New legal music services and their reduction on consumers’ intentions to pirate

(1)

New legal music services and their reduction on consumers’ intentions

to pirate

University of Groningen, Faculty of Economics and Business

Master Thesis. MscBA, specialization Strategy and Innovation

Robert Jeroen Messchendorp

Studentnumber: 1625942

Bankastraat 41a

9715 CD Groningen

Tel: 0653739611

(2)

2

Abstract

This study is first to investigate the effect of legal music services on consumers’ intention to pirate. The study investigates whether the perceived value of a legal service, measured by making use of the TAM framework combined with the literature on value research, influences consumers’ intentions to use and willingness to pay for music at the respective services. Moreover, it is investigated whether this influences consumers’ intentions to pirate. The results indicated that streaming music services are a more effective means rather than online music stores in decreasing consumers’ intentions to pirate. Furthermore, the results show it is of importance to increase the perceived value of the legal service in order to decrease consumers’ intentions to pirate. Lastly, the study shows indications that offering a large amount of content on a freemium base with some restrictions, or for a monthly subscription fee is perceived as being a fairer offer.

(3)

3

1. INTRODUCTION

During the 1990s the music industry enjoyed enormous growth with sales increasing from $7.5 billion in 1990 to a top of $14.6 billion in 1999. Thereafter the decline began, with global sales of recorded music falling by 7% in value and 8% in units in 2002 (Chiou, Huang and Lee, 2005) leading to cumulative sales of $8.5 billion in 2008. The advent of legal online systems rose during that year. In 2008, U.S. music sales through digital download grew 30 percent relative to 2007, whereas shipment of physical CDs fell 28 percent (IFPI, 2009).

Since the use of the Internet has become embedded in society, it has been debated for infringing intellectual property rights of record labels. Numerous authors have investigated the phenomenon of illegal downloading of copyrighted music by consumers all over the world (Kretschmer, 2000; Levin, Dato-on, and Rhee, 2004; Sinha and Mandel, 2008). Research has focused on areas such as the causes of illegal file-sharing (Halttunen, Makkonen, and Frank, 2010; Sinha and Mandel, 2008), the effects of file-sharing on record sales (Oberholzer-Gee and Strumpf, 2007), and on possible solutions for fighting piracy (Amberg and Schröder, 2007; Maltz and Chiappetta, 2002).

Nonetheless, digital piracy has proliferated during the last ten years, as new technological innovations in the areas of telecommunication and social media provide a large opportunity to use digital content illegally (Halttunen et al. 2010). Napster for instance, was home to approximately 60 million users which shared nearly 40 million songs,1 when the program was on its peak in terms of number of active users.

With these improvements in online infrastructure (e.g. wide spread high-speed internet) the possibility for entrants to set up legal, profit-realizing alternatives for sharing music such as iTunes and Spotify has increased. Nevertheless, artists and record labels still complain about these intermediaries, how they distribute their music digitally, and how they let consumers share their digital content. For example, Spotify is surrounded by controversy since some artists complain that Spotify is not open towards the public regarding the division of profits between their service and the artists. Especially the way that the company gives the feeling to their users that the artists are fairly remunerated for their efforts has led to criticism, as an individual artist needs their song be played millions of times to get substantial royalty revenues.2

(6)

6

Extant research focuses on illegal downloading and the possible consequences it has for record labels, artists and the recording industry as a whole. However, the current literature neglects the contribution of legal intermediaries to the decrease of the piracy rate. This paper focuses on recent legal online systems that provide a legal alternative to consume music via the internet. It focuses on streaming services such as Spotify, and whether they can contribute to the reduction of the piracy rate. Hardly any attention has been given so far to streaming services (Pfanner, 2009), and most of these contributions focus solely on the technical side of streaming services (Aurelius, Lagerstedt, and Kihl, 2011; Kreitz and Niemelä, 2010). Furthermore, a recent contribution of Waldfogel (2010) suggested that after the introduction of iTunes there has not been a significant drop in piracy rates. It needs to be investigated whether Spotify can be a more effective way of tackling piracy.

This paper contributes to the literature as it improves our understanding regarding the characteristics of new ways of obtaining digital content, how they affect consumer choice, and whether these new legal online systems can decrease the piracy rate in the music industry.

Considering the above, the following research question guides this paper:

To what degree can different legitimate music services reduce the intention to illegally share files? This research question leads to the following sub-questions:

- What are the motivations determining user intentions for music services? - What factors are of importance in determining the willingness to pay?

- Do legitimate offerings vary in the degree to which they can reduce the intention to illegally share/download files?

Although the list of digital music services is virtually infinite, it is beyond the scope of this research to include initiatives such as Last.fm and 22Tracks. Within the current research the focus will be on those intermediaries that allow for complete customer control in deciding which tracks to acquire and listen to, which is the case with initiatives as Spotify and iTunes, therefore these services are the guideline for this research.

(7)

7

2. THEORETICAL FRAMEWORK

2.1 Effect of (il)legal file-sharing on record sales

Since the advent of services such as mp3.com and Napster, record labels argued that the decline in record sales is solely due to lost sales by people downloading the albums. However, despite the fact that these record labels have won eventually the majority of the court rulings considering these kinds of cases, confusion exists in the academic world whether or not the record labels actually due suffer losses as a result of an increasing number of people downloading their albums.

Some studies, such as the study of Oberholzer- Gee and Strumpf (2007) showed different outcomes in this field. They analyze whether file-sharing has reduced the legal sales of music. The authors claim they are the first to study this subject employing empirical data on downloads of music files. They match an extensive sample of downloads to U.S. sales data for a large number of albums. In the U.S., households swap more than 300 million files each month, a figure that grew 50 percent during 2002 and 2003 (Karagiannis et al. 2004). Furthermore, in 2006 there were about 10 million simultaneous users on the major peer-to-peer (P2P) networks. Oberholzer-Gee and Strumpf (2007) conclude that file sharing has had only a limited effect on record sales.

Surprisingly, Levin et al. (2004) found that respondents who downloaded music have also acquired a greater number of CD’s in their collection than those who did not download. The authors concluded that downloaders bought more CD’s in the six months before the research than non-downloaders.

Waldfogel (2010) researches whether the availability of a convenient and widely adopted outlet for purchasing legal music has affected either the volume of unpaid consumption or the rate of displacement of paid by unpaid consumption. The author found that in general the respondents had more illegal than legal music, but the music obtained via file-sharing was, for the largest part, low-valuated music which respondents expected to not have bought anyway. Regardless of the strictness of approach chosen in the paper of Waldfogel (2010), the displacement estimate below 1 (absolute value) shows that much of the music people consume without paying for would otherwise not have been purchased.

