s with other individuals. Social, organizational, Situational,
and economic elements were implicated, all the while beneﬁting ﬁ'om the ineffectual
presence of the criminal justice system. Finally, the behavior was reciprocally supported
and reinforced, causing network elements to coalesce into a context that provided a
breeding ground for the phenomenon.

To reiterate, ANT suggests that the properties of computer criminality are a
product of the practices involved in the construction and maintenance of a networked
pattern of heterogeneous entities. If, as highlighted by the concept of heterogeneous
engineering, the material is equally as important as the nonmaterial, and the social and
technical are indissoluble components of any innovation or phenomena, then scholars,
researchers, and practitioners have been moving towards the best solution to certain
crime problems in a very piecemeal, inefﬁcient, and inappropriate fashion. Accordingly,

perhaps a different worldview is requisite to more fully understand and explain crime -

239

one that appreciates the active and contributory role of nonhuman elements and that does
not disproportionately focus on individual agency or on societal structure to account for
behavior. The convergence of the social and technical in the etiology of crime appears to
necessitate such a novel approach.

Actor-Network Theory and Characteristics

The preceding discussion makes no mention of latent and directly unobservable
factors which may play a role in effectuating a phenomenon. Certain knowledge, skills,
abilities, predispositions, inclinations, traits, and contributive thought processes seem
imperative to enable a person or group of persons to commit a crime. These, however,
are not “actors” according to ANT, and cannot be conceptualized as such because they do
not act or interact. Instead, they are considered characteristics of one or more actors, the
existence of which are only identiﬁable when manifested in action. For example, if an
individual is plagued with narcissistic qualities, such characteristics are unobservable
until that person reveals them through verbal or physical actions. Moreover, if a person
is mathematically gifted, that characteristic will not be known - and cannot interact with
other elements to produce a phenomenon - unless it is demonstrated via a conspicuous
action (as that person interacts with other actors).

Latour (1987) Speciﬁcally argues that the nonmaterial (i.e., characteristics) Should
only be invoked if material elements are inadequate in their explanatory power. Only the
material are accessible; for instance, legal proceedings are based primarily on evidence
that is tangible and which results from individual action, rather than on evidence that is
intangible and indirectly (and perhaps incorrectly) ascertained and measured.

Accordingly, it appears that these inaccessible characteristics may not be useful in

240

explaining a criminal phenomena. In line with Latour’s contention, they exist but are
only observed and measured through a person’s actions.

In contrast, criminologists often argue that in order to most completely understand
their respective behaviors, the context from which each human actor comes must be
examined. From their perspective, individuals have certain characteristics which they
bring as nodes; these include predilections, motivations, and experiences that
cumulatively affect their decision to act in an antinormative manner. Criminologists
would also argue that these persons are of a certain mindset, which speaks to elements of
the psychological, cognitive, and even emotional that inﬂuence or regulate their behavior.
Knowledge, skills, and abilities are also seemingly implicated, as they endow the
participant with the capacity for crime commission. These characteristics would warrant
close attention and examination by researchers who subscribe to, and are ensconced in,
traditional approaches of conceptualization.

In sum, criminologists have attempted to indirectly assess these latent
characteristics as predictors of crime through both quantitative and qualitative methods.
Perhaps disproportionate attention to factors not readily observable has caused the “big
picture” to be missed — a big picture that can be directly appraised. ANT scholars argue
that inquiry should be directed to the manifestation of these unobservable characteristics
in the network — the actions that are made by those who possess the characteristics —
rather than the characteristics themselves. ANT in its purest form demands focus on the
importance of every visible, tangible social and technical actor and its relationship to
other actors without consideration of underlying and indirectly measurable characteristics

that they bring to the table. The interactions reﬂect and represent the constitution of the

241

human actors, and therefore are sufﬁcient (and more accurate) proxies for a person’s
mindset, motive, and makeup Moreover, they can be directly and more accurately
measured.

This detailed breakdown of music piracy illustrates that together the
aforementioned constituent elements comprise an ordered and organized heterogeneous
network that cumulatively produces the phenomenon under analysis — music piracy.
Each component has its own necessary place, and its absence or malfunctioning can
cause the network to fail. Crime is generally viewed and conceived of as a social
phenomenon (in terms of it being primarily expedited and deﬁned by humans).
However, violent crimes often occur with the use of a weapon. Property crimes
necessarily involve tangible property, such as a home when considering burglary, or a
vehicle in the context of automobile theft. Corporate resources are implicated in all
white-collar crimes. With so-called ‘Victimless” crimes such as substance abuse and
gambling, drugs and money are requisite. Very little attention is given to these factors
because of the misconception that they are available and uncontrollable, do not act, and
cannot be affected through the use of policy. ANT argues that these nonhuman actors do
act, and that their interaction with other nonhlunan and human elements is what reiﬁes
the phenomenon.

The baggage of criminological tradition has encumbered and restricted novel
approaches to crime. AS a consequence, strategies to address criminality have
disproportionately focused on conventionally-deﬁned “social” factors, not taking into
account that every single human and nonhuman element in the heterogeneous network

mediates and even deﬁnes the “social.” That is, all factors that comprise a crime result

242

from the interconnectivity, cooperation, and synthesis of these constituent nodes. Each
actor, then, affects other actors and therefore merits analysis in its own right (Latour,
1987). Callon (1987:99) states that studying the evolution and maturation of technology
from a sociological perspective “means recognizing its proper object of study is neither
society itself nor so-called social relationships but the very actor networks that
Simultaneously give rise to society and to technology.”
Further Application Of Actor-Network Theory

Before proceeding to a discussion of policy that logically stems from this
theoretical paradigm, it is useful to probe deeper into the process behind the solidiﬁcation
of the network representing any phenomenon. By examining how actors act and form
relationships to accomplish their intentions, ways in which the network can be hindered
or thwarted become visible. It was stated earlier that each element might experience a
desire to “make off on its own” due to its disassociative properties (Law, 1992).
Accordingly, the network then must have the capacity to overcome the individual
resistance of each of its constituent parts. Through a process termed translation, they
must be concertedly enrolled and mobilized in sponsorship of the activity by participants,
who also must vigilantly maintain their allegiance to the grand design or purpose (in this
example, the piracy of music).
Translation

Translation, refers to the method in which an actor is able to speak for and
represent the interests and voice of another actor (human or nonhuman) to accomplish
certain purposes (often selﬁsh in nature) (Callon, 1987). Ultimately, in order to obtain

the cooperation, support, and greatest likelihood of succeeding in the action or behavior,

243

all elements must be linked together in a network where the “identity of actors, the
possibility of interaction and the margins of manoeuvre (sic) are negotiated and
delimited” by the participant of music piracy (Callon, 1987:203).

Four stages exist to this translation: problematisation, interessement, enrolment,
and mobilization. During the problematisation phase, participants must render
themselves indispensable to the other actors —- each having desires to attain their own
aims. This can occur by establishing themselves as a node on the network (an
“obligatory passage point” (19872204) that must he traveled through by the other actors if
those others are to achieve their respective goals. The second stage - interessement -
concerns the imposition and stabilization of the identities which the participants have
given to the other actors (Callon, 1987). This is accomplished by heading off any
problematisations - attempts to deﬁne their identities for other purposes - by still other
actors in the network. “To interest other actors is to build devices which can be placed
between them and all other entities who want to deﬁne their identities otherwise” (Callon,
1987:208). In other words, it is the thwarting of attempts to appropriate the actors for
purposes other than those of the participants.

The third phase - enrolment, where negotiations, trials, compromises, techniques,
and tricks are utilized to actually engage the actors into alliances. Certain “enemy
forces” (Callon, 1987:211) must be addressed and conquered before the reciprocal
relationship can be solidiﬁed — at least for a period in time. With mobilization - the
fourth and ﬁnal phase - the participants are ﬁnally able to speak for and represent the
other actors, thereby “mobilizing” their individual voices and integrating them into one

stance. Their positions, initially separate and divergent, have now been brought into a

244

relational, cohesive unity. This integration is transient due to the continued necessity to
maintain allegiance and cooperation among entities, but has settled for the time being into
a period of stability.

How is this applicable to the phenomenon in question? It appears that translation
of technology and the criminal justice system is necessary to successfully pirate music.
Participants who want to obtain digital music with complete freedom and impunity need
to give identities to these two actors, deﬁne their associations (the roles they will or will
not play), and extract compliance from them (or noncompliance if it furthers the cause).
Any resistance posted by these actors must be overcome; in fact, they must be assimilated
into a set of relational alliances which condones or even fosters the activity through
interactions with other actors. The participant must negotiate the positions they perceive
the actors are in, extract their compliance, arrive at a point where their disassociative
tendencies are curbed or rendered impotent, and bring them into a uniﬁed position that
directly or indirectly supports the crime. Their allegiance, moreover, is not automatically
stable and therefore must be maintained if the activity is to continue. “The bits and
pieces in the network. . .their form, their content, and their properties are not ﬁxed.
Rather their identity emerges — and changes — in the course of interactions” (Callon &
Law, 1997: 171). As such, the network remains dynamic in nature and requires effort on
the part of participants to perpetuate its viability should anything about the nodes or
relationships change.

1) Technology.
To effectuate music piracy, human actors must enlist the assistance of computer

networks, hardware, and software - devices which can be used to partake in normative or

245

deviant behaviors. They must also control or manage any technical schemes or
mechanisms which are in place to thwart piracy - either by invalidating or circumventing
them. Technology therefore must be interessed and enrolled. Speciﬁcally, technology
must be used in a more efﬁcacious manner than its potential employment by antagonists
of the MP3 phenomenon, and possibly even prevented from full exploitation by such
antagonists (who might seek to develop various technical controls on digital music to
restrict its reproduction, transfer, or use). Finally, technology must be mobilized to
accomplish the ends intended for it by the participant. This can be achieved by exploiting
the technology’s functionality to create, reproduce, and distribute MP3s, and thereby
dictate its primary use.
2) The criminal justice system.

To facilitate music piracy, it is essential that participants experience no resistance
(i.e., action) from the actors that comprise the criminal justice system. These include law
enforcement, the legal system that has Speciﬁed laws against copyright infringement and
the punitive measures that are due to those who violate such laws, the courts who attempt
to apply the statutes, the prosecutors who partner with the music industry to criminally
pursue offenders in this situation (e.g., issuing “cease and desist” letters through ISPs to
users) (Borland, 2002; Bowman, 2003; Dean, 2003a; Musil, 2003). In principle, the
system is decidedly opposed to participation in the MP3 phenomenon. In practice,
though, the system has many other pressing responsibilities that take precedence over
online intellectual property theft, and therefore are unable (or unwilling) to utilize their
resources to stem the tide of copyright abuse. The actors (both human and nonhuman)

that comprises the criminal justice system, then, do not work towards interessement,

246

enrolment, and mobilization of the technology to accomplish their selective purposes and
goals (ensuring compliance to copyright law) — of course a tremendous beneﬁt to
. participants.

Notwithstanding, participants must incorporate the system into the network
representing music piracy, because ignorance of its presence (however slight or
perceivably ineffectual) is often a recipe for disaster that has led to civil and criminal
penalties for some (Davis, 2003; Healy, 2003; IFPI, 2002; Luckenbill & Miller, 1998;
Patrizio, 1999; RIAA, 2000a; Roth, 1999). Music pirates can then do what is necessary
to acknowledge the system’s presence while acting in a manner that complicates its
ability to deter or restrict their activity, and that perpetuates a defeatist mentality among
them. Interessement by the participant may occur by preventing the criminal justice
system ’ﬁ'om amassing the authority to surveil or track more Internet users, by
championing privacy concerns online, and by heading off partnerships with other
political or economic powerhouses which might strengthen the antagonist’s platform and
effect. Enrolment of the system is not possible for participants unless mandated
restrictions against copyright violations are somehow overturned — which seems highly
unlikely. Mobilisation is also irrelevant due to its biased position of law preservation

(rather than law breaking). Still, its presence in the network is essential because the

-.‘4

offender(s) must acknowledge and accommodate it.

While the co-participants of music piracy translate, interesse, enroll, and mobilize
in the same way as participants, the recording industry does so quite differently. The
recording industry, by controlling the production, marketing, and dissemination of music,

has established itself as a network node through which the music-buying consumer must

247

travel in order to obtain that desired commodity on a physical medium. Problematisation
has thus been accomplished and maintained for decades, and this has allowed them to
grow into a $40 billion industry (RIAA, 2003). Interessement takes place as efforts are
made to prevent or head off the endeavors of other actors to take music and deﬁne it as a
digital commodity that can be freely exchanged over the Internet. Media campaigns,
legislative lobbying, cease-and-desist letters, and initiatives to develop technological
solutions such as secure digital music formats and digital rights management are all ways
in which the industry has attempted to stabilize their deﬁnition and conception of music
and its legitimate acquisition by customers.

Enrolment occurs as the recording industry teams up with IT companies to offer
digital downloads for purchase online, in order to regain control over music and to
reinstitute an alliance with the consumer population. Partnerships with civil and criminal
authorities and agencies; heavy lobbying to convince members of Congress to propose
bills related to music piracy; educational campaigns with universities, secondary schools
and media outlets; and the initiation of lawsuits against individuals who facilitate music
piracy through peer-to-peer ﬁle exchange programs are also methods to enroll individuals
into legitimately purchasing music, thereby acting in a manner that supports their
position. Finally, mobilization is sought as the industry attempts to dictate what is in the
best interests of the commodity of music and those who seek to obtain it, for the purposes
of bringing all divergent stances into one stable and cooperative position. Both of these
positions illustrate the collective process of translation as articulated by Callon

(1987:224), but from contrasting perspectives.

