Piracy Website Blocking – Effective or Not?
My coauthors and I recently released a paper entitled “The Effect of Piracy Website Blocking on Consumer Behavior.” In it, we ask whether the strategy of having Internet Service Providers (ISPs) block access to websites that facilitate distribution of illegal content is effective in increasing traffic to legal sites for that content.
To determine this, we exploit two natural experiments in the UK – the court ordered blocking of The Pirate May in May 2012 and the subsequent court-ordered blocking of 19 major piracy sites in November 2013. Rather than simply looking at time trends (which are affected by many factors), we ask whether users of the blocked sites (the “treated group”) increase their consumption of legal content more than non-users of the blocked sites (the “control group”, since the blocks had no real impact on them).
Our results are interesting. We find that blocking The Pirate Bay had no causal impact on legal consumption. Rather, former Pirate Bay users either found other piracy websites or they turned to technologies that allowed them to circumvent the blocks, such as VPN’s. This finding is consistent with a recent study by Aguiar, Claussen, and Peukert, who found that shutting down a single piracy linking site in Germany did not meaningfully increase legal consumption there for the same reason.
In contrast, we find that the near simultaneous blocking of 19 major piracy sites in November 2013 caused users of those sites to increase their usage of legal video streaming sites like Netflix or Viewster by 12% (over and above the control group). The lightest users of the blocked sites only increase their legal consumption by 3.5% while the heaviest users of the blocked sites increased their legal consumption by 23.6%, which is exactly what one would expect if the blocks caused the increase.
The fact that the Pirate Bay block did not migrate consumers to legal channels while the 19 site block did suggests that only persistent blocking of multiple piracy sites can be effective in converting pirates to legal channels. The most likely explanation is that when you block one major site, it is easy for pirates to find their second favorite site. But when you block 19 major sites, not all pirates have a “20th favorite site” and the cost of finding another reliable site is prohibitive enough to convert some of those pirates. Other aspects of the data support this interpretation (read the paper to find out!).
These findings are especially timely in Australia, where the House and Senate have recently passed a controversial bill that will allow for piracy website blocking as a form of antipiracy enforcement. Opponents of the bill argue that such actions will be ineffective – our research suggests that these measures are only ineffective if they do not sufficiently increase the cost of piracy to the individual. If enough websites are blocked, our study implies that it could meaningfully increase legitimate consumption in Australia.
Of course, while our research is a step toward quantifying the potential benefits of piracy website blocking, it does not allow for a full cost-benefit analysis. First, we do now know how or if increased industry revenues from reduced piracy will change incentives to produce quality content. Second, we do not know the direct or indirect costs associated with website blocking activities.
That said, if this new bill is utilized in Australia, it may provide more opportunities to study the effect of website blocking (and it appears as if there are some data available on the cost of such activities there.)