What Is Reasonable?
Sharing copyright-protected media with family and friends is widely accepted in both the US and Germany.
In the US 80% of those who have music files say it is “reasonable” to share them with family members, and 60% approve of sharing with friends. Concerning movie and TV files 73% view sharing them with family as reasonable, and 59% approve of sharing them with friends.8 In Germany 76% view sharing music files with family as reasonable, and 55% approve of sharing with friends. For TV and movies the numbers are 80% and 63%. Germany shows a sharper differentiation in attitudes by age and broader acceptance of sharing TV and movies.
In contrast, making unauthorized copies available through online networks receives low levels of support in both countries.
Our results suggest that a loose distinction between private and public sharing continues to organize public opinion on these issues: the former is reasonable, the latter is not. In this respect, public opinion mirrors and informs legal traditions that have historically distinguished private, non-commercial contexts for infringement from public ones. The results also point to the breakdown of this consensus as young adults redefine the scope and scale of personal networks. In the Facebook era, the concept of ‘friends’ has become very elastic, and is increasingly mediated by the Internet. We will see the difference this makes when we come to Penalties.
8. Our preliminary Research Note mislabeled these findings. It presented data based on all Americans, rather than the narrower category of those who have music or video files.
US Results
Few Americans with music files think it is “reasonable” to upload copies to websites where anyone can download them (15%), post links to “pirated copies” on websites such as Facebook (10%), or sell unauthorized copies (4%). Overall numbers are similar for TV/movies but differ in the age distribution. Significantly more younger people (18–29) view sharing, uploading, and linking to TV/movie files as reasonable—especially uploading (24%) and linking to unauthorized videos (22%). Such results are not surprising; we live in an audiovisual culture that is not only shared, in the passive sense signaled by the term “audience,” but also increasingly and actively sharing.9 In all cases, approval of these practices sharply exceeds the percentage of respondents actually engaging in them.
Music Files: It Is Reasonable To… (US, Among The 50% Of Adults Who Have Music Files)
Movie/TV Files: It Is Reasonable To… (US, Among The 14% Of Adults Who Have Movie/TV Files)
The views of P2P users track closely with those of the wider under-30 group—including the relative lack of tolerance for uploading. (P2P software typically makes available to others any files that are being downloaded.) This ambivalence is striking and confirmed at several points in our findings. Clearly many P2P users have some sympathy for arguments against file sharing outside personal circles—at least when these arguments are made explicit. Whether these are ethical concerns about “theft” or fairness or broader uncertainty about the systemic effects of large-scale file sharing is unclear. It is clear that, in P2P communities, such concerns are subordinated to other factors such as cost, availability, and convenience.
What about the 50% of Americans who don’t have music files? These are predominantly older Americans and non-Internet users. Surprisingly, these groups show very little divergence in attitudes. Seventy percent think it is reasonable to share with family, 52% to share with friends, 18% to upload, 6% to post links to pirated copies, and 9% to sell unauthorized copies.
The trend is very similar for movie and TV files. We conclude that the important shift in attitudes is among young adults—the natives in a culture of digital sharing. The views of older groups, in contrast, are relatively stable, independent of participation in digital media culture. This stability likely reflects the persistence of an older set of norms grounded in practices of sharing through face-to-face relationships—via books, discs, and tapes—but not through online networks.
- Australia, 2012: According to the film and TV industry association IPAF (2012), 49% of “persistent illegal downloaders” and 54% of “casual illegal downloaders” agreed with the statement that “movie/TV piracy” is “stealing/theft.” When asked Internet whether they contribute to “the problem of TV/movie piracy,” 74% of persistent illegal downloaders agreed with the statement that “It’s not something I give a lot of thought to” (vs. 65% of other groups).
9. Both US and German surveys generated unexpectedly high rates of approval among the oldest cohorts on the issue of sharing with family and friends. In the US >65 group, high approval for “share with friends” appears to reflect a conflation of sharing with collective time-delayed viewing of TV via a TIVO or set-top box. We do not think this conflation figures significantly in the other age groups. This possible explanation does not apply to the German results, where “vertretbar, Kopien Ihrer Dateien an Freunde weiterzugeben” unambiguously means “passing along copies.”
German Results
Results among German digital media owners are broadly similar. An expansive view of what is reasonable correlates strongly with youth—especially on the issue of sharing with friends. Among 18- to 29-year-olds, 73% view sharing music files with friends as reasonable, compared to 55% in the general population. For TV/movies 79% of young adults and 63% of the general population view sharing as reasonable.
Music Files: It Is Reasonable To… (Germany, Among The 42% Of Adults Who Have Music Files)
Music Files: It Is Reasonable To… (Germany, Among The 15% Of Adults Who Have Music Files)
In Germany broad views of what is reasonable are also common among those 70 and over. Here our results are qualified by small sample sizes: only 7% had music or video files. But our confidence is increased by the extent to which this tolerance is echoed in other results that have larger samples, notably on questions of enforcement. Compared to the two younger cohorts (50–64 and 30–49), the German postwar generation is more disposed toward sharing and more distrustful of enforcement and other exercises of state power.
There is, accordingly, an Internet politics of the old as well as the young in Germany, with support for a culture of IP enforcement anchored by the middle aged.