248

Implications

Now that this theoretical foundation has been laid and applied in a practical
manner to the crime of digital music piracy, what is the next step? How can we proceed?
In a piece on the social construction and evolution of technological systems, Hughes
(1983279) articulates the concept of a “reverse salient,” which is an element that has
lagged behind or appears out of phase with the development of a system, and which must
be addressed to ensure the continued viability of the phenomenon. Traditionally, salients
are protrusions in geometrical ﬁgures, lines of battle, or an expanding weather front
(Hughes, 1987:72). A reverse salient, put simply, is something that protrudes backwards
— in the opposite direction of which it should go. The primary idea is that problems arise
in networks that are not in sync or in step with the remainder of the entities, and which
must be corrected to ensure continued stability. This is very similar to the previously-
described concept of “enrolment”, except that the actors actually stray (rather than just
pose the threat of straying). For phenomena which must be curtailed or obviated (such as
music piracy), it appears that policy should enhance or exaggerate the likelihood of
reverse salients, which will ultimately effectuate a breakdown in the network if not
addressed in a timely fashion. Indeed, there are various ways to cripple the network and
ANT would argue that policy solutions intended to address criminality should identify
those elements most susceptible to being interessed, enrolled, and mobilized.

For instance, how and to what degree the legal system affects the action or
inaction of potential and actual music pirates is extremely relevant when considering
policy issues. The fact that laws are in place which outline the illegality of copyright

infringement and which prescribe punitive sanctions for their transgression is largely

249

meaningless unless interpreted in the light of their effect on the populace for who they
were written. The lack of jurisdictional authority in the boundless realm of cyberspace,
the difﬁculty to coordinate efforts between private- and public-sector organizations, the
insufﬁcient ﬁscal and temporal resources of law enforcement, and the relative
unfamiliarity of police in addressing hi gh-tech deviance largely render the legal system’s
“bark” absent of any meaningful “bite.” If they are useful in some capacity, certain
legislative mandates and penalties may have more of a deterrent effect than others; it
would be crucial to determine which are more potent in their impact on a population, and
how the strength of actions (or inactions) between statutes (a nonhuman actor) and
persons (a human actor) might be indirectly or directly augmented (or, at the very least,
preempted from being weakened) on account of other actors (and perhaps even other
networks representing other phenomena).

Without the interaction of actors, the networked pattern of elements that
comprises music piracy would collapse. Likewise, without the presence of legal
stipulations condoning or condemning certain behaviors, the concept of “crime” would be
nonexistent and the phenomenon would verily lose its illegality — and its attractiveness
for study and response. This underscores the mutually dependent and sustaining
inﬂuence that each actor (both human and nonhuman) has on others, albeit to varying
degrees. These notions also depict the possibility of some entities truly fulﬁlling the role
of reverse salients — a role that is seemingly poised for exploitation by those who seek to
combat a particular criminal activity.

Law (1987b: 131) asserts that the network is only inﬂuenced by those actors who

are able to make their presence known. That is, the actor must act in a way that results in

250

a deliberate and concerted force to inﬂuence the creation or maintenance of the network.
For example, victims of music piracy have historically done very little to exact change in
the frequency and prevalence of the crime. Through vociferous condemnation of the
activity, the music industry has asserted their position, but has not sought to markedly
inﬂuence the elements which are conducive to its occurrence. Admittedly, they have
championed tough criminal penalties and have partnered with ISPS to identify and bring
to justice certain transgressors (e.g., Healey & Leeds, 2003), but their impact has not
been felt on the network that represents music piracy in a substantive way and may
arguably have augmented the phenomenon due to their actions.

The criminal justice system has similarly failed to act on the network in such a
way that music piracy is reduced or stiﬂed. Despite the fact that additional bills are being
proposed to give law enforcement the authority and ability to combat digital intellectual
property theft (Dean, 2003a, 2003b), no substantive improvement has resulted. The
system has also occupied a position diametrically opposed to the ethos that permeates the
social and organizational spheres. Accordingly, a supportive infrastructure to successfully
preempt the illegal behavior is conspicuously missing because of the absence of strong
links between the criminal justice actors, the participants, the technology of MP3, the
associated computer—related elements, and to social, organizational, situational, and
economic elements.

The utilization of nonhuman actors to detect copyright violations and discover
and apprehend the transgressors appears to be most promising for law enforcement. For
instance, surveillance software solutions have been created by a handful of private sector

companies to search for and document evidence of intellectual property theft, trigger the

251

dissemination of take-down notices to the responsible parties, and to monitor for
copyright compliance”. Partnership with these businesses - whose clientele tend to be
the creators and owners of copyrighted material - may lead to strategies that police can
implement. The system must also hasten a sea-change in the predominant social
sentiment towards music piracy (perhaps through general deterrence strategies that
demonstrate their Skill and interest in decreasing participation), and must work to inhibit
the construction and stability of organizational contexts that provide venues supportive of
the crime (perhaps through logging and tracking illegal behavior in these environments).
The elements and consequently their interconnections must be redeﬁned, possibly
introducing new elements and relationships and transforming their overall purpose or
function to one that coheres with criminal justice objectives (Callon, 1987:93).

The ﬁnal ANT term deﬁned here is “transformation,” which means the
modiﬁcation or cessation of a network by testing the resistance of each element to its
limit (Callon, 1987297). Perhaps encouraging resistance among one or more of the
heterogeneous components that comprise a particular crime like music piracy — or
completely breaking the link between them - may cripple the architecture and
consequently preempt the occurrence of the phenomenon. If the individual elements
possess disassociative tendencies and hostile forces which must be overcome, the
objective must be to augment the damaging nature of those hostile forces, or prevent the
network from thwarting their tendencies (Law, 1987b). To control or prevent the
occurrence of a phenomenon, then, the durability of the network must be questioned and

compromised. Notably, once the contribution of one element is crippled or removed, the

 

69 See e.g., BayTSP (www.baytsp.com) and Envisional (www.cnvisional.com) for more information.

252

other elements will adjust accordingly; such adjustment, from the point of view of those
aiming to preclude the phenomenon, must somehow be insufﬁcient to sustain the
network. Future research Should build upon this theoretical conceptualization by testing
the resistance of nonsocial and social actors that participate in the network. Attempted
transformations will determine if their removal or weakening has any substantive effect
on curtailing or terminating the criminal activity.
SUMMARY

In order for music piracy to occur, a host of heterogeneous elements or nodes are
essential for inclusion in a network architecture, ordered and organized in a fashion that
makes it possible for music piracy to occur. All of these entities must be considered as
important constituent parts of the phenomenon, and all of their relationships and
interactions are crucial for its existence. American inventor Thomas Edison thought
carefully and critically about engineering, economics, political issues, and social
ramiﬁcations when creating the electric light (Hughes, 1983). His success was largely
due to his ability to shape these elements into a cohesive and stable network. This
network was the aggregation and juxtaposition of constituent parts that played (and
would play) a role in the construction and maintenance of the incandescent lamp. Such a
scenario effectuated the phenomenon as it was, and also introduced new technical and
social relations (that heretofore did not exist) into the picture. Intemet-related criminality
such as music piracy has unquestionably done the same.

The theoretical and methodological richness of ANT lie in its refusal to separate
the social from the technical, its broad-based perspective on contributive factors, its

ability to render irrelevant any distinction between the macro- and the micro-level, and its

253

illumination of points and links where interventions can occur to weaken or break the
cohesiveness of the network. This chapter has sought to suggest and present a different
way to conceptualize high-tech wrongdoing. Empirical testing of its application must
now follow to clarify whether the theory is a suitable framework to use in efforts to

understand and respond to crime and deviance.

254

CONCLUSION

The present study has sought to test the applicability of three “general”
criminological theories to online intellectual property theft in the form of digital music
piracy. It has determined that self-control and social learning theory are extensible to
crimes that are nontraditional in content and in context, while general strain theory is not.
Furthermore, it has fostered a comprehension of the factors that result in adoption of the
behavior and assimilation into the social group that supports and perpetuates it. Policy
solutions intended to curb the prevalence of rampant copyright violation have also been
suggested and discussed, and a new model in which to conceptualize Internet crimes such
as digital music piracy has been proposed. This model views the phenomenon through
the lens of Actor-Network theory and heterogeneous engineering, and gives equal
attention to both the social and the technical aspects of computer criminality.

The digital music phenomenon has achieved unprecedented media coverage,
public accolade and adoption, and legislative attention Since its introduction into the
popular culture. The technology and associated infrastructure which has developed to
support its growth and pervasiveness has augured great promise for the future of
intellectual property distribution. However, the wild proliferation of copyright
infringement in the face of traditional business models and extant copyright law has
demonstrated a weakness which may negatively affect the innovation, development, and
value of intellectual property and creative works as the Internet plays a larger role in our
information-based society. Thus, it is important to approach the issue in an informed
manner so that negative ramiﬁcations can be foreseen and prevented, or identiﬁed and

curtailed. Further, intelligent analysis of the expropriation of music online will shape

255

determinations of how valuable digital works of other kinds can and Should be
disseminated over the Internet. As society becomes further ensconced in the Information

Age, this is critical.

256

APPENDICES

257

APPENDIX A - SURVEY INSTRUMENT
Questionnaire on Participation in and Attitudes Towards MP3s

Thank you for taking the time to ﬁll out the following questionnaire. Its purpose is to obtain
an understanding of college students’ perceptions of, familiarity with, and attitudes toward, their
use of digital music (MP3s) ﬂom the Internet.

Your input is valuable to us and will aid in:

l. assessing the extent to which the Internet has become an integral part of students’ lives.
2. examining your ideas of acceptable and unacceptable conduct on the Internet.

Please select an answer for each of the following questions based on your personal
circumstances/knowledge. Also, don’t spend too long on any one statement; just input your initial
reaction on the scantron form provided.

This survey should take approximately 20 minutes to complete. Risks to subjects in this study are
minimal, and only concern emotional or psychological harm when requested to contemplate and
reveal participation in certain deviant behaviors. With regard to any and all information provided
by you as a resnondent, your nrivacy will be protected to the maximum extent allowable by law.

This survey is completely voluntary and anonymous. You are ﬂee to skip any question. Do not
write your name or any other identifying information on the questionnaire or scantron.

I would sincerely appreciate Lour honest answers in order to obtain a reliable measure.

My methodology is as follows: I am going into an assortment of classrooms ﬂom varying
disciplines with the permission of the professor in charge, and administering this questionnaire. I
will verbally inform the students of the conﬁdentiality and anonymity of the survey, as well as the
fact that participation is voluntary. This information is also provided at the top of each
questionnaire. Additionally, when I introduce myself to the classes I visit, I will explain the
purposes of the research, the expected time it should take for them to ﬁll out the survey provided,
and that there is no cost associated with participating except for the time spent in composing a
response. I will also make potential respondents cognizant that only group totals will be
consolidated and released at the culmination of the project. This is primarily to protect the rights
of the respondents and to garner a reliable cross-section for measuring the relevant constructs.

. If you have questions about the study, please feel ﬂee to contact Dr. Mahesh Nalla by phone:
(517) 355-2228, fax: (517) 432-1787; email: nalla@msu.edu, or regular mail: 560 Baker Hall,
East Lansing, MI 48824. In case you have questions or concerns about your rights as a research
participant, please feel ﬂee to contact Ashir Kumar, MD, Michigan State University's Chair of
University Committee on Research Involving Human Subjects by phone: (517) 355-2180, fax:
(517) 432-4503, email: ucrihs@msu.edu, or regular mail: 202 Olds Hall, East Lansing, MI 48824.

You indicate your voluntary agreement to participate by beginning this questionnaire.

FEEL FREE TO TEAR OFF AND TAKE THIS COVER SHEET HOME IN CASE YOU
HAVE ANY QUESTIONS IN THE FUTURE.

258

Negative events often occur in our lives. For the following questions, please answer A for

TRUE and B for FALSE

 

OVER THE LAST SIX MONTHS, I HAVE:

TRUE

FALSE

 

1. Received a bad grade in a class

 

2. Broken up_ with an intimate partner

 

3. Experienced weight gain or loss

 

4. Been ﬁred or laid off ﬂom a Eb

 

5. Had money problems (i._e., had difﬁculty paying tuition, rent, bills)

 

 

6. Been a victim of a crime

 

>>>>>>

 

wwwwww

 

Take a moment to reﬂect on your personality, and for each of the following questions, please
respond as follows: A = STRONGLY DISAGREE, B = DISAGREE, C = NEUTRAL, D =

AGREE, E = STRONGLY AGREE

 

 

 

 

 

 

 

 

 

 

 

 

 

SD D N A SA

7. I often do what brings me pleasure here and now. A B C D E
8. When things get complicated, I tend to quit or withdraw. A B C D E
9. I ﬁnd no excitement in doing things I might get in trouble for. A B C D E
10. I try to look out for others ﬁrst, even if it means making things A B C D E
difﬁcult for myself

11. I don 't lose my temper very easily. A B C D E
12. I feel better when I am on the move rather than sitting and A B C D E
thinking.