(8)

8

This argument for file-sharing is also put forward by Andersen and Frenz (2010) who make a distinction between a substitution effect (where free music downloads are substituting for purchases) and a penetration effect (increased exposure through P2P file sharing leads to increased purchasing of those albums). They found that the substitution effect is strongest for well-known artists, while the penetration effect is strongest for unknown artists. The authors argue that a substitution effect depends, among other factors, on the ability to listen to the downloaded file in as many locations as the original version.

Liebowitz (2005) and Chellapa and Shivendu (2005) attack the possibility of a positive sampling effect by arguing that, although consumers may learn more about the music and make superior choices, record companies are not necessarily better off since the consumer is more interested in compiling a customized album with single songs from different artists rather than buying a full album of one artist which features a number of songs that are of a lower quality than the loose singles. At the same time record companies rather sell these full albums since their margins are much higher on the sale of a full album.

To conclude, although there are ambiguous results whether or not file-sharing is harmful for the record sales, previous research does not see file-sharing as the only cause of declining record sales.

2.2 Causes of illegal sharing of music files

Unauthorized use of music that can be downloaded or directly streamed from the Internet has changed the traditional forms of piracy like making copies of games or burning music CDs. While the latter traditional form of piracy still was perceived as innocent and done only by a few ‘absolute pirates’, digital piracy is seen as a serious threat to intellectual property right holders. Digital piracy has seen wide spread as a result of the easiness of copying and sharing files (Halttunen et al. 2010). Therefore, digital piracy is considered as more threatening, since it barely costs any effort to copy and share these files, as opposed to traditional pirating. Well known P2P networks such as Kazaa and Bittorent emerged at the end of the 90’s, which allowed many users to download and share music files.

Researchers seem to agree that the increase of copyright infringement is caused by several different effects, pertaining to both technical developments such as the increasing speed with which the Internet developed, as well as consumers’ ethical perception regarding obtaining music illegally (Halttunen et al., 2010; Sinha and Mandel, 2008). Consumers decide to illegally share files based on monetary and non-monetary perceptions (Doucet, 2012). In the following section the non-monetary, more ethical causes of illegal file-sharing are highlighted, while the section thereafter threats the monetary perception.

(9)

9

2.2.1 Causes of illegal file-sharing of non-monetary nature

Levin et al. (2004) found that consumers’ ethical decision making follows three steps. Firstly, a person will judge whether s/he thinks the act of illegal file-sharing is wrong. Secondly, a person will measure the level of harm done to the other party as a result of the first decision. Lastly, a person will weigh the relationship with the harmed party and will act more ethically towards the artists s/he has a more favourable association with. Despite the fact that Chiou et al. (2005) found that artists’ favouritism has a negative effect on the willingness to pirate, they also stated that when the general customers’ attitude towards music piracy is positive, favouritism is seen as less a boundary not to download.

Previous research demonstrates that consumers rationalize unethical behaviour by denying injury (Fullerton et al., 1996; Strutton et al. 1994). Consumers rationalize the behaviour with for instance the thought that downloading one song does not harm a multimillion business. Moreover, many respondents find that record labels are making excessive profits in comparison with the recording artists. Thus, in the eyes of the respondents the ones who created most of the value capture relatively less value.

Although the number of interviewees is rather small, the study of Halttunen et al. (2010) can be considered an important indication concerning the ethical values of consumers regarding downloading illegal content. In general, most interviewees were unconscious regarding what is legally right or wrong. Most interviewees were not knowledgeable about copyright laws, and the little knowledge they possessed was often faulty and/or confusing. This was caused by 1) the different national versions of copyright laws and 2) purposeful ethical interpretations. Furthermore, as the law on this subject becomes stricter the prosecution risk will become a more direct reason to switch to legal alternatives as well (Chiou et al., 2005).

Several authors (Albers-Miller, 1999; Coyle et al. 2009; Levin et al. 2004) argue that charging high prices leads to consumers feeling being ripped off. When the consumer perceives that the company’s profit is much larger than their own benefits, they are often hesitant to do business with such a company.

(10)

10

aligned to the absolute price, which refers to the monetary reason to take into account and which is explained in the following section.

Reason Author(s) Effect on willingness to

pirate G E N E R A L M O T I V A T I O N S F O R P I R A C Y

Singer idolization/favouritism (Psychic costs)

Chiou et al. (2005); Wilkes (1978) Negative

Prosecution risk (Psychic costs) Chiou et al. (2005) Positive (i.e. customers see

the chance of being caught as minimal)

Magnitude of consequences (Psychic costs)

Chiou et al. (2005)

Social Consensus (Psychic costs) Chiou et al. (2005) Positive

More loose in general attitude (Psychic costs)

Levin et al. (2004) Positive

Believe that illegal file-sharing hurts the record companies (Psychic costs)

Confusion about copyright laws (Search costs)

Halttunen et al. (2010) Positive

S P E C I F I C M O T I V

Unfair division of profits (Psychic costs) Coyle et al. (2009) Positive

Feeling of being ripped off (Psychic/Monetary costs)

Albers-Miller (1999); Coyle et al. (2009); Levin et al. (2004); Gupta et al. (2004); Kwong and Lee (2002)

Neither positive nor negative

Feeling of being ripped off (Psychic/Monetary costs)

Huang (2005); Klein et al. (2004) Positive

Price of a unit / Individual budget constraints (Monetary costs)

Possibility for

discovery/sampling/sharing (Time costs, search costs)

Andersen and Frenz (2010);

Liebowitz(2005); Peitz and Waelbroeck (2006); Waldfogel (2010)

Positive

(11)

11

2.2.2 Monetary cause of illegal file-sharing

Levin et al. (2004) argue that absolute price of music CDs is a point of concern for especially youngsters which often hesitate to pay €15 - €20 for just a single CD since such an amount is often a big blow out of their budget. Consumers may find the price unfair, as the price of a CD has not been reduced, even though it is much cheaper to record a CD these days, and promotion can be done cheaper than before as well. iTunes already passes some of the costs advantages on to the consumers by charging less than the traditional record store. Downloading a new album costs around €10 (Bockstedt, Kauffman, and Riggins, 2006), depending on the exclusivity of the album and the license Apple has obtained from the record labels or artist.

Nevertheless, the overall thought of the discussion in most of above papers was that legal alternatives would be preferred to illegal ones, when user friendliness was the same. Sirohi, Mclaughlin, and Wittink (1998) argue that when a product is perceived as an appropriate substitute for a similar product in terms of utility, the absolute price becomes less of a concern.