- Germany, 2011: Some surveys ask, instead, about perceptions of legality—a question that measures understanding of copyright law rather than acceptance of practices. According to GfK 15% of Germans think downloading from P2P networks is legal, and 8% think making media content accessible on social networks is legal (GfK 2011).
- Mexico, 2012: A Parametría survey (2012) asked about “sharing various content on the Internet without the authors’ authorization.” Thirty-eight percent said that such sharing was a “crime,” 34% said it was part of one’s “right to obtain information,” and 28% did not know.
- Poland, 2011: A Centrum Cyfrowe study asked “active Internet users” for their views on the statement that “using pirated movies and music is theft.” 28% agreed; 47% disagreed.
Penalties
Only a slim majority of Americans (52%) say “people should face punishment if they download an unauthorized copy of a song or movie from a website or file-sharing service.” Thirty-four percent oppose penalties altogether, 7% say it depends on the circumstances, and 7% do not know or did not answer.
Among American 18-29 year olds, only 37% support penalties. 53% oppose them.
German support for penalties is somewhat higher: 59% support penalties, 26% oppose, and 9% would consider the circumstances. Here, the youth factor is much less pronounced. Among 18-29 year olds, 56% support penalties.
Should Downloading An Unauthorized Song Or Movie Be Punishable? (US, Among All Adults)
Should Downloading An Unauthorized Song Or Movie Be Punishable? (Germany, Among All Adults)
Among both Germans and Americans, support for penalties is limited to warnings and fines. In the US, 51% percent of adults support warnings or fines for unauthorized downloading.
Twenty-eight percent support limits on the speed or functionality of the Internet (e.g., “bandwidth throttling” or restricting access to certain sites).
Only sixteen percent support disconnection from the Internet as a penalty (72% oppose). Among those supporters, most (58%) would drop their support if it meant disconnecting households rather than individuals—which it does.
Interestingly, 41% of US P2P users support penalties for unauthorized downloading—just 11% below the average and marginally higher than among young adults. Among German P2P users, support for penalties comes in well below the average at 38%. Clearly, the practice is a matter of some ambivalence within the file-sharing community—something we see, too, in the shift toward legal streaming solutions and the stated willingness of P2P users to pay for cheap, convenient access. See Legalizing File Sharing.
Among Germans, only warnings receive majority support (59%). Half of the population supports fines, 28% support limits on the speed or functionality of the Internet, and only 22% support disconnection from the Internet as a penalty. (Due to time constraints, we did not ask the question about household-level disconnection.)
Among young Germans, Internet disconnection is very unpopular, even among those who favor penalties. Only 24% of penalty supporters under 30 support disconnection. Overall, 14% of those under 30 do).
Opposition to disconnection also rises sharply with income (which in turn correlates with the propensity to buy media). Among penalty supporters who make more than €3000/month, 20% support disconnection; 74% oppose it.
What Is An Appropriate Penalty For Downloading An Unauthorized Song Or Movie? (US, Among All Adults)
What Is An Appropriate Penalty For Downloading An Unauthorized Song Or Movie? (Germany, Among All Adults)
Should Infringers Face Disconnection? (Germany, By Age)
Should Infringers Face Disconnection? (Germany, By Income)
- Canada, 2009: In an Angus Reid survey (2009), 45% of Canadian Internet users described P2P users as “doing what people should be able to do on the Internet.” Only 3% “believe file-sharers are criminals who should be punished by law.”
- UK, 2010: Wiggin’s 2010 Digital Media Survey asked questions about penalties in the wake of passage of the 2009 Digital Economy Act. Among other provisions, the DEA created a framework for “graduated response” in cases of repeat infringement, leading to disconnection from the Internet. Among Wiggin’s respondents, strong penalties for repeat offenders attracted minimal support: bandwidth throttling, 7%; suspended Internet access, 11%; criminal prosecutions and fines, 14%.
- Poland, 2011: A Centrum Cyfrowe study (Filiciak, Hofmokl, and Tarkowski 2012) asked whether “the law should punish people who illegally share movies and music more severely.” Fourteen percent agreed with the statement; 60% disagreed.
- Germany, 2012: GfK asked about penalties for the narrower act of “making available” infringing content online—a term most likely equated with uploading. Seventy-seven percent supported fines; 50% supported bandwidth throttling; 47% supported Internet disconnection “for 1-3 months.”
Fines For Downloading
Among Americans who support fines, 68% support amounts under $100 for the unauthorized downloading of a movie, and 19% support fines of “less than $10.” This contrasts sharply with US copyright law, where the statutory penalty for willful infringement ranges from $750 to $150,000 per act. Fines over $1000 attracted only 9% support. Fines over $100,000 were supported by less than 1% of Americans.
US supporters of fines were also asked about appropriate penalties for unauthorized downloading of a song. In this case, recommendations shifted lower: 31% supported fines of “less than $10,” 44% supported “$11 to $100” fines, and 15% supported “$101 to $1000” ones. Support for higher fines was minimal.