 

Below are some questions related to certain behaviors in which some students have

participated. For each of the following questions, please respond as follows: A = TRUE, B =

 

 

 

 

 

 

FALSE

OVER THE LAST YEAR, I HAVE: TRUE FALSE

13. I have skipped more than 10 class periods in the past year. A B

14. I have lied to a professor/instructor either via email, telephone, or in A B
erson at least once in the past year.

15. l have plagiarized on a school assignment at least once in the past A B

year.

16. I have drank alcohol before I turned 21. A B

17. I have driven a vehicle while under the inﬂuence of alcohol at least A B

 

once in the past year.

 

 

 

Take a moment to reflect some more on yourself, and for each of the following questions,
please respond as follows: A = STRONGLY DISAGREE, B = DISAGREE, C = NEUTRAL,

D = AGREE, E = STRONGLY AGREE

 

 

18. I am optimistic about my future.

 

19. I have difﬁculty maintaining long-term relationships.

 

 

20. I actively expect the best ﬂompeople and situations.

 

21. My emotional life is unstable.

 

22. I am able to express the feelings I have, whether happy, sad, angry,
ﬂustrated, or confused.

m
>>>>>U

mmmmmﬁ’,’

 

23. I am not comfortable with myself when around others.

 

24. I have difﬁculty achieving long term goals.

 

 

 

25. I am happy.

 

>>>

meow wwwwwu

 

 

000 000002

 

CCU UUUUU>

 

111111111

 

259

 

 

 

 

26. How many student organizations (like the Debate Team, Campus Crusade, Outing Club, etc.)
did ‘ ' in over the ?

    

     

0
l
2
3
4

01' more

27. How many Sports did you regnlarly participate in (including running/working out) over the

0
l
2
3
4

01' more

 

28. On average each month, how many times do you participate in religious activities such as
a or ' session?

       

O
1
2-3

4-5
6 or more

29. I have a amount of friends in the area.
A. V low
B. Low

C. moderate
D. '
E. V

 

30. I would rate self-esteem as:
A. V low

B Low

moderate

C
D.
E

V

 

31. On a scale of 1-5 (with 1 = "cold, distant, and completely dysfunctional" and 5 = “healthy and
warm how would rate the of ' with s or s

       
  

         

32. On a scale of 1—5 (with 5 = very strongly), how strongly have your parents shaped your
' on life?

 

260

33. On a scale of 1-5 (with 5 = very strong”). how strongly have your friends shaped your
' on life?

    

261

Consider your participation with MP3s, and for each of the following questions, please respond as
follows: A = STRONGLY DISAGREE, B = DISAGREE, C = I DO NOT PARTICIPATE WITH MP3s,
D = AGREE, E = STRONGLY AGREE

 

SDDM-FBASA

 

34. It is a great beneﬁt to sample new music through MP3s. A B C D E

 

35. It is a great beneﬁt to be able to transfer assorted MP3s onto an A B C D E
audio/data CD or a portable MP3 player so that I can have music on-the-
__go

 

36. I feel practically no threat of sanction or punishment for use of MP3s.

 

WWW

37. It makes me feel good to download a song that l have wanted.

 

38. I have learned the techniques of using MP3s ﬂom television or print
media.

 

W

39. l have learned the techniques of using MP3s ﬂom online sources (web
pages, chat rooms, etc).

 

40. I was introduced by another person online to MP3s.

 

41. MP3 use is excusable and justiﬁable.

 

42. One of the reasons I download MP3s is because 1 will not purchase the
music.

 

 

> >>> > >>>
0 000 0 000
U UUU U CCU

W WWW

43. One of the reasons I download MP3s is because I feel the recording
industry has been overcharging the general public for music tapes and
CDS.

 

>
W
O
U
111

44. One of the reasons 1 download MP3s is because many musicians and
the recording industry make millions of dollars anyway, and downloading
MP3s of their songs does not really cut into their income.

 

45. My ﬂiends support my MP3 usage.

WW

 

46. I associate with others in real life (offline) who are supportive of MP3
usage.

 

47. I have introduced others in real life (ofﬂine) to MP3s.

 

48. l was introduced by another person in real life to MP3s.

 

49. I am embarrassed that I use MP3s.

 

50. I am proud that I use MP3s.

 

51. l associate with others online who exchange MP3s with me.

 

52. I do not care what others think of me.

 

53. I enjoy participating in a new, controversial technolgy.

 

54. I feel good about myself if I am able to help or beneﬁt someone with
an MP3.

 

55. I have learned the techniques of using MP3s ﬂom my family

 

56. l have learned the techniques of using MP3s from my ﬂiends

 

57. In general, I tend to do what the majority does.

 

58. It is a great beneﬁt to me to be able to access music ﬂeely.

 

59. MP3s do not really hurt musicians or the record industry.

 

60. Musicians and the record industry should embrace MP3 technology
and use it to their advantage.

 

61. One ofthe reasonsldownload MP38 isbecause I cmpurchase the
music.

 

W W WWWWWW WWWWWWWW

62.0ne of the reasons I download MP3s is because 1 think music should be
ﬂee

 

63. One of the reasons 1 download MP3s is because people 1 know do it.

 

64. One of the reasons I download MP3s is so 1 can sample new music
without having to buy the CD.

 

WWW

65. People 1 know ofﬂine (in the real world) like me, appreciate me, or
beneﬁt ﬂom me because I use MP3s.

 

66. People I know ofﬂine (in the real world) ﬂown on my use of MP3s

 

67. People I know online ﬂown on my use of MP3s.

 

68. People 1 know online like me, appreciate me, or beneﬁt ﬂom me
because 1 use MP3s.

 

> >>> > >> > > >>>>>> >>>>>>>> >>
W WWW W WW W W WWWWWW WWWWWWWW WW
0 000 0 00 0 0 000000 00000000 00
U CCU U DO U U UUUUUD UUUUUUUU DU

W 'WWW

69. Transferring MP3s in general should be allowed as long as individuals
use the music for personal purposes, and are not making money off of
them.

 

 

 

 

 

 

 

U
m

70. Use of MP3s is a "cool” my. A B C

 

 

262

Regardless of whether you participate with MP3s, please consider the situations and

circumstances which would make you more likely to do so. For each of the following questions,
please respond as follows: A = STRONGLY DISAGREE, B = DISAGREE, C = NEUTRAL, D =

AGREE, E = STRONGLY AGREE

 

I WOULD BE MORE LIKELY TO DOWNLOAD/UPLOAD MP3s .....

SA

 

71. if I could not afford the purchase price of the music on CD?

E

 

72. since numerous sources offering MP3s for free download are readily
available online?

WWW

E

 

73. since there are no clear-cut rules, laws, regulations, or even guidelines
when it comes to MP3 ﬁle cxchamxs?

 

74. if all my ﬁiends and classmates were doing it?

 

75. if it were known that the recording industry "could afford it" and would
never miss the tiny amount of proceeds lost ﬂom just a few MP3s here or
there?

CA
>> > >>U

WW

00 0 002

CC C 00>

 

76. if it were known that law enforcement agencies, universities, and
authorities in general couldn't care less about MP3 ﬁle exchanges, lack
adequate abilities to detect or combat the activity, and have bigger things to
worry about?

>

0

U

 

77. if it were held that the music industry, to some extent, deserves to have
their music distributed freely online considering the fact that they rip off
consumers?

 

78. if it were held that no one is really getting hurt ﬂom the downloading
and uploading of MP3s online?

 

79. because any rules or laws that seek to prevent individuals from
exchanging MP3s are misguided and ill-conceived?

 

80. because hardly anyone has been caught or punished or has been subject
to even the slightest repercussions for downloading and/or uploading MP3s
online?

>

0

U

 

81. if I needed the music wouldn‘t be able to obtain it any other way?

 

82. if a family member, friend, or sigriﬁcant other needed the music?

 

83. if the music will be used to complete a project for school or work, or to
achieve other school -related and career-related goals?

 

84. since it is okay if I do something questionable every now and then - it is
better than a frequently dishonest person engaging in questionable deeds
over and over again?

> >>>

W WWW

0 000

U CCU

W WWW

 

85. because I deserve somethirg for ﬂee sometimes?

 

86. if it were prevalent all over the Internet, and if a lot of people were
doing it?

WW

 

87. if it were held that no one else seems to care whether or not they get
calmht?

['11

 

88. if it were held that others are beneﬁting ﬂom it, and so why shouldn‘t 1?

WW

 

89. because I can't afford to waste money on a music CD that might only
have 1 or 2 good songs?

 

90. because without the ability to evaluate the music, 1 will not be able to
determine if 1 really want to purchase it on CD?

 

91. because the anonymous nature of the Internet affords privacy and
somewhat of a shield ﬂom detection; and so, why not take advantage?

 

 

92. because no one really cares about what I do online - it is just too
removed ﬂom the "real world"?

 

> > > >> > >>

 

W W W WW W WW

 

0 0 0 00 0 00

 

U U U DO U DU

 

 

263

 

Consider your CURRENT participation with MP3s, and for each of the following questions,
please select from the answer choices provided.

93. How many MP3 ﬁles have you personally downloaded in the last week?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A. 0

B. 1-5

C. 6—10

D. “-20

B. More than 20

94. How many MP3 ﬁles have you personally downloaded in the last month?
A. 0

B. 1-25

C. 26-50

D. 51-100

E. More than 100

95. How many MP3 ﬁles have you personally downloaded since the beginning of 2003?
A. 0

B. 1-10

C. 1 1-50

D. 5 1-250

B. More than 250

96. How MP3s do on download month?

   

        

     

0
1-25
2650

51-100
More than 100

Consider your participation with MP3s exactly ONE YEAR AGO from today, and for each
of the following questions, please select from the answer choices provided.

97. Approximately how many MPB ﬁles did you personally download in an average week exactly
one ?

0

1-5

6—10
11-20
More than20

 

98. Approximately how many MP3 ﬁles did you personally download in an average month
one ?

    

0
1-25
26-50

51-100
More than 100

264

Consider your participation with MP3s exactly TWO YEARS AGO from today, and for
each of the following questions, please select from the answer choices provided.

99. Approximately how many MP3 ﬁles did you personally download in an average 1115315 exactly
9

0

1-5

6-10

1 1-20

More than 20

M0
A.
B.
C.
D.
E.

 

100. Approximately how many MP3 ﬁles did you personally download in an average month
two ?

       

0
1-25
26-50

51-100
Morethan 100

Consider your participation with MP3s in years past, and for each of the following
questions, please select from the answer choices provided.

 

101. How many MP3 ﬁles did you personally download in 2002?

 

 

 

 

 

 

 

A 0

B. 1-10

C. 11-100

D 101-1000

B More than 1000

 

I—I

02. How my MP3 ﬁles did you personally download in 2001?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A. 0

B. 1-10

C. 11-100

1). 101-1000

E. More than 1000
103. How many MP3 ﬁles did you personally download in 2000?
A. 0

B. 1-10

C. 11-100

D. 101-1000

B. More than 1000

 

 

 

104. How many total complete music albums in MP3 format have you obtained online?

 

 

 

 

 

 

 

 

A. 0

B. 1-5

C. 6-10

D. 11-20

B. More than 20

 

265

105. How many total MP3s have you downloaded over the course of your life thus far?

 

 

 

 

 

 

 

A 0

B. 1-100

C. 101-500

D 501-2000

E 2001 or more

 

 

106. Of the total MP3s you have, what percent are ﬂQI personally created ﬂom CDS you own, or
are NOT of songs that you deﬁnitely own on CD?

0% (they are all ﬂom CDs I own or are of $035 that I own on CD)

l%-30% (a small amount are not ﬂom CDS I own or of songs that I own on CD)

31%-60% (a moderate amount are not ﬂom CDs I own or of songs that I own on CD)
61%-90% (a large amount are not ﬂom CDs I own or of songs that I own on CD)

Over 90% (almost all are not ﬂom CDs I own or of sggs that I own on CD)

 

 

 

 

 

 

1119.09”?

 

 

 

 

107. How many hours each week do you spend lookinmr or obtaining MP3s?

 

 

 

 

A I don't look for or obtain MP3s (zero hours)
B. Less than 1

C. 1-2 hours

D 3-4 hours

B 5-6 hours

 

 

 

 

108. The breakdown of my time spent online downloading MP3s (to your computer) and
ﬂom is .

I do not with MP3s

0% of the time ' and 100%

25% of the time and 75%

     

75% of the time and 25%
100% of the time and 0%

109. I have:

A. Created an audio CD ﬂom MPB ﬁles

B. Made an MP3 ﬁle myself (ﬂom an audio CD or ﬂom another sound source)
C. Both of the above

D. None of the above

 

 

 

 

 

 

 

 

 

110. With my MP3 ﬁles, I do the following:
. Listen to them on my computer

 

A
B. Listen to them after burning them to CD or transferrinLthem to a portable MP3 player
C. Both of the above

D

E

 

 

None of the above (but I do have MP3 ﬁles)
I don't have any MP3 ﬁles

 

 

 

 

 

 

111. With my MP3 ﬁles, I do the following
. Share them with others

 

A
B. Sell them

C. Both of the above
D

E

 

 

None of the above (but I do have MPB ﬁles)
I don't have any MP3 ﬁles

 

 

 

 

 

112. Do you believe that receiving or providing MP3s should be illegal?
A. Yes
B. No

 

 

 

 

 

 

266

 

113. From ' ' MP3s is:

  

Unethical/Immoral but still

Unethical/Immoral and

but Ethical/Moral and
' and

114. Do you refrain ﬂom obtaining MP3s because you believe it is illegal?

 

 

 

 

 

 

A. Yes, I reﬂain because I believe it is Mal

B. No, I participate even though I believe it is illegal

C. Yes, I reﬂain but not because I believe it is illegal, but for other reasons such as the fact it
hurts artists/bands, recording labels, and the music industry, or the fact that it does not sit
well with me

D. No, I participate because I do not believe it is iuﬁal

E. I do not obtain MP3s because I am not familiar with them or have no need/desire to do so.

 

 

 

For each of the following questions, please select from the answer choices provided.

