The majority of academic literature has found that piracy has a negative impact on movie and music revenues.1 This is accepted by all but a few filesharing denialists. The big question is, what does this mean for copyright policy? Because it’s one thing to establish that online piracy reduces sales, it’s another to figure out what to do about it.

And there has been no shortage of those whose answer to what to do about piracy is: nothing. Or, rather, authors and creative industries need to learn how to “compete with free”, piracy enforcement doesn’t work.2 A litany of justifications for this have arisen over the past decade and a half, but generally, the idea seems to be that most unauthorized downloading or streaming of media comes from individuals who would not have paid in the first place. Thus, while anti-piracy efforts might reduce piracy, they won’t increase sales. Right?

Wrong, according to a growing body of empirical research. Last year, a group of researchers found that HADOPI, France’s recent graduated response program, led to a 20-25% increase in music sales in that country.3 And in 2011, researchers concluded that physical music sales increased 27% and digital music sales increased 48% in Sweden following copyright protection reform that increased the risks of unauthorized downloading.4

This week, Brett Danaher and Michael Smith, working at the Initiative for Digital Entertainment Analytics (IDEA) at Carnegie-Mellon University, have released another study looking at this question. The study, Gone in 60 Seconds: The Impact of the Megaupload Shutdown on Movie Sales, found that digital movie revenues from online sales and rentals increased by 6-10% following the January 2012 shutdown of the popular cyberlocker site (Megaupload execs, including Kim Dotcom, are of course currently facing criminal charges in the U.S. for copyright infringement).

The abstract says:

The growth of Internet-based piracy has led to a wide-ranging debate over how copyright policy should be enforced in the digital era. While some enforcement approaches involve policies designed to deter consumers from filesharing though incentives or penalties, other approaches target the supply of piracy by shutting down Internet sites that serve as major conduits for pirated content. In this paper we analyze how one such anti-piracy intervention, the shutdown of the popular Megaupload site, affected the digital sales of movies for two major studios.

Simply examining changes in sales after the shutdown would produce an inaccurate measure of its actual effect as sales are changing over time for a variety of reasons. Instead we exploit cross-country variation in pre-shutdown usage of Megaupload as a measure of treatment intensity. Controlling for country-specific trends and the Christmas holiday, we find no statistical relationship between Megaupload penetration and changes in digital sales prior to the shutdown. However, we find a statistically significant positive relationship between a country’s Megaupload penetration and its sales change after the shutdown, such that for each additional 1% pre-shutdown Megaupload penetration, the post-shutdown sales unit change was 2.5% to 3.8% higher, suggesting that these increases are a causal effect of the shutdown.

Aggregating these increases, our analysis across 12 countries suggests that, in the 18 weeks following the shutdown, digital revenues for these two studio’s movies were 6-10% higher than they would have been if not for the shutdown. Thus our findings show that the closing of a major online piracy site can increase digital media sales, and by extension we provide evidence that Internet movie piracy displaces digital film sales.

As the paper notes, not only does this research add to the empirical studies on this subject mentioned above, but it makes a couple new contributions. For one, Danaher and Smith note that this is the first paper to look at “supply-side” enforcement. The previous studies focused on “demand-side” enforcement — both the French and Swedish laws studied policed individual downloaders. Each has its pros and cons, its costs and benefits, but it seems to me that the trend in the U.S. and many other countries is to move away from demand-side enforcement and toward supply-side enforcement. For another, this is the first paper to measure digital movie sales. Both physical and digital sales of music have been measured numerous times since the late 90s, as well as DVD sales and box office receipts, but until now there has been no data on digital movie sales and rentals.

Perhaps most interestingly, Danaher and Smith were able to show that the shutdown of Megaupload did not just correlate with the subsequent increase in digital film revenues but actually caused it. Filesharing denialists are quick to discredit the numerous studies showing negative effects of piracy on sales by spouting that “correlation does not imply causation”, but by looking at the data country-by-country and comparing sales with Megaupload’s penetration rate in each country, the researchers were able to conclude that the shutdown “caused some customers to shift from cyberlocker-based piracy to purchasing or renting through legal digital channels.”

Also interesting is the discussion from Danaher and Smith at the end of the blog post announcing the study. The authors say, “While some have argued that you can’t compete with free, we think a more productive view is that competing with free (pirated) content is just a special case of price competition.” They note that convenience is sometimes more important that prices to consumers and explain that “we would expect that some consumers would be willing to buy through legitimate channels if content in those channels is more valuable than the ‘free’ pirated alternative.” But then they apply the results of their study to this line of thinking:

[W]e believe that another key part of competing with free piracy can be making content on illegal channels less valuable to consumers. In this regard, our finding of a 6-10% increase in digital movie revenue suggests that even though shutting down Megaupload didn’t stop all piracy, it was successful in making piracy sufficiently less reliable, less easy-to-use, and less convenient than it was before, and some consumers were willing to switch from piracy to legal channels as a result.