German supporters of fines aimed a bit higher. Only 5% supported fines under €10; 45% supported fines between €11 and €100; 28% supported fines of €101–€1000. Since 2008, German copyright law requires that prior to suing, plaintiffs must first ask defendants to sign a cease-and-desist declaration (§97a UrhG), which typically triggers damages and lawyers’ fees in the range of €300–€1,000 for single titles. Unlike in the US, infringement claims have not been bundled into much larger claims against individuals, although there have been cases of serial actions against individuals who prove willing to pay. The majority of noncommercial infringement cases are settled out of court. German law does not allow for statutory damages, but rather makes determinations of damages based on the cost of an appropriate license. In commercial cases plaintiffs can alternatively demand the profits from the infringing use.
Proposed Fine For Downloading One Movie (US, Among The 51% Who Support Fines)
Proposed Fine For Downloading One Movie (Germany, Among The 50% Who Support Fines)
Will It Work?
There is a lively debate, at present, about the effect of warnings and penalties on online infringement. Several surveys have found file sharers to be at least potentially receptive to warnings: GfK in Germany (2011) found that 45% of file sharers claimed that they would be dissuaded by warnings; HADOPI in France (2011a) put the number at 34%; Wiggin 2010 in the United Kingdom found that warnings would make 34% “take measures to protect themselves,” including (but not limited to) reduced file sharing.
Warnings take many forms. In 2009 France implemented a public system (HADOPI) for issuing warnings and—upon a third notice—initiating legal proceedings alleged infringers. By early 2012 HADOPI had sent 822,000 first warnings and 68,000 second warnings. By June 2012, 314 alleged infringers faced the final strike, where a court can issue fines and/or an Internet ban of up to one year. The first and to date only conviction came in September, against a man who everyone agrees did not personally infringe. The new Socialist government has announced its intention to “redesign” and possibly defund HADOPI, casting serious doubt on the future of the system (Guess 2012; Farivar 2012).
A US version of HADOPI—the Copyright Alert System—is supposed to come into effect in early 2013. This is a private system set up by rights-holder groups (especially the MPAA) and the major ISPs. It envisions five or six strikes leading to bandwidth throttling and other “mitigation” measures but not—according to current statements—disconnection. Under the current guidelines, further penalties can be pursued through the courts. Proponents prefer to describe the alert system as an educational initiative, which is true provided one understands education primarily in terms of fear of punishment.
Germany does not have an official warnings regime, but it has become home to one of the world’s most active “cease-and-desist” industries. In 2010 at least half a million cease-and-desist letters were received by alleged infringing P2P file-sharers, demanding payments of half a billion euros in damages and lawyer’s fees. In mid-2011 the German ISP association ECO reported that ISPs were receiving court orders to hand over the personal data of 300,000 Internet subscriptions per month. A 2012 survey commissioned by the German association of consumer organizations found that 4.3 million Germans, or 6% of the population, have received a letter, demanding damages and lawyers’ fees of on average €800 (VZBV 2012). Under threat of civil litigation, the majority of recipients reportedly sign the agreements and pay the requested amount (Gulli, Abmahnwahn, and Abmahnwahn-Dreipage 2011).
Law-firm fronts like the US Copyright Group and Copyright Enforcement Group have adopted similar strategies in the US (following the precedent set in the early 2000s by the RIAA), but US courts have been increasingly hostile to the practice and have repeatedly dismissed efforts to bundle cases, use the courts to force settlements, pursue mass requests for Internet subscriber data, or—more recently—claim that an IP address is a foolproof means of identifying an individual. Particularly aggressive groups like Rightshaven have been crippled by adverse judgments and countersuits. Nonetheless, the cost-benefit equation continues to favor the firms. By mid-2011 an estimated 200,000 defendants had been named in suits in the US, with many more likely the recipients of prelitigation threat letters. There is no public data on how many have settled. No recent cases against individuals have been brought to trial.
- UK, 2012: Wiggin’s most recent survey (Wiggin 2012) asked respondents about their anticipated behavior as the United Kingdom prepares to introduce three-strikes enforcement measures. Among UK respondents, 27%–28% indicated that they plan to do more downloading of unauthorized music, film, and software in the next 12 months; 16%–18% plan to do less.
Organizations like HADOPI and IFPI have claimed success for surveillance and penalties based on apparent declines in P2P use in the wake of stronger enforcement. But, assuming these declines are real, they can reflect two factors: an overall decline in file sharing due to fear of enforcement or a shift toward less observable means, such as direct downloads or streaming. One way to explore this difference is to survey users on their downloading activities over time. Another is to look at what happens to markets for the affected goods in the wake of changes in enforcement.
In Germany GfK has sufficient long-term data for the first approach. GfK surveys have found a significant absolute decline in illegal music downloading in all forms since 2003 and a parallel (though in absolute terms, slight) rise in legal sales. These data points have been used by a variety of parties to justify the cease-and-desist efforts. For our part, we find it entirely plausible that 4.3 million legal threats have had a dissuasive effect. According to GfK, 25% of 20-29 year olds claim to know someone who has been “legally affected” (GfK 2012) by these letters. That is an astonishingly wide-ranging culture of enforcement.