 

 

 

 

 

 

 

 

l 15. Race:
Caucasian/White
African American/Black
Asian/Paciﬁc Islander
Other

1 16. Sex:

A. Female

B Male

117.

17 or
18-19
20-2 1
22-23
24 or older

 

1 18. Year of Studies:

119. What is your parents' annual household income?

Freshman

Junior
Senior
Graduate Student

 

 

 

 

 

 

 

A. $0 to $19,999

B. $20,000 to $29,999
C. $30,000 to $39,999
D. $40,000 to $49,999
E. $50,000 or more

 

 

267

 

120. My employment (job) status:
. I do not have ajob

 

I work approximately 10 hours a week

 

 

I work approximately 30 hours a week

 

 

 

 

A
B.
C. I work approximately 20 hours a week
D
E

I work approximately 40 hours a week

 

121. I live inan:
A. Residence Hall dorm
B.

C. Off or House
D. Other

 

122. Where I reside during the school year (dorm room, apartment, house, etc.), I am generally
connected to the Internet via:

 

A. high speeds, on the Ethernet network or with a Cable or DSL connection

 

B. slower speeds, where l dial in through my telephone line using my computer modem

 

 

 

 

C. I cannot connect to the Internet at my place of residence during the school year

 

123. In the following list, please count up the number of activities for which you regularly
use the Internet, and answer accordingly.

_Email, Chat/IRC

_Research for school work

_File Transfer

_Using the Newsgroups

_Product and Travel Information

_Online Stock Trading

_Online Shopping

_Online Auctions

_Online Games

_Online Banking

_To collect inforrmtion related to news, sports, or the weather
_To collect information related to personal interests and hobbies
_Web Design

0 items

1-2 items

3-5 items

6-8 items

9 or more items

 

A.
B.
C.
D.
E.

268

124. In the following list, please count up the number of activities that you have ever done
online, and answer accordingly.

_changed my browser's "startup" or "home” page

_nmde a purchase online for more than $100

_participated in an online game

_participated in an online auction

_changed my ”cookie" preferences

_participated in an online chat or discussion (not including email, ICQ, or AOL Instant
Messenger, or similar instant messaging programs)

_listened to a radio broadcast or music clip online

_made a telephone call online

_created a web page

_set up my incoming and outgoing mail server preferences

0 items
1-2 items

3-5 items
6-8 items
9 or more items

 

Q the "SECTIQN" section of your Scantron, in the section where you would usually record
your identifying information (DO NOT do so on this survey as it is anonymous), please

bubble in one of the following three-digit numbers to indicate the college in which your
major is housed.

001. College of Agriculture and Natural Resources
002. College of Arts and Letters

003. The Eli Broad College of Business/Graduate School of Management
004. College of Communication Arts and Sciences
005. College of Education

006. College of Engineering

007. College of Human Ecology

008. College of Human Medicine

009. James Madison College

010. College of Natural Science

011. College of Nursing

012. College of Osteopathic Medicine

013. College of Social Science

014. College of Veterinary Medicine

015. I have not decided on a major as of yet

016. I do not know where my major is housed
017. None of the above

269

APPENDIX B: DISENTAN GLIN G SOCIAL LEARNING THEORY VARIABLES

Table A: Factor Analyses of all Social Learniﬁ Theory Variables

 

 

1 2 3 4
Differential Association (05774)
My friends support my MP3 usage .823 -.086 .040 -.003
I associate with others in real life (ofﬂine) who are
supportive of MP3 usage .750 -.l36 -.005 .089
I was introduced by another person in real life to MP3s. .585 -.183 -.076 .588
I have learned the techniques of using MP3s ﬂom my
ﬂiends .563 -.050 .065 .596
Differential Reinforcement (a=.862)
It is a great beneﬁt to sample new music through MP3s. .814 -. 147 .125 -.269
It is a great beneﬁt to be able to transfer assorted MP3s
onto an audio/data CD or a portable MP3 player so that I
can have music on-the-go .794 -.115 .142 -.288
It makes me feel good to download a song that I have
wanted .735 -.019 .137 -.141
It is a great beneﬁt to me to be able to access music
ﬂeely .762 -.107 .047 -.134
Imitation (05525)
I have learned the techniques of using MP3s ﬂom
television or print media -.085 .648 .494 .141
I have learned the techniques of using MP3s ﬂom online
sources (web pages, chat rooms, etc) .102 .597 .590 -.014
I associate with others online who exchange MP3s with
me .053 .465 .292 .043

Deﬁnitions (a=.658l
One of the reasons I download MP3s is because I I"will

not“ purchase the music .183 .517 -.425 .137
One of the reasons I download MP3s is because I feel

the recording industry has been overcharging the general

public for music tapes and CDs .407 .467 -.360 -.237
One of the reasons I download MP3s is because many

musicians and the recording industry make millions of

dollars anyway, and downloading MP3s of their songs

does not really cut into their income .326 .5 79 -.375 -.069
One of the reasons I download MP3s is because I think
music should be ﬂee .199 .558 -.301 .065

 

All ﬁfteen social learning theory variables a=.773

270

Table B: Differential Association and Differential Reinforcement Factor Analysis

 

 

 

 

 

Component 1

DA - My friends support my MP3 usage .828
DA - I associate with others in real life (ofﬂine) who are supportive of MP3 usage .760
DA - I was introduced by another person in real life to MP3s. .605
DA - I have learned the techniques of using MP3s from my friends .574 .
DR - It is a great beneﬁt to sample new music through MP3s. .830
DR - It is a great beneﬁt to be able to transfer assorted MP3s onto an audio/data CD or a
portable MP3 player so that I can have music on-the-go .808
DR - It makes me feel good to download a song that l have wanted .740
DR - It is a great beneﬁt to me to be able to access music freely .771
a=.878
Table C: Imitation and Definitions Factor Analysis

Component 1 Component 2
I have learned the techniques of using MP3s from television or print .172 .825
media
I have learned the techniques of using MP3s from online sources (web .185 .830
pages, chat roonn, etc)
I associate with others online who exchange MP3s with me .209 .554
One of the reasons I download MP3$ is because I think music should .638 .254
be free
One of the reasons I download MP3s is because I *will not‘ purchase .656 .153
the music
One of the reasons I download MP3s is because I feel the recording .736 .111
industry has been overcharging the general public for music tapes and
CDs
One of the reasons I download MP3s is because many musicians and .775 .203

the recording indusz make millions of dollars anyway, and
downloading MP3s of their songs does not really cut into their income
a=.659

271

APPENDIX C — SAMPLE LETTER FROM THE RIAA TO CORPORATIONS

It“
\"

 

   

October 25, 2002

[COMPANY NAME]

[TITLE FIRST LAST NAME]
[ADDRESS]

[CITY], [ST] [ZIP]

Dear [CEO/President]

We write on behalf of the creative community regarding the use of wrporate networks to infringe
copyrights using online peer-to-peer (P2P) systems such as Kazaa, Grokster, iMesh and Gnutella Because
this letter references legal issues pertaining to the use of corporate networks, we ask that you forward a copy
of this to your company’s General Counsel/Chief Legal Oﬂicea

We, like you, invest our time and resources to create products or services that our customers will ﬁnd of
value. And, like you, we do not want our valuable assets to be stolen. Therefore, we must aggressively
enforce our rights in cases of copyright infringemart Many companies have already implemented employee
policies and technical measures to prevent copyright infringements on their corporate networks. If your
company is among this group, we commend your action. If it is not, we urge you to take whuever steps
necessary to ensure that your network is not being misused to nge copyrighted works.

By now, you and your colleagues are well aware that software piracy is illegal and ﬂrat your company is at
risk when your employees copy and use software without buying additional limes. You may not have thought
about unauthorized distribution of music and movies in the same way. But, in hot, allowing anployees to use
your corporate network to illegally distribute copyrighted music and movies is no different ﬁ'om software piracy.

We have recently become aware that piracy of music, movies, and other creative works is taking does at a
surprisingly large number of companies - including a number of For-time 1000 companies. It appeas that many
corporatenetworkusers aretakingadvantage ofﬁrsthrtanetconnectionsatworkbypubliclyuploedingand
downloadinginﬁingmgﬁlesonPZPservices,arrialsocishibuﬁngarrdstoringsuchﬁlesoncorpcateintranets.
In the past few decades, companies have built sophisticated digital networks - at great expense to shareholders
and investors—to ensure competitiveness in an increasingly networked world. The use of your digital network to
pirate music, movies, and other copyrighted works both interferes with the business purposes ‘yurr network wu
built to serve and mbjects yom' employees and your company to signiﬁcant legal liability unda the Federal
copyright law. A great many irrstitutiom—ccrporations, government agencies and universities - have prevented
unlawﬁrluses,savedagreatdeal ofmoney onbroadbandfeesarrdenhancedthe securityoftha'rintemal
networks by conﬁguring ﬁrewalls to pevent access to particular networks or limiting the use of certain software
or ﬁles. We are aware ofa few companies that are develop‘ng tools thatmay assistin detecting and/or blocking'
infringing activityonadigital network Foryourreferarce,wehaveattaehedalistofthesetypes ofeomparries.

272

Page 2

Copyright Letter
Octoba‘ 25, 2002

The creative community welcomes technological innovations such as peer-to-peer technologies. Legally
downloading, streaming, and otherwise experiencing music, movies, and other forms of entertainment over
networked environments is exciting, and our companies are working aggressively to make this experience
increasingly available. However, using technology to steal music and movies is no different from walking
into a store and shoplifting a CD or DVD.

The creative community commends those companies that have already implemented employee policies and
technical measures that serve to eliminate copyright piracy from their networks. We stand ready to work with the
corporate community to ﬁnd ways to eradicate digital piracy of copyrighted works on business networks, and
would welcome the opportunity to assist your company in this eﬂort

W

  

Jack Valenti
, President and CEO,
Recording Industry Association of America Motion Picture Association of America
‘
Edwwd P. Murphy Rick Carries
President and CEO, President.
National Music Publishers’ Association The Songwriters Guild of America
Ibschoakds’gmia Jim'sldlerd:
RIAA MPAA NMPA SGA
1330 Connecticut Ave, NW 1600 Eye Street, NW 475 Park Avenue South 1500 Harbor Blvd.
Suite 300 Washington. DC 20006 29th Floor Wednwken, NJ 07086
Washington, DC 20036 www.mpsaorg New York. NY 10016 www.mgwritas.org
wwwriaacom wwwnnrpaorg

273

ATTACHMENT

Companies that offer management tools for corporate networks:

(1‘ ' rat'
Vance Ikezoye
CEO
Audible Magic Corporation

985 University Ave, Suite 35
Los Gatos, CA 95032

Phone: (408) 399-6405 ext. 106
www.audiblemagic.com
v_ikezcye@audiblemagic.com

WW

Pierce Lcdbetter

CEO

Transparency Software, L113
1019 Mission Street

San Francisco, CA 94103-2812
Phone: (415) 294-4949
wwwpeervucom
emsil:pierw@peervueom

)Yshsmlne.

John Carrington

CEO

Websense Inc.

World Headquarters

10240 Sorrento Valley Road
San Diego, CA 92121

Tel: (800) 723-1166
wwwwebsenseccm
ssles@websense.corn

274

 

 

r

pagan 28:85.00
......... I...
0.338 Awe—Ewe; nausea
zeros—«am no? .naekmmaoz .EE
420 5.: Buscom
Baéom 8&5:— Due—F—

J

 

£22.. tom
as. 52%
53:50

   

Ii...

 

 
    

 

 

 

 

35> 05 925.5 85
Ammo—0:50: E93
5:555: can 58::

     

8.58.5 05 259.8
8 oﬁovcu 0:3 £88 :35:

30>» he gonna—Emac
.8625 05

2 233:8 :25»

8.55.0 553.52

5

 

ﬂag & ESQ he 5.53635
96:3 68:? use—8&8 a 2
8.5580 35 828 58382

 

 

 

     

        

          
      

 

 

MZOPU<h A<ZO_.—.<N_Z<U¢O\U_EOZOUm—Jcsucw t
>U<~=m UmmD—Z deg—"HG ho UZ—MHHZ—UZH mDOm—ZH

Emhm>m HOP—.9... J<Z=>=¢U

f

COM—HRH: HEP "D EQZﬂmm<

  

 

275

REFERENCES

276

REFERENCES

A & M Records Inc. et al. v. Napster Inc. (2001). No. 00-16403. Retrieved February 12,
2001, from http://www.riaa.com/pdf/napsterdecision.pdf

Adams, M. S. (1996). Labeling and differential association: Toward a general social
learning theory of crime and deviance. American Journal of Criminal Justice,
20(2), 147-164.