To conclude, some characteristics of the legal intermediaries might have the possibility to counteract some of the reasons why people use illegal file-sharing programs. Although these intermediaries will probably have no effect on the general ethical attitude of a person it does give consumers some incentives to make a switch to the legal intermediaries.

2.3 Ineffective measures against piracy

At the beginning of the 2000’s the music industry introduced Digital Rights Management (DRM) as a preventive measure against copyright infringement. This technique is used to encrypt content or restrict the access to the file. Restrictions range from limiting the number of plays, duration of access, or the number of devices the file can be played on.

Consequently, many advocates contend that DRM is ‘harmful to consumers because it tilts the balance of control in favor of copyright holders’ (Einhorn and Rosenblatt, 2005, p.1).

(12)

12

In 2001, the major labels cooperated somewhat in two attempts to sell digital music legally themselves. In an effort to provide consumers with appealing options to counter unpaid illegal file-sharing Sony and Vivendi Universal came with ‘Pressplay’ and AOL Time Warner, Bertelsmann, and EMI came with ‘MusicNet’. Consumers of ‘Pressplay’ were initially allowed 50 downloads and 500 streams. In 2002 they changed the offering to allow consumers to download without a numerical restraint at a price of $179.40 per year.

With the latter option, ‘MusicNet’, consumers paid $9.95 per month and than they were allowed to download 100 songs, out of a library of 75000 tracks. Furthermore, these tracks could not be put on a CD or any portable device, and the working of them expired after 30 days. Both efforts never really became popular alternatives. Reasons for this cited in the literature are the complex structure of payment for the services (Einhorn and Rosenblatt, 2005), impossibility to burn the tracks on CD or transfer them to a portable device, and the absence of a cooperative strategy among the major record labels by starting two independent services. As a result of this latter strategy both services were judged as unattractive, as their music-catalogues were perceived as incomplete (Moreau, 2009).

2.4 Which legal music services have the potential to decrease the intentions of

music piracy?

Although effectively fighting piracy is often seen as being difficult, some researchers in the past made constructive suggestions about how digital piracy might be brought to a halt. These authors, for instance, called for business models that better accommodate consumers’ desires (Amberg and Schröder, 2007; Chiou et al., 2005).

Goel et al. (2010) explain that the strategy favoured by the recording industry used to consist of seeking stronger copyright laws with wider coverage, because the recording industry cites consumers’ fear of being sued as a significant factor lowering the amount of P2P activity. This resulted in a complex system that arguably benefits neither the recording companies nor its users. Currently rights are distributed across multiple industry players, with each player demanding a royalty, resulting into increasing transaction costs. In contrast, obtaining and sharing digital files illegally is easy.

(13)

13

argue that the cost efficiencies of electronic production and distribution must be recognized and passed on to the purchasers, leading to price reductions.

Amberg and Schröder (2007) aim with their paper to evaluate to what extent present e-business models for digital music distribution meet the consumers’ expectations. The authors hold two criteria with regards to the examination of the business models. The first criterion is the method of payment (e.g. flat rate or pay per download) for digital music, and the second criterion is the dependency on the supplier or its technology (e.g. files are delivered in a specialized format that makes it more difficult to share them after obtaining them). With these criteria Amberg and Schröder (2007) identified four categories of business models, named category A till D which are shown in table 2.4.1.

Amberg and Schröder (2007) conclude that category B e-business models came closest to the consumers’ expectations, while they also acknowledged the viability of flat rate payment models, by stating that the monthly fee is not higher than the budget of the interviewees in the survey. Furthermore, they point out those services are offering the largest amount of content to their users.

In a similar vein, Goel et al. (2010) came up with four alternative future scenarios and business models appropriate to each scenario in the context of the overall business strategy. Two scenarios describe a situation with a non-enforceable copyrights situation. In such a situation revenue is claimed to be gathered through the products complementary to the music, whereas in the scenario with enforceable copyright a call is made for new business models. New technologies have increased the power of the individual consumers who can download singles (as opposed to relatively more expensive albums favoured by the record companies) for free over the Internet or listen to music over social networks or sites that stream music on demand.

Einhorn and Rosenblatt (2005) predicted that the industry would switch to streaming in the long run, since profit margins are considerably higher through licensing fees that can be adjusted more appropriately. This offers both an incentive to the organization offering the service, and the consumer, since streaming services come in a monthly subscription deal with libraries available of millions of songs. Consumers receive a large variety of music for a low monthly fee, or even for free with just a few limits.

Method of payment Dependency on supplier/technology

Pay per download Flat Rate

Dependent Category A (iTunes) Category C (Spotify)

(14)

14

However, Jeff Price, founder of TuneCore, a company that acts as a liaison between artist and the digital marketplace doubts the capacity of streaming services such as Spotify to eliminate piracy. He argues that cloud-streaming services, such as Spotify, are related to listening on demand while people download because they want to have the music on their hard-drive so that they can transfer it to portable devices. Nevertheless, although streaming services are unlikely to convert pirates into paying customers, it might reduce the tendency for future piracy as the free streaming service might limit the incentive to engage in piracy (Greenwood, 2010).

Although no convincing empirical evidence suggests that streaming content decreases or even eliminates piracy, the signs are positive, as reported by the 2012 Digital Music Report from the IFPI. According to Moore (2012) ‘the number of paying subscribers to services such as Spotify and Deezer3 has leaped in the last twelve months, from an estimated 8 to more than 13 million.’ This has lead to an eight percent increase in digital revenues in 2011 compared to 2010.

Pfanner (2009) argues that more people are starting to use these new legal intermediaries, while at the same time the percentage of users that admits they are engaging in illegal file-sharing diminishes. Moreover, it is argued that rather than cannibalizing existing digital sales on for instance iTunes, streaming intermediaries attracts new customers, the ones that were previously engaged in illegal file-sharing4.

Chu and Lu (2007) and Kwong and Park (2008) predict that important factors in the valuation of legal digital music services in the minds of the consumers are the ease of use of a service, the usefulness of a service and the playfulness of such services. Other authors (Chiang and Assane, 2007; Halttunen, 2010; Sinha and Mandel, 2008) use similar terms such as services should be ‘simple and easy to use’, and ‘provide ease of navigation.’ Moreover, the majority of these authors argue that a large selection of content should be available, and that utility prices should be lowered.