But even in the context of this massive campaign, it is unclear how much of the decline can be attributed to enforcement. Much of the drop preceded the expansion of the cease-and-desist industry, with the largest year-to-year decline occurring in 2004. Free streaming services grew throughout the period and now represent the primary means of listening to music for 9% of those under 30. Legal digital sales are indeed up in Germany, but from a very low baseline and—as IFPI has emphasized in other contexts—less than in neighboring countries like France. The CD remains far more popular than in peer countries—but so do physical books, which have not been part of the piracy/enforcement dynamic. Because GfK does not share detailed findings or data, its results cannot be examined in any detail.
The second approach takes us to France. With the French HADOPI three-strikes model moving toward its final penalty phase, the first relevant market-based analyses are beginning to emerge. Based on iTunes sales data in Europe, Danaher et al. (2012) describe a positive and (so far) persistent “HADOPI effect” on iTunes sales as the warnings regime began to be publicly discussed in 2009, resulting in a 22.5% increase for songs and a 25% increase in album sales (versus a control group of European countries). Again, claims of causality have a prima facie logic, but, as in Germany, we think more work needs to be done to isolate it from other changes in digital markets and enforcement practices. As we noted above, we are skeptical of frameworks that assume direct substitution between pirated and purchased products. Media markets have become very complex, with many partial substitutes competing for money and attention. Notably, the same period saw the introduction of major competitors to iTunes (e.g., Spotify) in most of the control countries, but not France—a difference that could contribute to France’s higher relative iTunes growth rates. Nor is it clear that HADOPI’s warnings regime produced an enforcement climate significantly different from those of other European countries, such as Germany, the United Kingdom, Sweden, and Spain, which have all seen major increases in public and private enforcement activity in recent years—including a similar 2010 announcement in the United Kingdom of impending three-strikes measures (as yet unimplemented). In terms of actual punishments, France is well behind the European curve.
Our study does little to illuminate these questions directly, but it does provide evidence of the diversity of the file-sharing ecology, marked by direct downloading, streaming, private networks, and offline copying. Given the low costs of transition between these methods, we would be surprised if P2P-specific measures such as HADOPI have more than a marginal impact on either file sharing or sales over time. And of course the major copyright stakeholders know this. As HADOPI officials have made clear, P2P surveillance is not an end point in the enforcement effort but a step on the way toward more comprehensive measures (Roettgers 2010).
As we shall see below, many Germans (and some Americans) use online privacy technologies that thwart current surveillance measures. One impact of warnings is to push file sharers toward anonymization and encryption technologies, setting up a direct conflict between future enforcement measures and privacy. Larson and Svensson (2010) have found ample evidence of this shift in Sweden, where strong enforcement measures were introduced in 2009.
Who Judges?
Who Should Adjudicate Charges Of Copyright Infringement? (US & Germany, Among All Adults)
Charges of copyright infringement are typically handled in three ways: (1) through the sending of takedown notices to services hosting content uploaded by their users, (2) by sending cease-and-desist letters to Internet subscribers from whose IP addresses infringing content was made available, and (3) failing those measures, through civil litigation. Only a handful of cases in the US and Germany have been prosecuted under criminal law, which has generally been reserved for infringement on a “commercial scale.” The most recent case in Germany targeted the movie streaming and download site kino.to, and led to sentences from two-and-a-half years to four-and-a-half years against the group’s operators. The site’s founder will also have to hand over €3.7 million allegedly earned from advertising on the site.
Germans have clear views about what constitutes due process in copyright infringement cases. Sixty-seven percent indicated that the courts are the proper authority in determining the truth of charges of infringement.
American views are somewhat softer. Fifty-four percent say that a court should make the determination of whether a person has committed copyright infringement. Only 18% percent say the music companies or movie studios should make that decision; 15% percent propose that ISPs play that role. (The rest responded that “it depends on the situation” or did not answer.)
The recent six-strikes agreement between US music and film companies and ISPs to warn and then punish infringers does not provide for judicial review.
Should Intermediaries Block Infringing Material?
To date, we have lived with a mostly ex post enforcement model, in which rights holders react to online infringement by filing complaints with site or service owners. This was the approach established by the 1998 Digital Millennium Copyright Act (DMCA) in the US and the 2000 Electronic Commerce Directive in the EU, which specified the responsibilities of site owners in regard to infringement claims and provided “safe harbor” from liability if they met them—chiefly, responsiveness to “takedown notices” for infringing content. The boundaries of the safe harbor have been a subject of debate ever since, with US courts generally asserting a broad understanding of the concept, while European courts have produced more diverse interpretations that have significantly clouded its meaning (Edwards 2011).
The goal of much recent enforcement legislation is to shrink the safe harbor. US and European content industries are trying to move Internet enforcement from an ex post model to an ex ante one, in which ISPs, search engines, social networking companies, and other intermediaries assume responsibility for screening and blocking infringing content on their networks—or face liability. This shift was a major stake of the 2011 Stop Online Piracy Act (SOPA).
What do people think about these issues? The truth is that they have rarely been asked. Because these proposals raise a mix of relatively new political and technical concerns, public opinion is, we think, still very much in formation on these issues—susceptible to the way in which questions are framed. We tested a number of functionally equivalent terms for ‘blocking’ infringing materials, including “screening” and “censoring.” We also asked whether people would support the monitoring of their Internet use to prevent infringement, and we explored their comfort levels with the different types of organizations that might play these roles.