Agnew, R. (1985). A revised strain theory of delinquency. Social Forces, 64, 151-167.

Agnew, R (1989). A longitudinal test of the revised strain theory. Journal of
Quantitative Criminology, 5, 373-387.

Agnew, R. (1992). F cundaticn for a General Strain Theory of Crime and Delinquency.
Criminology, 30(1), 47-87.

Agnew, R. (1997). Stability and change over the life course: A strain theory explanation.
In T. P. Thomberry (Ed.), Developmental Theories of Crime and Delinquency
(pp. 101-132). New Brunswick, NJ: Transaction.

Agnew, R. (1998). The approval of suicide: A social-psychological model. Suicide and
Life-Threatening Behavior, 28(2), 205-225.

Agnew, R. (1999). A general strain theory of community differences in crime rates.
Journal of Research in Crime and Delinquency, 36(2), 123-155.

Agnew, R., & Brezina, T. (1997). Relational problems with peers, gender, and
delinquency. Youth & Society, 29(1), 84-111.

Agnew, R., Brezina, T., Wright, J. P., & Cullen, F. T. (2002). Strain, personality traits,
and delinquency: Extending general strain theory. Criminology, 40(1), 43-71.

Agnew, R., Cullen, F. T., & Burton, V. S. J., et a1. (1996). A new test of classic strain
theory. Justice Quarterly, 13(4), 681-704.

Agnew, R., & Peters, A. A. R. (1986). The techniques of neutralization: An analysis of
predisposing and situational factors. Criminal Justice and Behavior, 13(1), 81-97.

Agnew, R., & White, H. R. (1992). An Empirical Test of General Strain Theory.
Criminology, 30(4), 475.

Agrawal, M., Gopal, R. D., Sanders, G. L., 85 Wagner, S. C. (2003). A Behavioral Model
of Digital Music Piracy (FORTHCOMING). Journal of Organizational
Computing and Electronic Commerce.

277

Akers, R. L. (1977). Deviant behavior: A social learning approach. Belmont, CA:
Wadsworth.

Akers, R. L. (1985). Deviant behavior: A social learning approach (3 ed.). Belmont, CA:
Wadsworth.

Akers, R. L. (1991). Self-control as a general theory of crime. Journal—of-Quantitative—
Criminology, 7((2)), 201-211.

Akers, R. L. (1992). Linking sociology and its specialties. Social Forces, 71, 1-16.
Akers, R. L. (1994). Criminological Theories. Los Angeles: Roxbury.

Akers, R. L. (1996). Is differential association/social learning cultural deviance theory?
Criminology, 34(2), 229-256.

Akers, R. L. (1998). Social learning and social structure: A general theory of crime and
deviance. Boston: Northeastern University Press.

Akers, R. L., & Cochran, J. K. (1985). Adolescent marijuana use: A test of three theories
of deviant behavior. Deviant Behavior, 6(4), 323-346.

Akers, R. L., Krohn, M. D., Lanza-Kaduce, L., & Radosevich, M. (1979). Social learning
and deviant behavior: A speciﬁc test of a general theory. American Sociological
Review, 44, 636-655.

Akers, R. L., & Lee, G. (1996). A longitudinal test of social learning theory: Adolescent
smoking. Journal of Drug Issues, 26(2), 317-343.

Angus Reid Worldwide. (20003). Despite the use of MP3 technology being widespread
among Internet-enabled Canadians, cd-buying behaviour remains strong.
Retrieved September 24, 2001, from
http://www.angysreid.com/media/content/displavpncﬁn?id to_view=985

Angus Reid Worldwide. (2000b). MP3 technology bridging Internet generation gap.
Retrieved October 1, 2001, from
http ://www.angusreid.com/media/content/displaypr.cﬁn?id_to_view=l 042

Archambault, D. (1999). MP3 format and the purchasing of music. Retrieved 2001,
September 30, from http://www.mp3.com/news/236.html

Ameklev, B. J ., Grasmick, H. G., Tittle, C. R., & Bursik, R. J. (1993). Low Self-control
and Imprudent Behavior. Journal of Quantitative Criminology, 9, 225-247.

Aseltine, R. H., Gore, 8., & Gordon, J. (2000). Life stress, anger and anxiety, and
delinquency: An empirical test of general strain theory. Journal of Health and
Social Behavior, 41(3), 256-275.

278

Babcock, L., & Loewenstein, G. (1997). Explaining bargaining impasse: The role of self-
serving biases. Journal of Economic Perspectives, 11, 109-126.

Bachman, R., & Paternoster, R. (1997). Statistical methods for criminology and criminal
justice. New York: McGraw-Hill.

Bandura, A. (1969). Principles of behavior modification. New York: Holt, Rinehart,
Winston.

Bandura, A. (1973). Aggression: A social learning analysis. Englewood Cliffs, NJ:
Prentice-Hall.

Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice-Hall.

Bandura, A., & Walters, R. (1963). Social learning and personality development. New
York: Holt, Rinehart, Winston.

Banerjee, D., Cronan, T. P., & Jones, T. W. (1998). Modeling IT ethics: A study in
situational ethics. Mis Quarterly, 22(1), 31-60.

Beccaria, C. (1968). On Crimes and Punishments (H. Paolucci, Trans). Indianapolis:
Bobbs-Merrill.

Benson, M. L. (1985). Denying the guilty mind: Accounting for involvement in a white-
collar crime. Criminology, 23(4), 583-608.

Benson, M. L., & Moore, E. (1992a). Are White-Collar and Common Offenders the
Same - an Empirical and Theoretical Critique of a Recently Proposed General-
Theory of Crime. Journal of Research in Crime and Delinquency, 29(3), 251-272.

Benson, M. L., & Moore, E. (1992b). Are White-collar and Common Offenders the
Same? An Empirical and Theoretical Critique of a Recently Proposed General
Theory of Crime. Journal of Research in Crime and Delinquency, 29, 251-272.

Bentham, J. (Ed). ([1781] 1970). An Introduction to the Principles of Morals and
Legislation. London: The Athlone Press.

Berst, J. (1997). Has your web site been stolen? Retrieved April 28, 1997, from
1m://www.zdnet.com/anchordesk/storv/storv 864.html

Bettig, R. V. (1996). Copyrighting culture. Boulder, CO: Westview.

Bhattacharjee, S., Gopal, R. D., & Sanders, G. L. (2003). Digital music and online
sharing: software piracy 2.0? Communications of the ACM (CACM), 46(7), 107-
1 l 1.

279

Black, J. (2003). Big Music's Broken Record. Retrieved February 13, 2003, from
httgﬂwwwbusinessweek.com/technology/content/feb2003/tc200302 l 3 9095_tcO
78.htm

Borland, J. (2000). US. sides with RIAA against Napster. Retrieved November 15, 2001,
from http://news.cnet.com/news/O-1005-200-273 1 198.html

Borland, J. (2002). Hollywood chases down campus pirates. Retrieved October 10, 2002,
from http://news.com.com/2100-1023-961637.html

Borrowman, 8., Young, A., Starnbaugh, H., Copping, T., Williams, W. P., Beaupre, D.
S., et al. (2001). Research report: Electronic crime needs assessment for state and
local law enforcement: National Institute of Justice. U. S. Department of Justice.
Ofﬁce of Justice Programs. National Institute of Justice.

Bowman, L. M. (2000). Musicians praise, bash Napster. Retrieved October 21, 2000,
from http://www.zdnet.com/zdnn/stories/news/0,45 86.264336 1 ,00.html

Bowman, L. M. (2003). RIAA to sue thousands of ﬁle swappers. Retrieved June 25, 2003,
from http://zdnet.com.com/21 00-1 105_2-1020876.html

Brand, W. (2003). Students warned on downloads. Retrieved September 1, 2003, from
http://www.oaklandtﬁbune.com/Stories/0,l413§2~1726~1575835,00.html

Breen, C. (2000). Steal this song. Retrieved November 8, 2001, from
http://www.cnn.com/2GOO/TECH/computing/07/07/digital.musicspecial.idg/inde
x.html

Brezina, T. (1998). Adolescent maltreatment and delinquency: The question of
intervening processes. Journal of Research in Crime and Delinquency, 35(1), 71-
99.

Brezina, T., Piquero, A. R., & Mazerolle, P. (2001). Student anger and aggressive
behavior in school: An initial test of Agnew's macro-level strain theory. Journal
of Research in Crime and Delinquency, 38(4), 362-386.

Broidy, L. M. (2001). A test of general strain theory. Criminology, 39(1), 9.

Broidy, L. M., & Agnew, R. (1997). Gender and crime: A general strain theory
perspective. Journal of Research in Crime and Delinquency, 34(3), 275-306.

Brownﬁeld, D., & Sorenson, A. (1993). Self-control and Juvenile Delinquency:
Theoretical issues and an Empirical Assessment of Selected Elements of a
General Theory of Crime. Deviant Behavior, 14, 243-264.

Brownﬁeld, D., & Thompson, K. (1991). Attachment to peers and delinquent behavior.
Canadian Journal of Criminology, 33, 45-60.

280

Buckley, M. R., Wiese, D. S., & Harvey, M. G. (1998). An investigation into the
dimensions of unethical behavior. Journal of Business Education, 73(5), 284-290.

Burgess, R., & Akers, R. (1966). A differential association-reinforcement theory of
criminal behavior. Social Problems, 14, 128-147.

Burton, V. S., Cullen, F. T., Evans, T. D., Alarid, L. F., & Dunaway, R. G. (1998).
Gender, self-control, and crime. Journal of Research in Crime and Delinquency,
35(2), 123-147.

Callon, M. (1987). Society in the Making: the Study of Technology as a Tool for
Sociological Analysis. In W. E. Bijker, T. P. Hughes & T. J. Pinch (Eds.), The
social construction of technical systems: New directions in the sociology and
history of technology (pp. 83-103). Cambridge, MA: The MIT Press.

Callon, M., & Law, J. (1989). On the Construction of Sociotechnical Networks: Content
and Context Revisited. Knowledge and Society: Studies in the Sociology of
Culture Past and Present, 8, 57-83.

Callon, M., & Law, J. (1997). After the individual in society: Lessons on collectivity
from science, technology, and society. Canadian Journal of Sociology, 22, 165-
1 82.

Calpo, G. (2000). MP3 bitrate comparison. Retrieved June 20, 2000, from
http://www.pinovware.com/ﬂiptech/bitrate.shtml

Campbell, D. T. (1988). Deﬁnitional versus multiple operationism. In S. Overrnan (Ed.),
Methodology and Epistemology for Social Science (pp. 31-36). Chicago:
University of Chicago Press.

Carroll, A., Juarez, M., Polk, J ., & Leininger, T. (2002). Microsoft "Palladium": A
Business Overview. Retrieved October 1, 2003, ﬁ'om
http://www.microsoﬁ.com/presspass/features/2002/iu102/0724Palladiumwp.asp

Catalano, R. F., Kosterrnan, R., & Hawkins, J. D., et al. (1996). Modeling the etiology of
adolescent substance abuse: A test for the social development model. Journal of
Drug Issues, 26(2), 429—455.

Chantler, N. (1996). Proﬁle of a Computer Hacker. Florida: Infowar.

Christensen, A. L., & Eining, M. M. (1991). Factors inﬂuencing software piracy:
Implications for accountants. Journal of Information Systems, 67-80.

Cloward, R. A., 8c Ohlin, L. E. (1963). Delinquency and opportunity: a theory of
delinquent gangs (Fourth ed.). New York: The Free Press.

281

CNN.com. (2000a). MP3.com beats street: Online music service reports loss of 9 cents a
share, revenue rises sharply. Retrieved August 3, 2001, ﬁom
http://cnnﬁ1.cnn.com/2000/ l 0/19/technology/eamsimp3/index.htm

CNN.com. (2000b). MP3.com ordered to pay up to $250 million in music copyright case.
Retrieved October 3, 2001, from
http://www.cnn.com/2000/LAW/09/06/mp3.lawsuit.01/index.html

CNN.com. (2000c). Napster heads to court. Retrieved October 5, 2001, from
http://www.cnnfn.cnn.com/2000/07/03/companies/napster/

CNN.com. (2004). More song swappers sued. Retrieved J anuary 24, 2004, from
http://money.cnn.con1/2004/01/2 l/technology/riaaisuits/index.htm

Cochran, J. K., Wood, P. B., Sellers, C. S., & Chamlin, M. B. (1998). Academic
Dishonesty and Low Self-control: An Empirical Test of a General Theory of
Crime. Deviant Behavior, 19, 227-255.

Cochran, J. K., Wood, P. 3., Sellers, C. S., Wilkerson, W., & Chamlin, M. B. (1996).
Academic dishonesty and low-self-control: An empirical test of a general theory
of crime. Paper presented at the American Society of Criminology, Chicago, IL.

Cohen, A. K. (1955). Delinquent boys: The culture of the gang. Glencoe, IL: The Free
Press.

Cohen, L. E., & Felson, M. (1979). Social change and crime rate trends: A routine
activities approach. American Sociological Review, 44, 588-607.