In table 2.4.2 (Appendix 1) an overview of the major legal music services available in the countries of interest can be found. As one can see the different services within a category (either pay-per-download or flat-rate services) do not differ very much. After analyzing previous literature and the already established new legal initiatives a division can be made between the new legal initiatives in terms of the payment model. Three different models are currently in place in the music industry, namely pay-per-download, flat-rate payment models which offer access to a large library of music, and a combination between the former two. The combination deal offers a monthly flat-rate for a specified number of songs. Once this limit has been reached, additional songs can be downloaded by paying extra per song.

3

Deezer is a web-based on demand streaming music service 4

(15)

15

2.5 Conceptual Model

This section describes the conceptual model that aims to explain consumers’ intentions to illegal file-sharing based on their perceptions regarding legal music services. It is proposed and visualized in figure 1 that the perceived value of a legal music service influences the intentions to use a legal service and the willingness to pay for a legal music service. Moreover, it is suggested that the motivations for illegal file-sharing as identified in earlier research have an influence on the willingness to pay for the legal services, just as the strictness of intellectual property law is of influence. Lastly, it is proposed that the intentions to illegally share/download files are reduced by both the intentions to use a legal service and the willingness to pay for a legal service.

H1a/b: + H5a/b: - H6a/b: -

H2a/b: + H4a H3: H4b

Figure 1. Conceptual model Perceived value of the legal services:

 Online Music Store (OMS)

 Streaming music service (SMS)

Intentions to use a legal service (OMS & SMS)

Willingness to pay for a legal service (OMS & SMS)

Motivations for illegal file-sharing:

(16)

16

Extant research (Childers et al., 2001; Chu and Lu, 2007; Davis 1989; Kwong and Park, 2008; Lederer et al., 2000; Van der Heijden, 2004) confirmed that consumers look at the value a particular product or service gives them when making a decision regarding the intentions to use such a product or service. The current research combines the theory of the Technology Acceptance Model (TAM) on information systems with perceived value research and assumes that the perceived value of both music services is determined by combining the perceived ease of use, perceived usefulness, and perceived enjoyment of the service. Ease of use refers to how easy consumers perceive it to consume and search for music, the easier and more convenient consumers can find and listen to the music, the lower the perceived time and search costs are. The perceived usefulness is related to quick access to music and quality of the service. Moreover, in hedonic shopping situations similar to the use of a legal music service the enjoyment perceived by consumers while making use of the service is also acknowledged as being an important factor in their intentions to use that service. Following the TAM framework users are more inclined to accept and/or use the service to the extent they believe it helps them to perform their task (in this case acquiring music) in a better way (Chu and Lu, 2007). The TAM framework argues that higher perceptions of ease of use, usefulness and enjoyment lead to higher intentions to use a certain technology. As described both in the literature on the causes of illegal file-sharing and the literature on the new music services consumers argue that legal alternatives would be preferred in comparison to illegal alternatives when these legal alternatives offer a similar functionality (i.e. perceived ease of use) and utility (i.e. perceived usefulness). Therefore, hypotheses 1a and 1b are:

 H1a: The perceived value – based on perceptions of ease of use, usefulness, and enjoyment- of an online music store (OMS) service has a positive effect on the intention to use an online music store to acquire digital music

 H1b: The perceived value of a streaming music service – based on perceptions of ease of use, usefulness, and enjoyment has a positive effect on the intention to use a streaming music service (SMS) to acquire digital music

Existent research (e.g. Homburg, Koschate, and Hoyer,2005; Pihlström and Brush, 2008; Zeithaml, 1988) makes use of (adapted) versions of the model of Zeithaml (1988) which describes that the perceived value of a product or service predicts the consumers’ willingness to buy/pay for a product or service. Higher perceptions of value translate into a higher willingness to pay for the alternative. The perceived value of the legal services is determined by making use of the three factors described by the technology acceptance model. Therefore, hypotheses 2a and 2b are:

(17)

17

 H2b: The perceived value of a streaming music service (SMS) – based on perceptions of ease of use, usefulness, and enjoyment- has a positive effect on the willingness to pay for a streaming music service

Several motivations, such as unfairness of price, no harm, and unfair division of profits, are identified as being motivations for illegal file-sharing, however, the current literature (e.g. Chiou et al. 2005; Levin et al. 2004; Sinha and Mandel, 2008) provides no clear answers regarding which and whether these motivations also influence the willingness to pay for a legal service. Since there is little empirical evidence for the direction of the relationships, this study tests the strength of relationships without a priori hypothesizing any direction.

 H3: The strength of motivations for illegal file-sharing influence the willingness to pay for legal services

When the strictness of intellectual property law increases, consumers tend to perceive a higher risk of prosecution, and the “value of the illegal services” will decrease in the minds of the consumers. This will, in turn, influence their willingness to pay for a legal service and their intention to use illegal services. The recording industry itself indicates that the strictness of the law and the perceived risk to face prosecution are leading factors in generating fear amongst consumers to be sued (Chen, 2007). Both a direct effect on the intentions to illegally share files, and an indirect effect through the willingness to pay are expected. Following this, hypotheses 4a and 4b are:

 H4a: The strictness of intellectual property law influences the willingness to pay

 H4b: The strictness of intellectual property law influences the intentions to illegally share files Consumers develop more positive intentions to use and a higher willingness to pay for the legal services when these services are delivering an appropriate value as perceived by the consumers (Chu and Lu, 2007; Davis, 1989; Pihlström and Brush, 2008; Zeithaml,1988). Most existing studies find at least some substitution effects between the use of legal services and the use of illegal services (e.g. Andersen and Frenz, 2010; Bustinza, Vendrell, and Parry 2013; Halttunen et al. 2010; Liebowitz, 2008; Mansell and Steinmueller, 2011). Therefore it is argued that when consumers have higher intentions to use a legal service, and/or a higher willingness to pay this will negatively influence their intentions to use an illegal service to share files. Following this, hypotheses 5 and 6 are:

 H5a: The intentions to use an online music store reduces the intentions to illegally share files  H5b: The intentions to use a streaming music service reduces the intentions to illegally share

(18)

18

 H6a: The willingness to pay for an online music store reduces the intentions to illegally share files

(19)

19

3. METHODOLOGY

The objective of this paper is to investigate consumers’ perceptions and intentions to use and/or pay for new legal music services, and what the effects of the legal alternatives are on the consumers’ intentions to illegally share files. The methodology section will explain how the research is executed, while addressing the reliability and validity of the research.

3.1 Data sources and methods of data collection

This study makes use of empirically collected data. The primary data concerning consumer attitudes regarding illegal/legal file-sharing is gathered by means of a survey, which is distributed among college students and also on localized music forums, as members of such a group often have a deeper interest in the topic. A complete list of all the internet blogs that were contacted, and the ones that responded positively can be found in Appendix 2.