Because of the relatively high degree of Internet literacy required to address this set of questions, we posed them to the largest relevant sample: home Internet users in some cases (US 72% of the population; DE 75%) and all Internet users in others (US 79%; DE 77%). Tracking on other questions suggests that the results are valid within our margin of error for the wider population.
What we see in the results is an emergent ordering of values online: a progression from simple judgments about whether people support copyright or law enforcement (they do) to efforts to rank values when they conflict. As the Internet becomes the platform for a wide range of forms of human activity and expression, it becomes home to more of these conflicts. It becomes necessary to make decisions about tradeoffs—about which values matter more.
In the US, majorities value both copyright and enforcement in general but will prioritize privacy, freedom of speech, and fear of government intrusion when there are conflicts. The SOPA debate was, arguably, the first political demonstration of this preference order.
In Germany our data show much higher support for online enforcement across the board—up to and including explicit government censorship of sites (52% support). But they also show a sharp reversal of support when such measures threaten privacy.
US Results
Blocking Copyright Infringement (US, Among Internet Users)
Predictably, the softest language evokes the highest support for enforcement. On the far left: Should websites like Facebook, Dropbox, Rapidshare, and others that allow links or files to be uploaded by users “try to screen all material and try to reject pirated music and video”? The question downplays conflict, costs, and difficulties. Sixty-one percent said yes; 32% no.
Support for enforcement drops slightly when the topic shifts to blocking and the language becomes more coercive: Should ISPs be “required to block access to sites that provide access to pirated songs and videos”? A 58% majority responded yes; 36% said no.
What about search engines? Should they “be required to block links to pirated music and videos online”? Fifty-three percent say yes; 42% say no.
Support for search engine and ISP blocking tracks with age: only 39% of 18- to 29-year-olds support requirements that search engines block sites, while 59% oppose them. Americans 65 and older, in contrast, support blocking by 59% to 31%. Other enforcement questions are less age sensitive. One possible explanation is that the young are less tolerant of blocking by services viewed as primary gateways to the Internet (such as ISPs) than of services perceived to be secondary or replaceable (Facebook).
When asked if the government should block access to sites that infringe, majority support for enforcement vanishes: 40% say yes; 56% no. When we replace the word “block” with the stronger “censor,” support drops further: 33% say yes; 64% say no. Americans do not like government involvement in these matters.
Should ISPs “censor” infringing content rather than “block”? Support drops 12%: 46% say yes; 49% say no.
What if efforts to block infringing files and links to infringing content also result in the blocking of some legal content (as has been the case with all large-scale efforts to blacklist sites or filter content to date)? In this case, support for blocking infringing materials drops sharply: 57% oppose blocking; 36% support it.
Finally, should corporations or the government monitor people’s Internet use “in order to prevent copyright infringement”? This is a privacy question. Sixty-nine percent said no; 27% responded yes or sometimes; 3% did not know.
What can we conclude from these results?
Solid majorities of American Internet users oppose copyright enforcement when it is perceived to intrude on personal rights and freedoms. Sixty-nine percent oppose the monitoring of their Internet activity for enforcement purposes; 57% oppose blocking or filtering if those measures also block some legal content or activity. In our view, these are the value choices that now define the enforcement debate in the US.
- UK, 2010: Wiggin’s 2010 survey asked if the government should “force the ISPs to police the Internet and block all illegal activity.” Only 12% said yes. Should the government “block access to websites that promote illegal activity”? Twenty-five percent said yes.
German Results
Blocking Copyright Infringement (Germany, Among Internet Users)
German support for third-party blocking of infringing content runs much higher than in the US across the board—until such actions infringe on privacy. Our data suggest that privacy is the paramount online value for Germans, capable of overriding much stronger support for the enforcement of copyright laws. And copyright is not the only domain in which privacy takes priority. Lawsuits over Google Street View and opposition to social plug-ins on Facebook and other social media services also reflect a different ordering of values than in the US and a different view of their applicability online. In both German and EU law, companies have less latitude to collect data and develop profiles of individuals without explicit consent than in the US. Our data suggest that large majorities support these data protection measures. Surveillance of citizens’ online behavior—the clear direction of current enforcement proposals—faces a steep uphill battle.
Age Effects
Blocking Copyright Infringement, By Age (US & Germany, Among Internet Users)
What does the future of these issues look like? On most questions support for blocking correlates roughly with age. Among young adults (18–29) in the US, only the Facebook question, using the softest language, generates majority support for blocking (57%). Here the gap between young adults and the top group (50–64) is 11 percentage points.
Stronger measures are unpopular among younger respondents. This age effect is clearest in regard to “requirements” that ISPs and search engines block access to pirated materials. In the US both questions show large gaps in support between younger and older respondents (28 and 25 points, respectively). These age effects are similar in Germany, although support there has a higher floor. In Germany the gaps between young adults and the top age group (50–64) on ISP and search blocking are 28 and 20 points, respectively.
At the same time, 18- to 29-year-olds in both countries show slightly less suspicion of government in the enforcement role, slightly less concern for privacy, and less concern for due process through the courts. Call this the “Facebook effect,” reflecting greater comfort with lives published online through corporate intermediaries, and with the relatively invisible role of the state in this sphere.