Coleman, J. W. (1989). The criminal elite: the sociology of White-collar crime. New
York: St. Martin's Press.

Congress, t. (1997). Bill summary and status: No Electronic Theft (NET) Act. Retrieved
2001, October 15, from http://www.thomas.loc.gov/cgi-
bin/bdquerv/z?dl 05:HR02265:@@@L

Consumer Electronics News. (2003). Compressed Audio Player Market Will Continue to
Surge Ahead as Consumers Snap Up Diverse Array of Devices, Says IDC.
Retrieved June 12, 2003, from http://www.prnewswire.com/c;&
birr/stories.pl?ACCT=SVBIZINK8.storv&STORY=/www/storv/06-l 2-
2003/0001964220&EDATE=THU+Jun+12+2003,+09:09+AM

Copyright Ofﬁce of the United States. (1998). Summary of Digital Millennium Copyright
Act of 1998. Retrieved September 24, 2001, from
http://www.loc.gov/copvright/lggislation/dmcapdf

Copyright Ofﬁce of the United States. (2000a). Copyright Basics, Circular 1. Retrieved
October 4, 2001, from http://www.chgov/copvﬁght/circs/circl .html

282

Copyright Ofﬁce of the United States. (2000b). Copyright law of the United States of
America and related laws contained in Table 1 7 of the United States Code.
Retrieved October 1, 2000, from http://www.loc.gov/c04_)yright/titlel 7/circ92.pdf

Cravotta, N. (2000). The Internet-audio (r)evolution. Retrieved February 3, 2000, from
http://www.ﬁndarticles.com/cf_0/mOEDN/3 45/59969530/printjhtm1

Crawford, W. (2000). New sounds for a new millennium? Online, 24(2), 82-84.
Cressey, D. R. (1953). Other people's money. New York: The Free Press.

Cross, P. (1977). Not can but will college teaching be improved. New Directions in
Higher Education, 1 7(1).

CyberAtlas. (2001). Piracy extends beyond software to digital content, 2001, November
9, from
http://cyberatlas.intemet.com/big picture/applications/arrticle/O,l323,130] 92068
1 00.html

 

Davis, M. (2003). Penn State seeks out student piracy. Retrieved May 1, 2003, from
http://www.redandblack.com/vnews/displav.v/ART/2003/05/01/3eb l 2b203a281

de F ontenay, E. (1999). The music industry enters the digital age. Retrieved September
24, 2001, from http://www.musicdish.com/survev/downloads/MP3report.pdf

Dean, K. (2003a). RIAA hits students where it hurts. Retrieved April 5, 2003, from
http://www.wired.com/news/di giwood/O, l 41 2,5 83 5 l .00.html

Dean, K (2003b). Upload a ﬁle, go to prison. Retrieved July 17, 2003, from
ht_tp://www.wired.com/news/digiwood/0,1412,59654,00.html

Denning, D. (1998). Information Warfare and Security. Reading: Addison-Wesley.

Difﬁe, W. (1988). The ﬁrst ten years of public-key cryptography. Proceedings of the
IEEE, 76, 560-577.

Difﬁe, W., & Hellman, M. E. (1976). New directions in cryptography. IEEE
Transactions on Information Theory, 22, 644-654.

DMA. (2000). Web Music. Retrieved September 28, 2001, 2001, from
http://www.diggrediaorg/whatsnew/web music.html

Dordick, H. S. (1986). Intellectual property: Protecting rights and privileges in an
electronic age. New York: Clark Boardman.

Dymess, C. (2002). Setting sites on nabbing pirates: Stolen web design a growing
problem, 2002, September 18, from .
http://newsobserver.com/business/story/1 741 1799-1 753793c.html

283

Eining, M. M., & Christensen, A. L. (1991). A psycho-social model of software piracy:
The development of a test model. In R. Dejoie, G. Fowler & D. Paradice (Eds.),
Ethical Issues in Information Systems (pp. 134-140). Boston: Boyd & Fraser.

Etzioni, A. E., O. (1999). Face-to-face and computer-mediated communities, a
comparative analysis. The Information Society, 15, 241-248.

Evangelista, B. (2003). Film industry crusades for piracy laws, 2003, May 12, from
http://wW.rochountainnews.com/drmn/technology/article/O,l 299,DRMN_49
1954561 ,00.html

Evans, T. D., Cullen, F. T., Burton, V. S., Dunaway, R. G., & Benson, M. L. (1997). The
social consequences of self-control: Testing the general theory of crime.
Criminology, 35(3), 475-504.

F laherty, L., Pearce, K., & Rubin, R. (1998). Internet and face-to-face communication:
Not functional alternatives. Communication Quarterly, 46(3), 250-268.

Frank, S. (2003). Downloads could result in discipline. Retrieved August 15, 2003, from
httpzl/wwwstatenewscom/pri nt.phtml?pk=l 8447

Garner, B. A. (Ed.). (1999). Black's Law Dictionary (7th ed.): West Group Publishing.

Gentile, G. (2003). Film industry trying to stop pirates. Retrieved April 18, 2003, ﬁom
http://www.dailvtexanonline.com/vnews/display.v/ART/2003/04/ 1 8/ 3 e9fb66f698
92

Gibbs, J. J., & Giever, D. (1995). Self-Control And Its Manifestations Among University
Students: An Empirical Test Of Gottfredson And Hirschi's General Theory.
Justice Quarterly: JQ, 12(2), 231.

Gibbs, J. J ., Giever, D., 81. Martin, J. S. (1998). Parental management and self-control:
An empirical test of Gottfredson and Hirschi's general theory. The Journal of
Research in Crime and Delinquency, 35(1), 40. .

Gilbert, S. W., 8: Lyman, P. (1989). Intellectual property in the information age: Issues
beyond the copyright law. Change, 21, 23-28.

Google. (2002). Year-End Google Zeitgeist: Search patterns, trends, and surprises.
Retrieved May 20, 2003, from http://www.google.com/press/zeitgeist2002.html

Gopal, R. D., & Sanders, G. L. (1997). Preventive and deterrent controls for software
piracy. Journal of Management Information Systems, 13(4), 29.

Gopal, R. D., & Sanders, G. L. (1998). International software piracy: Analysis of key
issues and impacts. Information Systems Research, 9(4), 380-397.

284

Gopal, R. D., Sanders, G. L., Bhattacharjee, S., Agrawal, M., & Wagner, S.
(forthcoming). A behavioral model of digital music piracy. Journal of
Organizational Computing and Electronic Commerce.

Gottfredson, M., & Hirschi, T. (1990). A general theory of crime. Stanford, CA: Stanford
University Press.

Grasmick, H. G., & Bursik, R. J. (1990). Conscience, signiﬁcant others, and rational
choice: Extending the deterrence model. Law and Society Review, 24, 837—861.

Grasmick, H. G., & Green, D. E. (1980). Legal punishment, social disapproval and
intemalisation as inhibitors of illegal behavior. The Journal of Criminal Law and
Criminology, 71(3), 325-335.

Grasmick, H. G., Tittle, C. R., Bursik, R J ., & Ameklev, B. J. (1993). Testing the core
empirical implications of Gottfredson and Hirschi's general theory of crime. The
Journal of Research in Crime and Delinquency, 30(1), 5.

Hagan, J ., McCarthy, 8., & Foster, H. (2002). A gendered theory of delinquency and
despair in the life course. Acta Sociologica, 45(1), 37-46.

Hall, J. (1935). Theft, Law and Society. Boston: Little Brown & Co.

Haney, C. (2000). Microsoft fears e-book piracy. Retrieved August 29, 2001, ﬁ'om
http://www.pcworld.com/news/article/0,aid, l 8262,00.asp

Harari, O. (1999). You say you want a revolution? Management Review, 88(10), 30-33.

Harmon, A. (2001). Software double bind, 2001, August 13, ﬁorn
http://www.nvtimes.com/2001/08/13/technology/ebusiness/13NECO.html

Harris, L. J. (1969). Nurturing new ideas. Washington, DC: Bureau of National Affairs.

Harris, M. (2003). Now playing: movie piracy crackdown. Film industry ﬁghts back
against illegal recording, 2003, May 2, ﬁ'om
http://www.ca_r_r_ada.com/edmonton/edmontonjoumal/storv.asp?id=F90C17FC-
7E1A-4954-9CAB-81 14B633B58A

 

Hawkins, J. D., Herrenkohl, T., Farrington, D. P., Brewer, D., Catalano, R. F., & Harachi,
T. W. (1998). A review of predictors of youth violence. In R. Loeber & D. P.
F anington (Eds.), In serious & violent juvenile oﬁ'enders: Risk factors and
successful interventions. Thousand Oaks, CA: SAGE Publications.

Healey, J ., & Leeds, J. (2003). Tone deaf to a moral dilemma? Retrieved September 2,
2003, from http://www.latimes.com/technology/la-ﬁ-
morals2sep02234423,0,777l l28.story?coll=la-headlines-technology

285

 

Healy, J. (2003). 4 pay steep price for ﬂee music. Retrieved May 2, 2003, ﬁom
http://www.latimes.com/la-ﬁ-settleZmayOZ,0,6007000.story

Heid, J. (1997). So long, CDs. Retrieved October 3, 2001, from
http ://www.macworld.zdnet.com/ l 999/07/features/M P3 .html

Heimer, K. (1997). Socioeconomic status, subcultural deﬁnitions, and violent
delinquency. Social Forces, 75, 799-834.

Himmelweit, H., Biberian, M., & Stockdale, J. (1978). Memory for Past Vote:
Implications of a Study of Bias in Recall. British Journal of Political Science, 8,
365-375.

Hinduja, S. (2001). Correlates of Internet software piracy. Journal of Contemporary
Criminal Justice, 17((4)), 369-382.

Hinduja, S. (2003a). Neutralizing Piracy.

Hinduja, S. (2003b). Trends and patterns among software pirates. Ethics and Information
Technology, 5(1), 49-61.

Hirschi, T. (1969). Causes of delinquency. Berkeley: University of California Press.

Hirschi, T., & Gottfredson, M. (1993). Commentary: testing the general theory of crime.
Journal-of-Research-in-Crime-and-Delinquency, 30((l)), 47-54.

Hoffman, J. P., & Miller, A. S. (1998). A latent variable analysis of general strain theory.
Journal of Quantitative Criminology, 14(1), 83-110.

Hofﬁnann, J. P., & Su, S. S. (1997). The conditional effects of stress on delinquency and
drug use: A strain theory assessment of sex differences. Journal of Research in
Crime and Delinquency, 34(1), 46-78.

Horvath, F. (1982). Forgotten Unemployment: Recall Bias in Retrospective Data.
Monthly Labor Review, March, 40-43.

Hosmer, D. W., & Lemeshow, S. (2000). Applied logistic regression. New York: John
Wiley and Sons.

HR2517. (2003). Piracy Deterrence and Education Act of 2003 . Retrieved September 20,
2003, from http :0www.thcmtor.comfbills 1 08/hr25 17.html

Hughes, T. P. (1983). Networks of power: Electrification in western society, 1880-1930.
Baltimore: Johns Hopkins University Press.

Hughes, T. P. (1987). The evolution of large technical systems. In W. E. Bijker, T. P.
Hughes 8: T. J. Pinch (Eds.), The social construction of technological systems.
Cambridge, MA: The MIT Press.

286

IFPI. (2002). IFPI Music Piracy Report. Retrieved September 15, 2003, from
hﬁttpzﬂwwwifpiorg/site-content/library/piracv2002.pdf

Irn, J. H., & Van Epps, P. D. (1991). Software piracy and software security in business
schools: An ethical perspective. Data Base, 22(3), 15-21.

International Standards Organization. (2003). Introduction. Retrieved October 20, 2003,
ﬁ'om http://www.iso.ch/iso/en/aboutiso/introduction/index.html

Jacobs, F., & Allbritton, C. (2001). E-book publishers face piracy panic. Retrieved
January 1, 2001, from
http://www.cnn.com/2001 /TECH/computing/Ol/01 /e.book.piracy.panic.idg/

Jay, E. D. (2000). Survey of napster users. Retrieved October 4, 2001, from
http://www.riaa.com/PDF/jav.pdf

Johnston, M. (2000). E-book industry warned of piracy threat. Retrieved September 25,
2000, from
http://www.idg.net/idgns/2000/09/2S/EBookIndusteramedOfPiracyThreat.shtm
1

Jones, C. (2000). RIAA sues MP3.com. Retrieved October 2, 2001, from
http://www.wired.com/news/politics/O,l 283,33634,00.html

J unger, M., & Tremblay, R. E. (1999). Self-control, accidents, and crime. Criminal
Justice and Behavior, 26(4), 485-501 .

Kaplan, H. B., Johnson, R. J ., & Bailey, C. A. (1987). Deviant peers and deviant
behavior: Further elaboration of a model. Social Psychology Quarterly, 50(3),
277-284.

Karagiannis, K. (1999, 1999, July). Digital audio and MP3. Electronics Now, 70, 22-24:

Katz, D., & Kahn, R. L. (1966). Organizations and the systems concept. In D. Katz & R.
Kahn (Eds.), The Social Psychology of Organization (pp. 14-29). New York:
Wiley.

Keane, C. M., Paul S; Teevan, James J. (1993). Drinking and driving, self-control, and
gender: Testing a general theory of crime. The Journal of Research in Crime and
Delinquency, 30(1), 30.