3.2 Sample

For the selection of the sample, this research focused on young people. Earlier research identified a specific target group of youngsters (age 19-31) as these are generally assumed to belong to the group that often engages in illegal file-sharing (Bounie et al. 2007; Halttunen et al., 2010; Levin et al., 2004). Coyle et al. (2009) state that people within this age range are generally more interested in digital music, and got acquainted with downloading and sharing digital music at an early age and are therefore very knowledgeable regarding the different techniques. The survey will follow some elements out of the surveys of earlier named authors, edited to fit the technological changes in file-sharing. However, as these papers are already five to six years old, the current research assumes that by nature this group has widened slightly and therefore the target group has widened slightly and for this research the initial target group was set at 18-36 years old.

3.3 Procedure

(20)

20

The survey is constructed with the utmost care to prevent respondents from giving socially desirable answers, by making use of neutralized terms in the survey (Nederhof, 1985). Therefore, the term illegal is refrained from use, and unpaid service is used instead.

The first part of the survey includes a few general questions about their experience with and usage of digital music. After these questions, the respondents are divided into one of three ‘market segments’, namely illegal file-sharers, people who predominantly use an online music store (e.g. iTunes), and people who are predominantly are using some form of the streaming monthly flat-rate alternatives (e.g. Spotify). By making use of such a division it is possible to test the acceptance of these streaming alternatives, where one ‘rents’ the music as opposed to owing the music. Furthermore, it is possible to test whether either of the legal business models leads to a reduction in the intention to illegally share files. As such, the questionnaire offers possibility to test between subjects and within-subjects as grouping variables can be made (Field, 2006).

3.4 Measurement of constructs: Principal component analysis

Field (2006) describes principal component analysis (PCA) as ‘a technique for identifying groups or clusters of variables (p.619).’ This analysis is performed in order to reduce the number of variables which are taken into account. Hereby a more manageable data set is created while still measuring the underlying dimensions. During the PCA the Kaiser criterion (eigenvalue criterion) and the percentage of variance explained criterion are mostly unambiguous in their judgment on how many components to extract and therefore these are followed.

3.4.1 Perceived value of the legal music services (OMS/SMS)

(21)

21

3.4.2 Motivations for illegal file-sharing

Motivations for illegal file-sharing consist out of several items that were identified from the literature as motivations for people to use an illegal file-sharing service. The items artist ‘favouritism’ which tests for the fact that people being a fan of a particular artist might withhold them from downloading music from this artist illegally, ‘no harm’, ‘unfair division of profits’ and ‘prosecution risk’ are measured by one item in the questionnaire, while ‘unfairness of price’ is measured by the two items ‘I consider a paid download (individual song or whole album) too expensive’ and ‘I believe that the price of a album is too expensive for what you get in return.’

3.4.3 Intention to use legal option (OMS/SMS)

The two items that are part of this scale form one neat component that explains respectively 92% and 86% of the variation. The intention to use is either one of the services is measured by two items, respectively ‘In the future I intent to make predominantly use of [service] to acquire digital music’ and ‘I intent to recommend others to use [service] to acquire digital music.’

To control that both components truly measure the intention to use their specific service the analysis is performed for both services combined, and as expected, the four items displayed two neat components measuring only for the service they intended to measure.

3.4.4

Willingness to pay (OMS/SMS)

The willingness to pay for a service is determined by making use of a hypothetical willingness to pay for a service (Voelckner, 2006). To be more specific the contingent valuation approach is used in letting the respondents decide whether they pay for a service at a certain price (Wertenbroch & Skiera, 2002). For the willingness to pay all five items loaded on one component in the case of the OMS users, however in the case of the SMS users the Kaiser criterion suggested extracting two components, while the percentage of variance explained was already sufficiently satisfied with one component. Therefore, the reliability analysis is consulted, which shows one component still delivers a very reliable scale, thus the choice was made to extract one component.

(22)

22

3.4.5 Intention to illegally share files

As expected the two items that measure the intention to illegally share files (‘In the future I intent to decrease the use of P2P services/ Google search to acquire unpaid digital music’ and ‘In the future I intent to stop the use of P2P services / Google search to acquire unpaid digital music’) both load on the same component, and this is confirmed both by the Kaiser criterion and the percentage of variance explained criterion. Furthermore both items had a communality value of above .5 thereby meeting the threshold.

3.4.6 Strictness of intellectual property law

(23)

23

Scale name Abbreviation Number of items

retained λ (OMS/ SMS) α (OMS/S MS) Sources

 Perceived value of the legal services (OMS/SMS)

PVOMS/PVSMS 9 items .69/.60 .94/.92 Deveraj et al.

(2002) Childers. et al.( 2001) Chu and Lu (2007) Kwong and Park (2008)  Motivations for illegal

file-sharing o Unfairness of Price o No harm o Unfair division of profits o Favouritism o Prosecution Risk Chiou et al. (2005), Wilkes (1978), Levin et al. (2004), Coyle et al.(2009) Unfairness of Price 2 items .70 .57 No harm Downloading a few songs or albums for

personal use does not harm the artist or record company

N.A. N.A.

Unfair division Record labels capture too much value compared to the artists, who created the value of the music

N.A. N.A.

Favouritism I only download from artists that

are not my

favourite (recoded)

N.A. N.A.

Prosecution risk Please assess the likelihood of getting in trouble for downloading/ uploading files in your country N.A. N.A.

Willingness to pay for legal option Online music store (OMS)/streaming music service (SMS) WTP OMS WTP SMS 5 items 5 items .69/.64 .89/.85 Voelckner (2006); Wertenbroch & Skiera (2002) Intention to use legal option

(OMS/SMS) Int. OMS Int. SMS 2 items 2 items .92/.86 .91/.84 Sirohi et al. (1998); Venkatesh & Morris (2000)

Intention to illegally share files (recoded)

IIFS 2 items .77 .70

Strictness of Intellectual property law The Netherlands:

55 texts Sweden: 70 texts The United Kingdom: 144 texts Ireland: 147 texts

N.A. N.A. Van

Kranen-burgh and

Hogen-brink (2005)

(24)

24

4. RESULTS

In this chapter the results of the statistical tests, that are conducted in order to be able to draw conclusions regarding the dataset, are displayed. The purpose of this chapter is to present and examine the results of the data. Firstly, some general descriptive statistics describing the data, such as the respondents’ age, gender and ‘user group’ are displayed. These are followed by various statistical tests such as t-tests, one-way analysis of variance post-hoc LSD tests, a correlation test and to test the model in the end a regression analysis is performed. The sample consists out of 154 returned questionnaires of which 103 were useful, as 51 respondents only gave information regarding the first four questions, thus these responses did not allow for further analysis.