Age effects are also visible at the other end of the spectrum, where enthusiasm for blocking falls off slightly among Americans older than 60. This result is consistent across several questions, but we have no particular explanation for it.
Among Germans, the oldest cohort (>70) shows sharper differences on questions related to enforcement and the role of the state. Compared to younger cohorts, support for private-sector blocking falls across the board among this group—in some cases sharply: 15% off the peak for Facebook and 16% for search engines. This postwar generation of Germans also consistently shows the lowest levels of trust in the state. Compared to the peak results, support for government action on enforcement falls 28% for “censorship” and 14% for “blocking.” Concern for privacy is highest in this group (14% above the bottom), and faith in the courts as a legitimate arbiter is the lowest (18% off peak).
Income Effects
Our income analysis divides the population approximately into thirds, comprising low-, middle-, and high-income households. Attitudes toward online enforcement do show some rough correlation with these categories, though not to a degree that would allow for strong conclusions.
In Germany, lower-income Internet users tend to see the government as responsible for enforcement while high earners tend to place this responsibility on ISPs and search engines. Higher-income Germans also offer the strongest support for privacy and the role of courts in adjudicating infringement charges.
US respondents show less differentiation on private-sector blocking but similar distributions on the other issues. High earners provide the strongest support for due process, privacy, and speech rights and the lowest support for government involvement. With allowances for some noisy data, low earners in the US show moderately more support for strong online enforcement in general.
Blocking By Politics
Blocking Copyright Infringement, By Politics (US & Germany, Among Internet Users)
Unlike age and income, political identification in the US plays only a small role in explaining differences in views on enforcement.
Self-identified Republicans are slightly more supportive of online enforcement measures than Democrats. The largest gap is 12%, on the question about ISP blocking. This result is almost certainly overdetermined by age differences: the young dislike ISP blocking and lean sharply Democratic. Among 18 to 29 year-olds, Obama beat Romney by 24%. Conversely, Republicans are slightly more likely than Democrats to rank privacy over enforcement.
Among independents, support for enforcement is generally slightly lower than for either party. The data strongly suggest that piracy and enforcement are first and foremost generational issues and secondarily income issues—not political issues.
In Germany political identification plays a larger role in differentiating attitudes toward enforcement, with conservatives generally showing more support for enforcement than those in the center or left.
Here respondents were asked to place themselves on a political spectrum of 0 to 6, with 0 representing the Far Left and 6 the Far Right. In Germany’s multiparty political system, this is a common method for representing political ideology. In our survey 20% of respondents identified themselves as “Left”—as 0, 1, or 2 along the numerical spectrum; 43% selected 3 on the scale; 23% chose 4, 5, or 6—the political Right. Fourteen percent were unsure or refused to choose.
This distribution does not map neatly onto party identification. In 2011 the center-right parties—the allied Christian Democrats (CDU) and Christian Social Union (CSU)—together polled at 35% at the time of our survey (Infratest 2011). The traditional centrist party, the Free Democrats (FDP), polled at 4%. The center-left parties—the Social Democrats (SPD) and the Greens—polled at 28% and 20% respectively. To their left, the party known as ‘the Left‘ polled at 7%.
For our survey this implies that the block of center-identified respondents vote across a range of center-left and center-right parties. Because centrists are also the largest block by a wide margin, centrist opinion plays a disproportionate role in defining public opinion more generally on these issues. This is the case in the US as well, where 37% of respondents described themselves as “independents.”
Views on blocking and filtering are relatively consistent across center- and left-identified respondents, with differences generally falling within or near the margin of error. Conservative support for such measures runs higher, typically by around 10%. This is clearest in regard to the “safe harbor” questions about ISP and search engine responsibilities: strong safe harbor positions have become part of the platforms of the Center and Left parties. In contrast, there is no significant divergence between Left, Center, and Right on private-company screening of user activity.
The alignment between center- and left-identified respondents breaks down on the Internet monitoring question, with left-identified respondents showing considerably more tolerance for monitoring than centrists (and slightly more than conservatives). Traditional centrist concern with privacy appears to carry considerable weight on this issue, tracking the longer-term priorities of parties like the FDP. The issue of data retention by ISPs—necessary to any comprehensive policing of Internet use—has become particularly contentious. The current minister of justice, Sabine Leutheusser-Schnarrenberger (FDP), has refused to implement the EU Data Retention Directive (2006), resulting in legal action against Germany by the European Court of Law.
Privacy & Countermeasures
A major goal of SOPA and similar enforcement measures is to make Internet intermediaries like ISPs and social networking companies bear responsibility for infringement on their networks or services. The content industry vision of strong ex ante enforcement—prior filtering of infringing content—will require extensive monitoring of user activity. Content providers and industry groups already monitor public P2P networks and can identify the titles of files transferred and the IP addresses of participants in the exchange.
A variety of other actors, both public and private, monitor Internet activity and communications. And a variety of tools has been developed to shield such activity from surveillance, including Virtual Private Networks and routing systems that obscure IP addresses (such as TOR). The debate about copyright enforcement leads directly into this larger debate about surveillance, privacy, and anonymity.