Kibbee, R. (1999). The choice conundrum. MP3: Digital Music for the Millennium?
Retrieved September 25, 2001, from
http://www.musicdish.com/survey/downloads/MP3reportpdf

Kievit, K. (1991). Information systems majors/non-majors and computer ethics. Journal
of Computer Information, 32(1), 43-49.

287

King, B. (2000a). Napster: Music 's friend or foe? Retrieved September 29, 2001, from
http://www.wired.com/news/print/O,1 294,3 696 1 ,00.html

King, B. (2000b). New school of thought on piracy. Retrieved September 29, 2001, from
http://www.wired.com/news/print/O,1294,36875,00.html

Knight Systems. (2000). The top ten. Retrieved July 7, 2000, from
http://www.searchtemrs.com

Krochmal, M. (1998). Music industry unprepared for MP3. Retrieved October 1, 2001,
from http:wawtechwebcom/story/TWB1998071680010

Krohn, M. L., Skiller, W. F., Massey, J. L., & Akers, R. (1985). Social learning theory
and adolescent cigarette smoking: A longitudinal study. Social Problems, 32(5),
455-473.

Lanier, M., & Henry, S. (1998). Essential Criminology. New York: Free Press.

Latonero, M. (2000). Survey of MP3 usage: Report on a university consumption
community. Retrieved October 5, 2001, from
http://www.entertz_rinment.usc.edu/publications/mp3.pdf

Latour, B. (1987). Science in Action: How to Follow Scientists and Engineers through
Society. Cambridge, MA: Harvard University Press.

Latour, B. (1997). On Recalling ANT - 'Actor Network and After' Workshop, Centre for
Social Theory and Technology (CSTT), Department of Sociology and Social
Anthropology, Keele University.

Law, J. (1987a). The Structure of Sociotechnical Engineering-A Review of the New
Sociology of Technology. The Sociological Review, 35(2), 404-425.

Law, J. (1987b). Technology and Heterogeneous Engineering: The Case of Portuguese
Expansion. In W. Bijker, T. P. Hughes 8; T. Pinch (Eds.), The Social
Construction of Technological Systems (pp. 111-134). Cambridge: MIT Press.

Law, J. (1992). Notes on the theory of the actor-network: Ordering, strategy and
heterogeneity. Systems Practice, 5, 379-393.

Law, J. (1994). Organizing modernity. Oxford: Blackwell.

Law, J. (1997). T raduction/T rahison: Notes on ANT. Retrieved July 10, 2003, ﬁ'om
http://www.cornp.lancs.ac.uk/sociologv/stslathtml

Law, J. (1999). After ANT: complexity, naming and topology. In J. Law & J. Hassard
(Eds.), Actor Network Theory and After (pp. 1-14). Oxford: Blackwell Publishers.

288

 

Law, J ., & Callon, M. (1988). Engineering and Sociology in a Military Aircraft Project:
A Network Analysis of Technological Change. Social Problems, 35(3), 284-297.

Leannonth, M. (2000). Finally, some numbers on Napster. Retrieved October 2, 2001,
from http://www.thestandard.com/article/displav/O,l 151,15502,00.html

Leibowitz, W. R. (1999). How law enforcement cracks cybercrimes. New York Law
Journal, 5.

Lersch, K. M. (1999). Social learning theory and academic honesty. International
Journal of Comparative and Applied Criminal Justice, 23(1), 103-114.

Lessig, L. (1997). Intellectual property and code. St. Johns Journal of Legal
Commentary, 11 (635).

Lessig, L. (1999a). Code and Other Laws of Cyberspace. New York: Basic Books.

Lessig, L. (1999b). Commons and Code. F ordham Intellectual Property, Media and
Entertainment Law Journal, 9(459).

Lipton, B. (1998). New music pirates face lawsuits. Retrieved 2001, September 18, from
http://www.news.com/News/Item/O,4,22060,00.html

Long, J. S. (1997). Regression Models for Categorical and Limited Dependent Variables.
Thousand Oaks, CA: Sage.

Long, P. D. (2003). iTunes and the technology beyond MP3. Retrieved July 1, 2003, from
http://www.syllers.coWanicle.asp?id=7895

Longshore, D., & Turner, S. (1998). Self-control and criminal opportunity: Cross
sectional test of the general theory of crime. Criminal Justice and Behavior,
25(1), 81.

Longshore, D., Turner, S., & Stein, J. A (1996). Self-control in a Criminal Sample: An
Examination of Construct Validity. Criminology, 34, 209-228.

Luckenbill, D. F ., & Miller, S. L. (1998). Defending intellectual property: State efforts to
protect creative works. Justice Quarterly, 15(1), 93-120.

Maddala, G. (1983). Limited dependent and qualitative variables in econometrics. New
York: Cambridge University Press.

Manning, W. W. (1997). Should you be on the net? The FBI Law Enforcement Bulletin,
66(1), 18.

Mariano, G. (2002). Apple: Play music at your own risk. Retrieved August 25, 2003,
from http://news.com.com/2100-1023-912695.htnrl

289

Marsh, H. W., & Hau, K. T. (1999). Conﬁrrnatory factor analysis: Strategies for small
sample sizes. In R. H. Hoyle (Ed.), Statistical strategies for small sample
research. Thousand Oaks, CA: Sage.

Matsueda, R. L. (1982). Testing control theory and differential association: A causal
modeling approach. American Sociological Review, 47, 489-504.

Matsueda, R. L., & Anderson, K. (1998). The dynamics of delinquent peers and
delinquent behavior. Criminolog, 36(269-308).

Mazerolle, P. (1998). Gender, general strain, and delinquency: an empirical examination.
Justice Quarterly, 15(1), 65-91.

Mazerolle, P., Brame, R., Paternoster, R., Piquero, A., & Dean, C. (2000). Onset age,
persistence, and offending versatility: comparisons across gender. Criminology,
38(4), 1143-1172.

Mazerolle, P., & Maahs, J. (2000). General strain and delinquency: An alternative
examination of conditioning inﬂuences. Justice Quarterly, 1 7(4), 753-778.

Mazerolle, P., & Piquero, A. (1998). Linking exposure to strain with anger: An
investigation of deviant adaptations. Journal of Criminal Justice, 26(3), 195-211.

McCullagh, D. (2002). DOJ to swappers: Law's not on your side. Retrieved September
12, 2003, from httpz//news.com.com/2100-l 023-954591.html?tLEfd_top

McKelvey, R., & Zavoina, W. (1975). A statistical model for the analysis of ordinal level
dependent variables. Journal of Mathematical Sociology, 4, 103-120.

Mendels, P. (1999). University craclm down on MP3 trade. Retrieved September 20,
2001, ﬁnal http://www.ou.edu/archives/it-fvi/0723.html

Menta, R. (2000). The merits in the mp3. com lawsuit. Retrieved October 5, 2001, from
http://www.mp3newswire.net/stories/2000/MP3courtpoint.html ,

Merton, R. K. (1938). Social structure and anomie. American Sociological Review, 3,
672-682.

Merton, R. K. (1968). Social Theory and Social Structure. New York: Free Press.

Messner, S. R, 8c Rosenfeld, R. (1994). Crime and the American Dream. Belmont, CA:
Wadsworth.

Michigan State University. (1998). Acceptable Use Policy. Retrieved August 10, 2001,
from http://www.msu.edu/dig/aup/msuaup.html

Michigan State University. (2003). Acceptable use of your account. Retrieved November
15, 2003, from https://mail.msu.edu/upgrade/pilotupgrade2.htrnl

290

Microsoft. (2000). Microsoft incorporates new anti-piracy technologies in Windows
2000, Office 2000. Retrieved April 5, 2000, from
http://www.microsoft.com/presspass/press/Z000/feb00/ap featurcsprasp

Middle Tennessee State University. (2001). Computer and Network Acceptable Use
Policy. Retrieved March 15, 2001, ﬁom http://www.mtsu.edu/misc/policy.html

Midgley, C. (2000). History of MP3? MP3 for the Mac - best of MP3 for Apple Macs.
Retrieved 2001, September 29, from
http://www.mp3mac.co.uk/Pagesfl-listorv of MP3.html

Miller, J. K., & Gergen, K. J. (1998). Life on the line: The therapeutic potentials of
computer-mediated conversation. Journal of Marital and Family Therapy, 24(2),
1 89-202.

Mindenhall, C. (2000). Music pho the masses, 2000, January 21-27, from
http://www.laweeklv.com/ink/00/09/cyber—mindenhall.php

Morgenstem, R., & Barrett, N. (1974). The Retrospective Bias in Unemployment
Reporting By Sex, Race and Age. Journal of the American Statistical Association,
69, 355-357.

MP3 Newswire. (2000). RIAA defeats MP3.com in court. Retrieved September 28, 2001,
from http://www.mp3newswire.net/stories/2000/MP3courtrulinghtml

MP3.com. (1998). RIAA wins restraining order against MP3 recording device. Retrieved
October 9, 2001, ﬁ'om http://www.mp3.com/news/112.html

MP3.com. (1999). Glut yourself on more SDMI dirt - Editorial. Retrieved October 2,
2001, from http://www.mp3.com/news/31 1.html

Musil, S. (2003). Week in review: RLAA rings twice. Retrieved May 16, 2003, ﬁ'om
http://news.com.com/2100-1083 3-1003664.htm1

Nagin, D. S., & Paternoster, R (1993). Enduring individual differences and rational
choice theories of crime. Law & Society Review, 27(467-96).

Nielsen Net Ratings. (2003). Nearly 40 Million Internet Users Connect Via Broadband,
Growing 49 Percent. Retrieved October 20, 2003, from http://www.nielsen-
netratingscomlnewsjsp

Nijmeh, G. (2001). Behind the files: History of MP3. Retrieved July 20, 2002, from
http://www.mp3-cdbumer.com/History-Of-MP3s.shtml

Nilsson, M. (2000). ID3v2 made easy. Retrieved October 20, 2001, from
http://www.id3.org/easv.html

291

Oakes, C. (2000). Copy-protected CDs taken back. Retrieved August 20, 2003, ﬁom
http:l/www.wired.com/news/technology/O,1282,3 3921 ,00.htnrl

Ofﬁce of Technology Assessment. (1986). Intellectual property rights in an age of
electronics and information. Washington, DC: US. Government Printing Ofﬁce.

Orcutt, J. D. (1987). Differential association and marijuana use: A closer look at
Sutherland (with a little help from Becker). Criminology, 25(2), 341-358.

Paternoster, R. (1989). Decisions to participate in and desist from four types of common
delinquency: Deterrence and the rational choice perspective. Law and Society
Review, 23, 7-40.

Paternoster, R., & Brame, R. (1998). The structural similarity of processes generating
criminal and analogous behaviors. Criminology, 36, 633-669.

Paternoster, R., & Brame, R. (2000). On the association among self-control, crime, and
analogous behaviors. Criminology, 38(3), 971-982.

Paternoster, R., & Mazerolle, P. (1994). General Strain Theory and Delinquency - a
Replication and Extension. Journal of Research in Crime and Delinquency, 31(3),
23 5-263.

Paternoster, R., & Simpson, 8. (1996). Sanction Threats and Appeals to Morality: Testing
a Rational Choice Model of Corporate Crime. Law & Society Review, 30, 549-
583.

Patrizio, A. (1999). DOJ cracks down on MP3 pirate. Retrieved September 27, 2001,
from hm)://www.wired.com/news/politics/O,1283,21391,00.html

Patterson, J. (1998). The history of computers during my lifetime - the 1980's, ﬁom
http://www.pattosoft.com.Mason/Articles/HistorvOfComputers/ 19803.html

Petersen, S. (2000). Music's new killer app: Napster. Retrieved March 21 , 2001, from
http ://comment.zdnet.co.ul_c/storv/0,,t479-s2l 13471 ,00.html

Petreley, N. (2000). How record companies could embrace Napster and maintain proﬁts.
Retrieved September 27, 2001, from http://www.idg.net/go.cgi?id=250024

Pew Internet & American Life Project. (2000). 13 million Americans 'ﬁeeload' music on
the Internet; 1 billion free music ﬁles now sit on Napster users' computers.
Retrieved 2000, October 5, from
lip://www.pewinternet.org/reports/pdfs/MusicReportFullpdf

Philipkoski, K. (1999a). The student jukebox sting. Retrieved October 1, 2001, from
http://www.wired.com/news/culture/OAZ84,32444,00.html

292

Philipkoski, K. (1999b). University snoops for MP3s. Retrieved October 2, 2001, from
http://www.wired.com/news/MP3/0,1285,32478,00.html

Philips, C. (2000). Recording chief critiques criticisms of her industry. Retrieved October
1, 2001, ﬁ'om http://www.lawtimes.com/business/20000427/t000039445.html

Piquero, A. R., MacIntosh, R., & Hickman, M. (2000). Does self-control affect survey
response? Applying exploratory, conﬁrmatory, and item response theory analysis
to Grasmick et al.'s self-control scale. Criminology, 38(3), 897.

Piquero, A. R., & Rosay, A. B. (1998). The Reliability and Validity of Grasmick, et al.'s
Self-control Scale: A Comment on Longshore, et a1. Criminology, 36, 157-173.

