Risks and Rewards of Unlawful File Sharing

Presented by Dr Piers Fleming
Unlawful file sharing (‘piracy’) remains a signi cant global issue with potentially severe punishments. However, increasing the perception of legal risk has had limited success in reducing unlawful content sharing by the public. This research examines to what extent file sharing is motivated by perceived benefits rather than perceived legal risks. It also examines the relationship between perceived risk and benefits as well as potential moderators including: trust in industry and legal regulators, and perceived online anonymity. A large two-part survey of consumers of music (n = 658) and eBooks (n = 737) was carried out online. Perceptions of benefit, but not legal risk, predict stated file sharing behaviour. As perceived benefit increases, perceived risk falls - indicative of an affect heuristic. This relationship increased under high regulator and industry trust (which actually increases perceived risk in this study) and low anonymity (which also increases perceived risk). Given the limited impact of perceived risk for this online behaviour more attention should be paid to perceived bene t of risky and safer alternatives.

Transcript

1. Risks and Rewards of Unlawful File Sharing Piers Fleming Funding from AHRC Grant Number AH/K000179/1, and from the University of East Anglia are gratefully acknowledged RCUK Centre for Copyright and New Business Models in the Creative Economy
2. Unlawful File Sharing • The Recording Industry Association of America • “had the music industry sat idly by and refused to enforce its rights against the sites that profited from facilitating theft, would there even BE a legal music marketplace today?” • US copyright law • making unauthorised copies of music recordings can lead to a civil law suit or criminal charges : 5 years in gaol and fines up to $250,000 • Approximately 50% of the US adult population share digital media unlawfully • It is estimated that this costs the music industry $12.5 billion/year* *These are highly disputed figures
3. Replace/ Generate • UFS may replace sales, e.g. anecdotally the emergence of Napster coincided with a decline in music sales (Liebowitz, 2006) • Most evidence suggests potential sales of music and movies are decreased by UFS. • However, UFS may generate sales by promoting interest. • There is strong evidence that for individuals more UFS is associated with more actual sales. Unlawful file sharing Sales 40-60% of evidence > 60% of evidence -ve association Interest Legal Sales vs Unlawful File Sharing Watson, Zizzo & Fleming (2015)
4. Replace/ Generate Risk / Benefit • Stronger laws appear to reduce unlawful file sharing, but effect may be temporary and limited behavioral data confirming causal legal effect • Many people are unaware of what is lawful (Ofcom, 2011) or it doesn’t feel like a crime (BMRB Social Research, 2009). • High prices appear to reduce sales, lower willingness to pay for content is associated with UFS Unlawful file sharing Sales 40-60% of evidence > 60% of evidence -ve association Lower WTP Less strict laws Lower Prices Less Perceived Deterrent Watson, Zizzo & Fleming (2015)
5. Risk, Benefit and Affect • Films, music and eBooks are affectively enjoyed • We know that judgements of risk and benefit can be psychologically linked for an affective judgement (Alhakami & Slovic, 1994; Finucane et al., 2000) • This could explain greater UFS with greater purchasing Risk Perception Benefit (Value/Utility) (-ve) (-ve/+ve) (+ve/-ve) Affect
6. Trust, Media Type and Anonymity • Trust • Increased trust typically associated with decreased risk perception (e.g. Flynn et al, 1992) • Increased trust is also associated with increased affect (Siegrist et al, 2007) • Anonymity • Expect reduced risk perception • Expect more disengaged, less affective if anonymous • Media type • Evidence suggests that all media types are not psychologically equivalent (Watson, Zizzo & Fleming, 2015) • Perceived risk may be greater for music – industry focus