4.1 Respondent profile

More than half of the valid responses (55.3 percent) answered that they are predominantly using an unpaid service to gather their digital music. Furthermore, 32 percent confirmed they already predominantly used a streaming media service such as Spotify, and the remaining 12.6 percent used pay per download services such as iTunes.

Frequency Percentage Country of origin NL SE UK IRL

Illegal file-sharers 57 55.3 49 2 2 4

Online music store users 13 12.6 8 3 2

Streaming media users 33 32.0 22 5 4 2

Total 103 79 7 9 8

Table 4.1. Division of sample per user groups (and per country)

The respondents generally ranged in terms of age between 13 years old and 36 years old (M=25.13) Furthermore, the division of men and women was quite equally divided between both genders as 59 percent of the respondents are men, and 41 percent women.

4.1.1 Listener type differences per service

(25)

25

4.2 Mean differences for PV, Int., and WTP of both services between user

groups

The one-way ANOVA compares the means of three or more groups. By performing a one-way ANOVA post-hoc LSD it is possible to test whether the difference in means between the three groups are significant, and thus whether these groups truly have different opinions on the two different legal services. In the following sections the author uses the terms PV to describe the perceived value of a particular service, Int. to describe the intentions to use a particular legal service, WTP to describe the willingness to pay for a particular legal service, and IIFS to describe the intentions to illegally share files.

Illegal file-sharers

Online Music Store users Streaming Music Service users

F

N

51

13

30 PVOMS

2.53a 4.36b 2.39a

23.14 N

48 10 33

PVSMS

2.87a 2.99a 3.82b

13.29

Table 4.2.1 Mean differences of the PV for both services between user groups

Mean differences in rows with differing subscripts are significantly different at the p <.05 level

The results show that on average the online music stores users also perceive the service they use as highest, whereas the streaming media service users have the lowest opinion from the three groups regarding the perceived value of online music stores. Both the illegal file-sharers and streaming music service users argue that the statements regarding the perceived value of an online music store does not reflect their opinions concerning the perceived value of an online music store.

(26)

26

Illegal file-sharers Online Music Store users Streaming Music Service users F

N 51 13 30

PeouOMS 2.56a 4.49b 2.49a 18.15

PuOMS 2.94a 4.49b 2.83a 11.26

PeOMS 2.08a 4.10b 1.83a 31.27

Table 4.2.2 Mean differences of the PV for OMS between user groups when displaying the separate components Mean differences in rows with differing subscripts are significant at the p <.05 level

N 48 10 33

PeouSMS 3.33a 3.23a 4.13b 7.34

PuSMS 2.95a 2.80a 4.01b 12.61

PeSMS 2.65a 2.43a 3.96b 15.90

Table 4.2.3 Mean differences for PV for SMS between user groups when displaying the separate components Mean differences in rows with differing subscripts are significant at the p <.05 level

In table 4.2.2 the mean differences are displayed for the separate components as defined in the technology acceptance model (TAM). The results show that both illegal file-sharers and the streaming music service users especially see the acquiring of music by making use of an online music store as not enjoyable. As the table shows, the differences in perceived value of online music stores from illegal file-sharers and streaming music service users are not significant, which means that the illegal file-sharers and streaming music service users share largely the same opinion regarding the perceived value of online music stores.

In table 4.2.3 the mean differences for the separate components concerning the perceived value of streaming music services are displayed. Although the streaming music service users are most positive regarding their own service, all three user groups see the streaming music services as quite easy to use. Moreover, although there are no significant differences in value perceptions of streaming music services between illegal file-sharers and online music store users, the former group seems to have a slightly more positive value perception of streaming music services than the latter user group have.

N 56 12 32

Int. OMS 1.75a 4.25b 1.36c 69.17

Int. SMS 2.40a 2.79a 4.28b 43.67

Table 4.2.4 Mean differences of the intentions to use a service between the user groups Means in rows with differing subscripts are significant at the p <.05 level

(27)

27

Illegal

file-sharers

Online Music Store users

Streaming Music Service users F

N

53

13

31 WTPOMS 1.94

a

3.63

b

2.21

a

17.63 N

52

13

32 WTPSMS 2.30

a

2.95

b

3.51

b

19.06

Table 4.2.5. Mean differences of the willingness to pay for a service between the user groups Means with differing subscripts are significant at the p <.05 level

Table 4.2.5 displays the differences in willingness to pay for either service between different user groups. The respondents who predominantly use an illegal file-sharing program and the respondents who predominantly use a streaming music service do not differ significantly from their willingness to pay for an online music store. Moreover, as expected the online music store users have the highest willingness to pay for their service, just as the streaming music service users have the highest willingness to pay for their service.

Illegal file-sharers Online Music Store Users Streaming Music Service Users F N 56 12 32 IISF 3.64a 2.67b 3.02b 6.45

Table 4.2.6 Mean differences of the intention to illegally share files between the user groups Means with differing subscripts are significant at the p <.05 level

(28)

28

4.3 Mean differences per user group in PV, Int., and WTP between OMS and

SMS

By means of a paired sample t-test it can be tested whether respondents from one particular group (e.g. illegal file-sharers) differ significantly in their opinion between both services. As such, this part of the results makes use of a within-subjects design as opposed to the between subjects-design used in the previous section.

N

PVOMS PVSMS

Illegal file-sharers

48 2.55

a

2.89

b

Online music store users

10 4.47

a

2.99

b

Streaming music service users 30 2.39

a

3.77

b

Table 4.3.1 Mean differences per user group between the perceived value of an online music store and a streaming music service

Means in rows with differing subscripts are significant at the p <.05 level

Table 4.3.1 shows that the mean differences for each individual group are significant, meaning that each group perceives a difference between the value of an OMS and the value of a SMS. The illegal file-sharers show a higher value perception with regards to the streaming music service. Moreover, the online music store users perceive the value of the streaming music service more positively than the streaming music service users perceive the value of the online music store.

N

Int.OMS Int. SMS

Illegal file-sharers

56 1.75

a

2.40

b

Online music store users

12 4.25

a

2.79

b

Streaming music service users 32 1.36

a

4.28

b

Table 4.3.2 Mean differences per user group between the intentions to use an online music store and a streaming music service

(29)

29

N Willingness to pay for OMS Willingness to pay for SMS

Illegal file-sharers 52 1.94a 2.30b

Online music store users 13 3.63a 2.95b

Streaming music service users 31 2.21a 3.47b

Table 4.3.3 Mean differences per user group between the willingness to pay for an online music store and streaming music service

Table 4.3.3 displays the results for the willingness to pay for either service per group. Naturally the user groups show the highest willingness to pay for using their own service. Moreover, the illegal file-sharers seem to have a higher willingness to pay for streaming services in comparison with their willingness to pay for online music stores.