Piquero, A. R., & Tibbetts, S. (1996). Specifying the Direct and Indirect Effects of Low
Self-control and Situational Factors in Offenders' Decision Making: Toward a
More Complete Model of Rational Offending. Justice Quarterly, 3, 481-510.

Piquero, N. L., & Sealock, M. D. (2000). Generalizing general strain theory: An
examination of an offending population. Justice Quarterly, 1 7(3), 449-484.

Ploskina, B. (2000). Record industry shoots itself with MP3 bullet. Retrieved October 1,
2001, from http://www.zdnet.com/intweek/stories/news/041 64,2551 103,00.html

Pratt, T. C., & Cullen, F. T. (2000). The empirical status of Gottfredson and Hirschi's
general theory of crime: A meta-analysis. Criminology, 38(3), 931-964.

Quantrun Corporation. (2000). The Growing Need for Gigabytes. Retrieved October 3,
2001, 2001, from http://www.quantum.coMsrc/ﬂ/sLneedshtm

Rahim, M. M., Seyal, A. H., & Rahman, N. A. (1999). Software piracy among computing
students: A Bruneian scenario. Computers and Education, 32, 301-321.

Reciprocal Inc. (2000a). Measuring the inﬂuence of music ﬁle sharing. Retrieved October
12, 2001, ﬁ'om http://www.reciprocal.com/pdf/ReciprocalﬁVNU reportpdf

Reciprocal Inc. (2000b). Reciprocal/VNU entertainment study reveals online ﬁle sharing
as likely cause of decline in college market album sales. Retrieved October 12,
2001, ﬁom http://www.reciprocal.com/pnn_re105242000.a§p

Reiman, J. H. (1995). The rich get richer and the poor get prison. (4th ed.). New York:
Wiley.

RIAA. (2000a). Anti-Piracy: Eﬂects. Retrieved July 20, 2001, from
http://www.riaa.com/Protect-Campaigr:3.cfrn

RIAA. (2000b). Complaint for copyright inﬁ'ingement against mp3. corn. Retrieved
October 10, 2001, ﬁom http://www.riaa.com/PDF/MP3_Complaint.p_df

293

RIAA. (2000c). Digital music laws. Retrieved September 24, 2001, ﬁom
http_://www.riaa.com/Copyright-Laws-4.cfnr

RIAA. (2000d). FAQ on Napster and Digital Music. Retrieved September 20, 2001, from
http://www.riaa.com/Napster.cfm

RIAA. (2000c). Federal Laws. Retrieved October 15, 2001, from
http://www.naa.conVqulm;ght-Lauts-2.cfm

RIAA. (20001). Napster lawsuit Q&A. Retrieved August 25, 2002, from
http://www.riaa.com/Napster.cﬁn

RIAAJ (2000g). The NET Act. Retrieved January 25, 2001, from
htfp://www.riaa.com/Protect-Online-2.cﬁn

RIAA. (2000h). Soundbyting: T op ten myths. Retrieved October 3, 2001, from
http://www.soundbyting.com/html/top 10 myths/myths index.html

RIAA. (2000i). Who we are. Retrieved August 4, 2001, from
http://www.riaa.com/About-Who.cfm

RIAA. (2003). Tracking music trends in America. Retrieved October 3, 2003, from
http://www.riaa.com/MD-Tracking.cfm

Robertson, M. (2000). MP3.com Response to RIAA Lawsuit Letter. Retrieved October 3,
2001, from http://www.MP3.com/response.html

Rogers, M. K. (2001). A social learning theory and moral disengagement analysis of
criminal computer behavior: An exploratory study. Unpublished doctoral
dissertation, University of Manitoba, Winnipeg.

Rondeau, E. (2003). Universities seek methods to curb illegal ﬁle sharing. Retrieved
September 5, 2003, ﬁ'om http://www.statenews.com/article.phtnrl?pk=l 8857

Rosen, H. (2000). RIAA Lawsuit Letter to MP3.com. Retrieved September 28, 2001, from
http://www.MP3.com/response2.html

Rosoff, S. M., Pontell, H. M., & Tillman, R. (2002). Proﬁt without honor: white-collar
crime and the looting of America. Upper Saddle River, NJ: Prentice Hall.

Roth, D. (1999). Desperately seeking MP3: A writer's quest. Fortune, 140(6), 264.

Royal Canadian Mounted Police. (2000). Computer crime, can it aﬁ'ect you? Retrieved
November 10, 1999, from httpzl/www3.sk.sympatico.ca/rcmpccs/cpu-crim.html

Schneier, B. (2000). Secrets and Lies: Digital Security in a Networked World. New York:
John Wiley & Sons.

294

Schneier, B. (2003). Beyond Fear: Thinking Sensibly about Security in an Uncertain
World. New York: Copernicus Books.

Schoen, S. (2002). BPDG one-page critique. Retrieved June 4, 2002, from
http://bpdg.blogs.efforg/archives/000121.html

Seale, D. A., Polakowski, M., & Schneider, S. (1998). It's not really theft! Personal and
workplace ethics that enable software piracy. Communication Abstracts, 21(4).

Sellers, C. S. (1999). Self-control and intimate violence: An examination of the scope and
speciﬁcation of the general theory of crime. Criminology, 37(2), 375-404.

Sharman Networks. (2003). Kazaa Media Desktop. Retrieved October 20, 2003, from
http://www.kazaa.com

Sherizen, S. (1997). Criminological concepts and research ﬁndings relevant for
improving computer crime. In R. C. Hollinger (Ed.), Crime, Deviance and the
Computer (pp. 298-305). Aldershot: Dartmouth.

Sims, R. R., Cheng, H. K., & Teegen, H. (1996). Toward a proﬁle of student software
piraters. Journal of Business Ethics, 15(8), 839.

Singleton, R. A. J ., & Straits, B. C. (1999). Approaches to Social Research (3rd ed.).
New York: Oxford University Press.

Skinner, B. F. (1953). Science and human behavior. New York: MacMillan.
Skinner, B. F. (1957). Verbal learning. New York: Appleton-Century-Crofts.
Skinner, B. F. (1971). Beyond Freedom and Dignity. New York: Knopf.

Skinner, B. F., & Fream, A. M. (1997). A social learning theory analysis of computer
crime among college students. Journal of Research in Crime and Delinquency,
34 , 495-5 18.

Smith, G. V., & Parr, R. L. (1989). Valuation of intellectual property and intangible
assets. New York: Wiley.

Solomon, S., & O'Brien, J. A. (1990). The effect of demographic factors on attitudes
toward software piracy. Journal of Computer Information Systems, 40-46.

Sony Corp. v. Universal Studios (464 US. 417; 220 U.S.P.Q. 665 1984).

Spring, T. (2000). Swap MP3s, go to jail? Retrieved October 12, 2001, from
http://www.cnn.com/2000/TECH/conrputing/04/ 1 4/MP3 .crackdownidg/ index .ht
ml

Stamp, M. (2002). Risks of digital rights management. Communications of the ACM,
45(9), 120.

295

Stamp, M. (2003). Digital rights management: The technology behind the hype. Journal
of Electronic Commerce Research, 4(3), 102-112.

Steffensmeier, D. (1989). On the Causes of White-Collar Crime - an Assessment of
Hirschi and Gottfredson Claims. Criminology, 27(2), 345-358.

Stenger, R. (2000). Campuses seek compromise over popular bandwidth hog. Retrieved
September 28, 2001, 2001, from
http://www.cnn.com/ 2000/TEC H/computing/OB/ 01 / napsterban/

Stenneken, J. (1999). MP3 Research Survey. Retrieved September 28, 2001, from
httpz/Avwwjoasten.webprovider.com/

Stone, R. (2000, 2000, June 16). MP3 Survey. Rolling Stone.

Sutherland, E. (1947). Principles of criminology (3rd ed.). Philadelphia: J .B. Lippincott.
Sutherland, E. (19493). White collar crime. New York: Holt, Rinehart, and Winston.
Sutherland, E. (1973). On analyzing crime. Chicago: The University of Chicago Press.

Sutherland, E. (Ed.). (1949b). The white collar criminal. New York: Philosophical
Library.

Swiatecki, C. (2000a).1 want my MP3. Retrieved September 20, 2001, from
http://www.statenews.com/editions/041400/ms MP31.html

Swiatecki, C. (2000b). Stiﬂ ﬁnes, jail can 't deter Napster users. Retrieved September 20,
2001, from http://www.statenews.com/editions/041400/ms MP32.html

Sykes, G. M., & Matza, D. (1957). Techniques of neutralization: A theory of
delinquency. American Sociological Review, 22, 664-670.

Tarde, G. (Ed.). ([1890] 1903). Gabriel T arde's laws of imitation. New York: Henry Holt.
Taylor, C. (2002, May 12). Burn, Baby, Burn. TIME.

Thong, J. L., & Yap, C. S. (1998). Testing an ethical decision-making theory: The case of
softlifting. Journal of Management Information Systems, 15(1), 213-237.

Tittle, C. R. (1980). Sanctions and social deviance: The question of deterrence. New
York: Praeger Publishers.

Tremblay, R. E., Boulerice, B., Arseneault, L., & Niscale, M. J. (1995). Does low self-
control during childhood explain the association between delinquency and
accidents in early adolescence? Criminal Behaviour and Mental Health, 5, 439-
45 1 .

296

Tyler, T. R. (1996). Compliance with intellectual property laws: A psychological
perspective. International Law and Politics, 29, 219-235. '

Vold, G. 3., Bernard, T. J ., & Snipes, J. B. (1998). Theoretical criminologr (Fourth ed.).
New York: Oxford University Press.

Wagner, S. C. (1998). Software Piracy and Ethical Decision Making. Buffalo: State
University of New York at Buffalo.

Walsh, A. (2000). Behavior genetics and anomie/strain theory. Criminology, 38, 1075-
1 107.

Walther, J. B. (1992). Interpersonal effects in computer-mediated interaction: A relational
perspective. Communication Research, 19, 50-88.

Walther, J. B., Anderson, J. F ., & Park, D. W. (1994). Interpersonal effects in computer-
mediated interaction: A meta-analysis of social and antisocial communication.
Communication Research, 21(4), 460-487.

Warr, M. (2002). Companions in crime: The social aspects of criminal conduct.
Cambridge: Cambridge University Press.

Warr, M., & Stafford, M. (1991). Inﬂuence of delinquent peers: What they think or what
they do? Criminology, 29, 851-866.

Webnoize. (2000). Napster university: From ﬁle swapping to the future of entertainment.
Retrieved September 21, 2001, from http://research.webnoize.com

Weekly, D. (1998). MP3 Summit Report. Retrieved October 3, 2001, 2001, from
http://www.MP3.com/news/064.html

Weekly, D. (2000). The MP3 Explosion. Retrieved October 8, 2001, from
http://david.weeklv.org/MP3book/ch1 .phn3#7

Weisbard, E. (2000). At Pho, a thousand e-mails a month track the great digital debate:
Keeping up with the Napsters, 2000, May 10-16, ﬁ'om
http://www.villagevoice.com/issues/OO19/weisbard.php

Wikipedia. (2003, July 10, 2003). Analog Hole. Retrieved August 29, 2003, from
http://www.wilgipedia.org/wiki/Analog_Hole

Winfree, T. L., Jr., Backstrom, T. V., & Mays, G. L. (1994). Social learning theory, self-
reported delinquency, and youth gangs: a new twist on the general theory of
crime and delinquency. Youth & Society, 26(2), 147-177.

Winship, C., & Mare, R. D. (1984). Regression models with ordinal variables. American
Sociological Review, 49(4), 512-525.

297

Wong, E. Y. H. (1995). How do we teach computer ethics? A short study done in Hong
Kong. Computers and Education, 25(4), 179-191.

Wong, G., Kong, A., & Ngai, S. (1990). A study of unauthorized software copying
among post-secondary students in Hong Kong. The Australian Computer Journal,
22(4), 114-122.

Wood, P. B., Cochran, J. K., Pfefferbaum, B., & Ameklev, B. J. (1995). Sensation-
seeking and delinquent substance use: An extension of learning theory. Journal of
Drug Issues, 25(1), 173-193.

Wood, P. B., Pfefferbaum, B., & Ameklev, B. J. (1993). Social psychological correlates
of delinquency. Journal of Crime and Justice, 16, 111-130.

Wood, W., & Glass, R. S. (1995). Sex as a determinant of software piracy. Journal of
Computer Information Systems, 36(2), 37-40.

Word, R. (2003). University's software kicks oﬂdownloaders. Retrieved November 22,
2003, from http://www.chron.com/cs/CDA/ssistory.mpl/tech/news/Z242112

Woudenberg, E. (2003). Minidisc FAQ. Retrieved October 30, 2003, from
http://www.minidisc.or,q/minidisc_faq.html#SCMS

Wright, B. R. E., Caspi, A., Mofﬁtt, T. E., & Silva, P. A. (2001). The effects of social ties
on crime vary by criminal propensity: A life-course model of interdependence.
Criminology, 39(2), 321-351.

Zeiler, D. (2003). Hackers bite Apple in its iTunes. Retrieved May 22, 2003, from
http://www.sunspot.net/technology/custom/pluggedin/bal-
mac052203,0,6600950.column?coll=bal-business-indepth

298

 

   

I"LillilljﬂljllﬂjljllﬁMW