Taken this way, one might say that enforcing against piracy is necessary to ensure a vibrant, functional marketplace for expressive works online (though this alone is not sufficient). That’s not to say that the goal must be the complete elimination of piracy (that has never been the reality), nor is it to say that there is not a point where the costs of enforcement (monetary or otherwise) overtake the benefits. Efforts to mitigate piracy also do not need to be limited to law enforcement efforts or new legislation. They can include voluntary, market-based solutions aimed at education, like the Copyright Alert System that began operations on Monday, or reducing financial support of infringing sites, like the Statement of Best Practices to Address Online Piracy and Counterfeiting adopted last year by the Association of National Advertisers and American Association of Advertising Agencies.

But simply ignoring online infringement, and the harm it causes creators of all levels, forestalls the development of that vibrant marketplace and impedes the progress of the arts and sciences.


  1. See Smith & Telang, Assessing the Academic Literature Regarding the Impact of Media Piracy on Sales (Aug. 19, 2012), “Based on our review of the empirical literature we conclude that, while some papers in the literature find no evidence of harm, the vast majority of the literature (particularly the literature published in top peer reviewed journals) finds evidence that piracy harms media sales.” []
  2. See, for example, Joe Karaganis, Media Piracy in Emerging Economies (SSRC 2011). Note that this report’s conclusion, that “After a decade of ramped up enforcement, the authors can find no impact on the overall supply of pirated goods,” does not contradict the findings of the study discussed here. Even if true, it would seem that supply of pirated goods is a meaningless metric, especially where, as here, the goods are capable of infinite reproduction. []
  3. Danaher, Smith, Telang, and Chen, The Effect of Graduated Response Anti-Piracy Laws on Music Sales: Evidence from an Event Study in France, Journal of Industrial Economics (forthcoming). []
  4. Liang & Adermon, Piracy, Music, and Movies: A Natural Experiment, Working Paper, Uppsala University, Uppsala, Sweden (2011). []


  1. “…in the 18 weeks following the shutdown, digital revenues for these two studio’s movies were 6-10% higher than they would have been if not for the shutdown.”
    1) What were the specific dates for the sampling?
    2) What were the tentpole releases during that period for those two studios that weren’t available before the MegaUpload closing?
    3) How were those factors incorporated into the findings?
    4) How was the projected estimate for the period (that was beaten by 6-10%) conceived?
    5) What about after the “18 weeks”? Have the digital revenues remained consistently-high since?

    • I will leave the specifics to the researchers, however, from my experience researching piracy on the web, this finding isn’t surprising. Obviously Megaupload was a major conduit for pirated films, but beyond that, the takedown had a big impact on the rest of the black market cyberlocker culture. Several of the other big sites like Filesonic and Wupload disappeared once their business model was exposed for the criminal enterprise it was.

      The vacuum created by their absence gave legit outlets a fighting chance. No matter how you spin it, it’s hard to compete with free. Megaupload not only offered free content, but did so in a user-friendly way. Once they and their ilk were gone, legit, user-friendly options became more attractive to consumers.

      Glad this study was able to quantify what common sense and experience tells us to be true.

    • All these questions are answered in the report, which I have linked to in the article. It would probably be easiest for you to just read the report.

  2. I’m not familiar with Megaupload’s specific terms of service, but if they were like other cyberlockers, the content was only free to an extent. In addition to advertising, most cyberlockers make money from selling unthrottled access to the files so people could download them faster. If someone was getting enough material from Megaupload, they probably would have been paying for this unthrottled access. When Megaupload went down, the money they were spending there would have been freed up to spend elsewhere, such as on sites offering legitimate material.

    In that sense, from the consumer’s point of view, it was never free. Depending on how much someone downloaded, Megauploads unthrottled access fees might have been less than what they would have paid to get the material elsewhere, but the monetary cost to them was never zero. So switching to a legitimate source is less a case of suddenly having to start paying for their movies/songs/television shows/books/etc. and more a case of redirecting the money they were already spending.

    The big difference, of course, is that some of the money spent would go to the creators of the material rather than all of it going into Megaupload’s coffers and Kim Dotcom’s pockets.

  3. Pingback: Friday Endnotes – 03/15/13 | Copyhype