4.4 Cross-country mean differences of PV, Int., and WTP between OMS and

SMS

Since there is also some data collected from respondents outside the Netherlands it might be of interest to see whether there are differences between countries in how they value the different services, their intentions to use, and their willingness to pay for either service. These differences are tested by making use of an ANOVA post-hoc LSD test. Results of these tests are displayed in tables 4.4.1 till 4.4.4. However, one should take into account that there is a large difference between the numbers of respondents from the countries, thus no definitive conclusions can be made from these figures.

The

Netherlands

Sweden The United

Kingdom

Ireland F

N 68 7 9 8

PVOMS 2.61cd 2.49 3.38a 3.61a 3.18

Table 4.4.1a Mean differences of the perceived value of an online music store between the different countries

a = significant difference with The Netherlands, b = significant difference with Sweden, c = significant difference with The United Kingdom, d = significant difference with Ireland, significant at the p <.05 level

The

Netherlands

Sweden The United

Kingdom

Ireland F

N 65 7 9 8

PVSMS 3.19 3.41 3.68 3.34 .85

Table 4.4.1.b Mean differences of the perceived value of a streaming service between the different countries

(30)

30

to the statements. The perceptions of the value of the streaming music services show no significant differences between the various countries.

The Netherlands Sweden The United

Kingdom

Ireland F

N 76 7 9 8

IntOMS 1.74cd 1.78d 2.5a 3.25ab 5.71

IntSMS 2.86 4.14a 3.44 3.68 3.71

Table 4.4.2 Mean differences of the intentions to use for both services between countries

a = significant difference with The Netherlands, b = significant difference with Sweden, c = significant difference with The United Kingdom, d = significant difference with Ireland, mean differences are significant at the p <.05 level

In table 4.4.2 the differences that are interesting are that the intentions to use an online music store are quite high for the UK and Ireland in comparison with the Netherlands and Sweden, whereas Sweden has by far the highest intention to use a streaming music service. These high intentions for the UK and Ireland are not very surprising as their perceptions of the value of these services (table 4.2.1) are also (significantly) higher than the perceptions of The Netherlands and Sweden regarding the perceived value of the online music stores.

Table 4.4.3a Mean difference of the willingness to pay for an online music store between countries

The

Netherlands

Sweden The United

Kingdom

Ireland F

N 73 6 8 8

WTPSMS 2.77 3.27 2.76 2.80 .40

Table 4.4.3b Mean difference of the willingness to pay for an online music store between countries

By taking a look at tables 4.4.3a and b it becomes clear that the separate countries are quite unanimous regarding their willingness to pay for either service. The only significant difference visible is that of the willingness to pay for online music stores, which is higher in Ireland than in the other countries.

The

Netherlands

Sweden The United

Kingdom

Ireland F (sig.)

N 74 6 8 8

(31)

31

The Netherlands Sweden The United

Kingdom Ireland F (sig.) N 77 7 9 8 Unfair division of profits 3.43b,d 4.43a,c 3.56b 4.13a 4.80 No harm 3.56d 2.86 3.33 2.75a 2.60 (.056) Prosecution risk* 1.69 1.43 1.56 1.50 .71 Unfairness of price 3.61 3.78 3.38 3.00 1.52 Favouritism 3.83 3.42 4.11 3.25 1.50

Table 4.4.4 Mean differences in the motivations for illegal file-sharing between countries

* measured on a four points scale from very unlikely to very likely

(32)

4.5 Assessment of correlations

No harm Unfair div. Favouritism Price unfairness Prosecution risk

PVOMS PVSMS IntOMS IntSMS WTPOMS WTPSMS Strictness of law Int. to ill sf. No harm Unfair division .11 Favouritism .08 -.24* Price unfairness .22* .13 -.06 Prosecution risk .02 -.15 .01 -.05 PVOMS -.13 .11 .02 -.40** .03 PVSMS -.10 .00 .06 .10 -.13 .27** IntOMS -.27** .07 -.12 -.43** .13 .71** -.04 IntSMS -.21* .16 -.10 .18* -.16 -.00 .50** .02 WTPOMS -.31** -.07 -.05 -.47** .14 .54** .19 .53** .10 WTPSMS -.29** -.07 .16 -.04 -.11 .02 .53** -.05 .48** .43** Strictness of law -.23* .32** -.09 -.14 -.16 .18* .09 .28** .27** .14 .03 Intentions to illegally share files .19* -.01 .19* .15 .06 -.22* -.31** -.32** -.48** -.33** -.33** -.06

Table 4.5 Results of bivariate correlations between constructs *= p <.05

(33)

A correlation analysis is performed to investigate the relationships between the constructs that are tested in the model. Table 4.5 displays the interrelationships existing between the variables that are of concern when testing the different relationships displayed in the conceptual model.

The table shows that there is high relationship between the perceived value of the online music stores and the intentions to use an online music store (r=.71, p <.01). This is also the case for the perceived value of the streaming music services and the intention to use these services (r=.50, p <.01). Other high correlations are among the perceived value of an online music store and the willingness to pay for an online music store (r= .54, p <.01), perceived value of a streaming music service and willingness to pay for a streaming music service (r=.53, p <.01).

Moreover, not surprisingly there are also relatively high correlations between intentions to use online music stores and willingness to pay for online music stores (r=.53, p <.01) and intentions to use streaming music services and willingness to pay for streaming music services (r=.48, p <.01).

The negative correlation between strictness of law and the independent variable ‘no harm’ might be home to a signalling effect, meaning that when the laws becomes stricter the respondents might be showing less agreement to the statement that downloading does not harm the record labels or artists.

(34)

34

4.6 Regression analysis

Due to the nature of the research the regression analysis has to be performed multiple times, with changing dependent and independent variables. Next to this, the analysis had to be performed two times, once for both services (i.e. online music stores and streaming media services).