Do You Think Your Internet Use Is Regularly Monitored By… (US & Germany, Among Internet Users)
Germans are slightly less likely than Americans to believe that their Internet use is monitored by government or private companies. Yet they are more than twice as likely to take measures to protect their online privacy: 41% make special efforts to encrypt their Internet traffic, and 11% use tools to hide their IP addresses online (such as TOR or a VPN). That is more than twice the percentage of Americans who encrypt traffic (19%) or hide their IP addresses (5%). It is also significantly higher than the 4% of Germans who admit to using P2P services, suggesting a commitment to privacy that extends well beyond fear of copyright enforcement.
But for obvious reasons, file sharers are particularly sensitive to this issue and have become early adopters of anonymizing tools. In the US 16% of P2P file sharers take measures to hide their IP addresses (under 2% of the general population). In Germany 36% hide their IP addresses (1% of the population).
It seems entirely likely that the use of anonymizing tools will grow quickly in response to stronger surveillance and enforcement. Our study provides no data on this trajectory and there has been very little investigation of the subject in general. An exception is the work of Larssen and Svensson in Sweden (2010; 2012), which has tracked VPN use since the passage of strict new enforcement legislation in 2009. In 2008 VPN use stood at 9% of Internet users and 12% of online file sharers. By 2012 it had grown to 15% and 24%, respectively.
Privacy Countermeasures (US, Among Internet Users)
Privacy Countermeasures (Germany, Among Internet Users)
Legalizing File Sharing
Penalization, censorship, and surveillance are not the only strategies on the table. There have also been proposals to legalize file sharing in exchange for a collectively managed levy paid to authors and artists. Although these proposals go by a variety of names,10 we will refer to them as “sharing licenses.”
Since the emergence of file sharing as a mass practice with the launch of Napster in 1999, sharing license proposals have been advanced by legal scholars, music industry stakeholders, consumer organizations, collecting societies, Internet user associations, and several European political parties (Grassmuck 2012). These proposals differ in detail but all call for legal permission to upload and download copyright-protected works accompanied by a monthly levy on broadband Internet access to recompense rights holders. Our survey explored support for such a model and asked how much people would be willing to pay for it.
“Some people have suggested that songs, TV shows and movies could be legally uploaded and downloaded as long as every broadband user in the country were willing to pay a relatively small monthly fee or levy to compensate authors and artists. Would you be in favor or would you oppose paying such a monthly fee for access to songs, TV shows and movies?”
Would You Pay A Small Broadband Fee To Compensate Creators If File Sharing Were Legalized? (US and Germany)
Forty-eight percent of Americans and 61% of Germans indicate support for the sharing license model. Support is higher among the young (US, 55%; DE, 68%) and among those who download music (US, 55%; DE, 70%) and videos (US, 49%; DE, 79%). It is higher still among P2P users: 60% in the US and 75% in Germany support the idea.
Support for the sharing license greatly exceeds the level of online file sharing reported in our survey. The discrepancy is most pronounced in Germany, where only 4% use P2P protocols and 29% have downloaded songs for free. In the US, where this marks the first time this question has been put to a representative panel, high support is surprising given the lack of history with copyright levies (with the exception of the largely inconsequential Audio Home Recording Act of 1992). On the other hand, sharing licenses were first proposed in the US and have played a recurring role in discussions among ISPs, music industry stakeholders, and public interest groups.
It is worth noting that penalizing and legalizing file sharing attract almost identical levels of support—and significant overlapping support. In both countries over half of those who support penalization also support legalization: 60% in Germany and 55% in the US. Clearly, some of the pro-penalization sentiment reflects support for law enforcement in general rather than opposition to file sharing in particular. It also seems likely that these issues lack sufficient definition in public discourse to generate consistent positions.
We asked how much those who favor a sharing license would be willing to pay.
The most frequently cited figure in this debate dates back to Fisher (2004), who estimated that a $5 monthly surcharge would cover what he estimated to be then-current music and film industry losses to file sharing. In our survey large groups in both countries indicated willingness pay over 15 dollars/euros for such an arrangement. The average proposed rates in the two countries were similar: $17.53 in the US and €16.26 in Germany. Median rates closely tracked the averages: $18.79 in the US and €16.43 in Germany.
As in other contexts, US and German P2P users reported a higher willingness to pay than their non-P2P-using peers. US P2P users proposed, on average, $20.28; German P2P users, €19.53.
Our results also show two peaks in the distribution of rates: one around 10–15 dollars/euros and a smaller peak around 20–25 dollars/euros. This clustering is consistent in the US and Germany and is roughly confirmed by other studies. It seems likely that these clusters reflect prices for existing commercial services: music services like Spotify, which typically hover around 10 dollars/euros, and video services like Netflix, which are generally priced in the range of 15–20 dollars/euros for streaming and DVDs.
Our evidence suggests, then, that there is considerable public willingness to pay to share files—and pay more than has been suggested in many earlier proposals. These sums are not trivial. In the US a $20 broadband fee, spread across the roughly 90 million current home subscriptions, would generate $21.6 billion—nearly as much as the home video and recorded music markets combined. In Germany a €15 broadband fee would generate €5 billion per year, nearly equaling 2011’s home video, recorded music, and pay TV revenues combined.