7. Panel Study eBooks (N = 1036) T2 (N= 737) 0 = 644, 1-3 = 57, 3+ = 36 Music (N = 959) T2 (N= 658) 0 = 540, 1-3 = 43, 3+ = 75 • Perceived Risk • Perceived Benefit • Trust in Industry • Trust in Regulators • Perceived Anonymity • Behaviour : How many downloaded in the past 2 months, what % lawful? 2 months Unlawful downloads
8. Method • Risk 6-items (likelihood & severity): e.g. If I was caught downloading eBooks/music unlawfully I think I would face a harsh punishment (Cronbach’s αMUSIC = .72, Cronbach’s αEBOOKS = .77). • Benefit 7-item e.g. I think getting books/music for free is a good reason to download eBooks/music files unlawfully (Cronbach’s αMUSIC = .80, Cronbach’s αEBOOKS = .76). • Trust Industry/Legal regulators 8-items (fairness, openness, care, competence) e.g. The book publishing companies/Legal regulators actions, with regard to unlawful downloading, are intended to help the public (Legal regulators - Cronbach’s αMUSIC = .77, Cronbach’s αEBOOKS = .72; Industry - Cronbach’s αMUSIC = .71, Cronbach’s αEBOOKS = .69). • Anonymity 5-item e.g. When you are on the internet you feel free to act in ways you normally would not (Cronbach’s αMUSIC = .62, Cronbach’s αEBOOKS = .61). • Unlawful file sharing : How many eBooks/music files have you downloaded in the past two months (of all kinds)?” (i.e., since Part 1), and “What percentage of those eBooks/music files were lawful?”
9. Results T1 T2 Risk Benefit Trust (industry) Trust (regulators) Anonymity UFS eBooks 23.6 21.5 33.5 33.8 15.4 14.6% N=737 Music files 24.1 22.6 31.8 32.8 15.4 21.9% N=658 Sig dif ** *** ** *** Greater trust for eBook industry and regulators Greater perceived benefit and UFS for music files
10. Ordinal logit regression to predict UFS Media Variable OR Lower 95% CI Upper 95% CI Wald χ2 (1df) p eBooks Risk 1.01 .97 1.05 .19 .666 Benefits 1.07 1.04 1.11 20.43 <.001* Music Risk 1.00 .96 1.04 .002 .965 Benefits 1.15 1.11 1.18 82.31 <.001* Benefit, and not risk, predicts unlawful file sharing
11. Risk and Benefit (affect heuristic) • Risk and benefit ratings correlate negatively • For music (r = -.153, p < .001) • For eBooks (r = -.202, p <.001).
12. eBooks Music Industry trust Regulator trust Anonymity 20 21 22 23 24 25 26 27 Low benefit High benefit Perceived risk Low trust High trust 20 21 22 23 24 25 26 27 Low benefit High benefit Perceived risk Low trust High trust 20 21 22 23 24 25 26 27 Low benefit High benefit Perceived risk Low trust High trust 20 21 22 23 24 25 26 27 Low benefits High benefits Perceived risk Low trust High trust 20 21 22 23 24 25 26 27 Low benefit High benefit Perceived risk Low anonymity High anonymity 20 21 22 23 24 25 26 27 Low benefit High benefit Perceived risk Low anonymity • The risk – benefit association was only found amongst high trust (industry/regulator) individuals for Music • This indicates greater affect-heuristic use with high trust
13. eBooks Music Industry trust Regulator trust Anonymity 20 21 22 23 24 25 26 27 Low benefit High benefit Perceived risk Low trust High trust 20 21 22 23 24 25 26 27 Low benefit High benefit Perceived risk Low trust High trust 20 21 22 23 24 25 26 27 Low benefit High benefit Perceived risk Low trust High trust 20 21 22 23 24 25 26 27 Low benefits High benefits Perceived risk Low trust High trust 20 21 22 23 24 25 26 27 Low benefit High benefit Perceived risk Low anonymity High anonymity 20 21 22 23 24 25 26 27 Low benefit High benefit Perceived risk Low anonymity • The risk – benefit association was only found amongst low anonymity individuals • This indicates greater affect-heuristic use with low anonymity
14. Conclusions • Perceived UFS benefits are more salient/important than perceived UFS risks • Legal impact may be low due to public focus on benefit • High trust increases perceived UFS risk • Low anonymity increases perceived UFS risk • Risk-Benefit correlations indicative of affect are primarily evident with high trust/low anonymity • Increasing trust may encourage affective judgment • Reducing anonymity may increase affective judgment • Music and eBooks are not equivalent: • Music has less trust and greater benefit • Music is more sensitive to trust
15. Thanks to: • Steven Watson and Daniel Zizzo