Hypothesis Accepted/Rejected (OMS/SMS)

H1a/b: The perceived value of an OMS/SMS has a positive effect on the intention to use an OMS/SMS to acquire digital music

Accepted/accepted

H2a/b: The perceived value of an OMS/SMS has a positive effect on the willingness to pay for digital music at an OMS/SMS

Accepted/accepted

H3: Motivations for illegal file-sharing influence the willingness to pay for legal services

a) No harm done with downloading b) Unfair division c) Favouritism d) Unfairness of price e) Prosecution risk Accepted/accepted Rejected/rejected Rejected/rejected Accepted/rejected Rejected/rejected

H4a/b: The strictness of intellectual property law influences the willingness to pay and the intentions to illegally share files

Rejected/rejected

H5a/b: The intention to use an OMS/SMS reduce the intentions to illegally share files

Rejected/accepted

H6a/b: The willingness to pay for an OMS/SMS reduces the intention to illegally share files

Rejected/rejected

(35)

35

4.6.1 Dependent variable: Intention to use legal option (OMS/SMS)

The first relationship to test is between the independent variable ‘perceived value of the legal services (OMS)’ and the dependent variable ‘intentions to use a legal option (OMS).’ The R2 coefficient has a value of .50 which means that 50% of the variance in the intention to use an online music store is explained by the perceived value of that online music store. Based on the F-test showing a p-value of .000 it can be concluded that there is indeed a relationship between the independent variable and the dependent variable. The unstandardized regression coefficient (b) has a value of .740. Moreover, the independent variable has a positive significant effect (p <.001). Thus H1a is accepted as the perceived value of the online music store has a positive effect on the intention to use an online music store.

When entering the separate components that TAM prescribes (perceived ease of use, perceived usefulness, and perceived enjoyment) as independent variables the adjusted R2 explains 52% of the variance in the intention to use a legal option. Moreover, the analysis shows that the perceived ease of use of an online music store has a positive significant effect (p<.05), just as the perceived enjoyment of the online music store (p<.01).

The same tests are performed in the same order for the intention to use a streaming media service. R2 displays that 25% of the variance in the intention to use a streaming media service is explained by the perceived value of the streaming media service. The F test has a p-value of .000, confirming there is a relationship between the independent and dependent variable. The unstandardized regression coefficient (b) has a value of .684, which is a positive effect on the intention to use a streaming media service, thus H1b is also accepted.

When entering the separate components of TAM (PeouSMS, PuSMS, and PeSMS) the adjusted R2 explains 28% of the variance in the dependent variable ‘intention to use a streaming media service.’ Perceived enjoyment has a positive significant contribution to the intention to use a legal service (p <.05).

Table 4.6.2a Relationship between Intentions to use online music stores and IV(s)

(36)

36

4.6.2 Dependent variable: Willingness to pay for a legal option (OMS/SMS)

The dependent variable willingness to pay is determined by a large set of independent variables. The willingness to pay is determined by ‘the perceived value of the legal option’, five motivations for illegal file-sharing, and the strictness of intellectual property law.

Testing H2, H3, and H4a WTP OMS WTP SMS H2a/b PVOMS/ PVSMS .358*** .566*** H3 No harm -.227* -.284** Unfair division -.019 .043 Favouritism -.074 .137 Prosecution risk .185 -.097 Unfairness of Price -.372** -.048

H4a Strictness of Law -.088 -.259

Adjusted R2 .392 .314

Table 4.6.3 Relationship between the willingness to pay and the IV’s *= p <.05

**= p < .01 *** = p <.001

The dependent variable willingness to pay for an online music store is for 39% explained by the various independent variables. The F-test results in a p-value of .000. The perceived value of an online music store has a positive significant contribution to the willingness to pay (p <.001), thus accepting H2a. Moreover, the IV ‘no harm’ measured by the statement that ‘downloading a few songs or albums for personal use does not harm record labels or artists’ has a negative significant effect on the willingness to pay (p <.05), accepting H3a. Lastly, unfairness of price measured by the two items ‘I consider a paid download (individual song or whole album) too expensive considering the cost savings in distribution, casing, etc.’ and ‘I believe that the price of a album is too expensive for what you get in return’ has a negative effect on the willingness to pay for an online music store (p < .01).

Testing H1 Intentions to use SMS 1 Intentions to use SMS 2 1 factor solution PVSMS .684*** 3 factor solution PeouSMS .158 PuSMS .236 PeSMS .290* (Adjusted)R2 .252 .279

Table 4.6.2b Relationship between Intentions to use streaming music services and IV(s) *= p <.05

**= p <.01

(37)

37

The analysis for the willingness to pay for streaming music services shows similar results. The dependent variable is for 31% explained by the various independent variables. Here the perceived value of a streaming media service has a positive influence on willingness to pay (p <.001). Therefore, H2b is accepted. Moreover, the IV no harm has a significant negative effect on the willingness to pay for streaming music services (p<.01), thereby accepting H3a. As table 4.6.3 displays none of the other hypotheses showed a significant effect in the case of a streaming music service.

4.6.3 Dependent variable: Intention to illegally share files

The intention to illegally share files is determined by three independent variables, namely the strictness of intellectual property law, the intention to use a legal service, and the willingness to pay for a legal service.

The first analysis including the strictness of intellectual property law, intention to use an online music store, and willingness to pay for music at an online music store explains 10% of the variance in the intention to illegally share files, with none of the variables having a significant effect at an alpha level of 5%. And thus, H4b, H5a, and H6a are rejected. However, when slightly loosening the assumptions to an alpha level of 10% the online music stores show a significant effect.

The second analysis including the strictness of intellectual property law, intention to use a streaming music service, and willingness to pay for a streaming music service explains 23% of the variance in the dependent variable. Strictness of intellectual property law and willingness to pay for a streaming music service shows no significant effect on the intentions to illegally share files, and thus H4b and H6b are rejected. The intentions towards the legal use of a streaming media service has a positive significant effect (p<.001), thereby accepting H5b.

Table 4.6.4 Relationship between the intentions to illegally share files and the IV’s *= p <.05 **= p < .01 *** = p <.001 Testing H4b, H5, and H6 Intentions to illegally share files Intentions to illegally share files H4b Strictness of intellectual property law .016 .167

H5a The intention to use

OMS

-.185

H5b The intention to use

SMS

-.389***

H6a Willingness to pay for

OMS

-.235

H6b Willingness to pay for

SMS

-.129

(38)

38

4.7 Sobel test: indirect effect test

In this section the interactive Sobel test is used to test whether the indirect effect of the independent variable on the dependent variable via the mediator is a significant effect. Tables 4.7.1 till 4.7.3 show that the only significant indirect effect is for the perceived value for the legal services on the intentions to illegally share files with the intentions to use a legal service as a mediating variable.

P-value

Online Music Stores .02

Streaming Music Services .00

Table 4.7.1 Sobel tests with intention to use legal service as a mediator P-value

Table 4.7.2 Sobel tests with willingness to pay as a mediator for perceived value P-value

New legal music services and their reduction on consumers’ intentions to pirate