Sharing licenses raise a number of implementation challenges and face considerable industry opposition. In the US, legalized file sharing would compete with commercial distribution channels that generate far more revenue for intermediaries. Cable home video revenues in the US hit $57 billion in 2011 with an average bill of $86/month (not including Internet access, and for content further subsidized by advertising) (NPD 2012). Record companies and movie studios still talk about peak CD and DVD sales as natural revenue levels for those industries, recoverable through stronger enforcement. With so much money at stake in the current system, the major US content stakeholders are a blocking constituency for any major copyright reform. By the same token, after SOPA and ACTA it is no longer clear that these stakeholders have the political power to effectively lock down the distribution channels or raise intermediary liability high enough to exclude disruptive competition. As a result, in the US, revenue and access models are far more likely to change as a result of business interventions than political ones.
In Germany the barriers to legal change may be lower. The history of private copy rules, the greater role of the public media sector, the mounting abuses of the cease-and-desist industry, the greater willingness of political actors to protect privacy on the Internet, and—not least—the smaller size of industry stakeholders relative to other interests have the potential to change the political calculus.
The main goal of a sharing license, it is worth reiterating, is not to preserve the revenue levels of the current intermediaries (though it could, in theory, do so), but to ensure funding for a cultural ecosystem that accommodates the basic functions of computers and the Internet while preserving freedom of expression and privacy. Our work suggests that large numbers of people in both countries (and a majority in Germany) are willing to explore such legal alternatives. As policy makers threaten to discount other rights to strengthen copyright protection, all options should be on the table.
- Germany, 2010: In a survey by Wöbken (2010), 43% of German “web actives” (regular Internet users) agreed that “a monthly flat fee should be levied on all Internet accounts in Germany to benefit the authors of music, books and films because their works are presumably being copied on the Internet against their will. Illegal file-sharing would thereby be legalized.”
- Sweden, 2009: STIM (2009) found that “almost nine out of ten music users on the Internet—86.2% —would be interested in paying for a voluntary subscription legally entitling them to file share music.” Those who had the largest digital music collections (more than 5,000 songs) registered even stronger approval (88%).
- Canada, 2009: Angus Reid (2009) solicited responses to the following statement: “Some people propose that a levy be placed on Internet Service Providers . . . to help compensate musicians and others in the music industry for lost revenue from people using free file sharing services.” Only 27% agreed that this was “a worthwhile initiative”; 73% called it “an unnecessary and/or inappropriate fee that would end up being passed along to consumers.”
- UK, 2008: Bahanovich and Collopy (2008) asked 14- to 24-year-olds about their interest in “a file-sharing service where you could download any music in the world to own and keep.” Seventy-four percent supported the idea, including 80% of illegal file sharers. In their 2009 follow-up survey, 85% of P2P downloaders said that they would be interested in paying for an “unlimited, all-you-can-eat music download service.” In their 2011 follow-up 74% offered support.
- UK, 2010: In Wiggins’ 2010 Digital Media Survey, 59% put the “reasonable fee” for unlimited access to music and movies at around £3.50 ($5.00, €4.00); 25% declared willingness to pay up to £14.50 ($20.94, €16.96).
- Sweden, 2009: A STIM (2009) survey found that 52% of respondents would consider paying between SEK 50 and 150 ($6.59–$19.77, €4.71–€14.14) per month. One in five people—19%—would consider paying between SEK 150 and 300 (US$19.77–$39.54, €14.14–€28.29).
10. This approach has been discussed using names such as alternative compensation system, artistic freedom voucher, noncommercial use levy, licence globale, global Internet licensing agency, culture flat rate, global dissemination treaty, etc.
Artist Compensation On iTunes
Given the prominence of iTunes in setting prices for music, we also asked what percentage of an iTunes song sale people think goes to artists. The actual answer is more complicated than it seems. iTunes and Amazon.com take 30% of the retail price—about $0.30 on a $0.99 single—and pay the balance to labels and collecting societies—not directly to artists. How much of that ends up with the song author, lyricist, and performing artist depends on contracts with the labels and collecting society rules. The artist royalty from “average” major label contracts is hard to determine, but most sources put the percentage around 15% of the label’s share, or roughly 9% of the retail song price. Reportedly, superstar artists can command a 20%–25% royalty.11 For some indie labels and many of the new digital distribution services, artist royalties can range from 40% to 80%, or 28 to 56 cents on a dollar sale after Apple’s cut.
Germans did reasonably well on this question. A majority came within shooting distance of 9% (18% of respondents indicated that they did not know).
Americans did less well. Over 30% dramatically overestimated artist revenues; 36% came relatively close; 14% indicated that they did not know.
What Percentage Of An iTunes Song Sale Do You Think Goes To
The Artist? (US)
What Percentage Of An iTunes Song Sale Do You Think Goes To
The Artist? (Germany)
11. And megastars can apparently do even better. According to Billboard Magazine (Christman 2011), the Beatles’ long-delayed appearance on iTunes in 2011 was conditioned on recharacterizing downloads as a licensing arrangement rather than a retail sale, which triggers a more advantageous split for artists. As a result, the band is paid directly by iTunes, rather than by its label, EMI. This change has been sought by artists since the early days of digital sales and aligns with the ways Apple, Amazon, and other vendors characterize purchases as licenses to consumers. To date, the major labels have opposed the change.